---
mi_id: MIMR-node_b7410a0
type: market_report
canonical_url: "https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/market-report/global-food-authenticity-and-adulteration-testing-market/"
md_url: "https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/market-report/global-food-authenticity-and-adulteration-testing-market.md"

schema:
  "@type": Dataset
  "@id": MIMR-node_b7410a0
  url: "https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/market-report/global-food-authenticity-and-adulteration-testing-market/"
  title: "Global Food Authenticity & Adulteration Testing Market Report, Size, Share and Forecast 2026–2033: USD 1.65 Billion to USD 2.98 Billion at 7.67 Per Cent — the Fraudster Buys the Test, and the Reference Database Sets the Price"
  description: "Global food authenticity and adulteration testing market from USD 1,650.0m in 2025 to USD 2,980.0m in 2033 at 7.67%. No statutory limit, a reference database as the asset, and buyers on both sides of the fraud. Full TOC, chapters, visuals and sources."
  datePublished: "2026-11-21T12:39:41+00:00"
  dateModified: "2026-11-21T12:39:41+00:00"
  keywords:
    - food authenticity testing market
    - food adulteration testing
    - food fraud testing market
    - honey authenticity testing
    - olive oil adulteration
    - DNA barcoding food
    - NMR food profiling
    - EU Directive 2024/1438
  isPartOf:
    - id: MISG-node_392a29f
      type: CollectionPage
      url: "https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/"
      name: "Food Safety & Quality Testing"
    - id: MIIN-node_b477d05
      type: CollectionPage
      url: "https://www.pheonixresearch.com/food-beverages/"
      name: "Food & Beverages"
    - id: MIWS-root
      type: WebSite
      url: "https://www.pheonixresearch.com/"
      name: Pheonix Research
  mentions:
    - id: MIEN-node_3d4cc765
      type: Organization
      url: "https://www.eurofins.com"
      name: Eurofins Scientific
      sameAs:
        - "https://www.wikidata.org/wiki/Q324846"
    - id: MIEN-node_ce3c8e4d
      type: Organization
      url: "https://www.sgs.com"
      name: SGS
      sameAs:
        - "https://www.wikidata.org/wiki/Q669620"
    - id: MIEN-node_5f71ae50
      type: Organization
      url: "https://group.bureauveritas.com"
      name: Bureau Veritas
      sameAs:
        - "https://www.wikidata.org/wiki/Q1974503"
    - id: MIEN-node_978fb2c5
      type: Organization
      url: "https://www.intertek.com"
      name: Intertek
      sameAs:
        - "https://www.wikidata.org/wiki/Q1136085"
    - id: MIEN-node_6a568266
      type: Organization
      url: "https://www.alsglobal.com"
      name: ALS Limited
      sameAs:
        - "https://www.wikidata.org/wiki/Q5028003"
    - id: MIEN-node_d3493845
      type: Organization
      url: "https://www.merieuxnutrisciences.com"
      name: Mérieux NutriSciences
      sameAs:
        - "https://www.wikidata.org/wiki/Q21015817"
    - id: MIEN-node_0d0e9ee2
      type: Organization
      url: "https://www.thermofisher.com"
      name: Thermo Fisher Scientific
      sameAs:
        - "https://www.wikidata.org/wiki/Q643674"
    - id: MIEN-node_2b0c35e6
      type: Organization
      url: "https://www.bruker.com"
      name: Bruker
      sameAs:
        - "https://www.wikidata.org/wiki/Q431454"
    - id: MIEN-node_0cf454d4
      type: Organization
      url: "https://www.qiagen.com"
      name: QIAGEN
      sameAs:
        - "https://www.wikidata.org/wiki/Q315287"
    - id: MIEN-node_b9fff8bc
      type: Organization
      url: "https://www.agilent.com"
      name: Agilent Technologies
      sameAs:
        - "https://www.wikidata.org/wiki/Q393762"
    - id: MIMR-node_c16abcd
      type: Dataset
      url: "https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/market-report/global-food-safety-testing-market/"
      name: "Global Food Safety Testing Market Report, Size, Share and Forecast 2026–2033: Eleven Amendments in Thirty-Six Months, and Seventy-Five Per Cent of the Growth Is More Tests, Not Dearer Ones"
    - id: MIMR-node_304b4df
      type: Dataset
      url: "https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/market-report/global-food-allergen-testing-and-control-market/"
      name: "Global Food Allergen Testing & Control Market Report, Size, Share and Forecast 2026–2033: USD 1.18 Billion to USD 2.18 Billion at 7.97 Per Cent — a Test Costs Less in 2033 Than It Does Today, and Every Dollar of Growth Is More Tests"
    - id: MIMR-node_9b87119
      type: Dataset
      url: "https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/market-report/global-mycotoxin-testing-and-feed-binder-market/"
      name: "Global Mycotoxin Testing & Feed Binder Market Report, Size, Share and Forecast 2026–2033: USD 3.28 Billion to USD 5.80 Billion at 7.39 Per Cent — the Test Is Priced Like a Commodity and the Binder Is Priced Like Insurance"
    - id: MIMR-node_49f8147
      type: Dataset
      url: "https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/market-report/global-food-plant-hygiene-monitoring-and-atp-verification-market/"
      name: "Global Food Plant Hygiene Monitoring & ATP Verification Market Report, Size, Share and Forecast 2026–2033: USD 1.15 Billion to USD 1.91 Billion at 6.51 Per Cent — A Third of the Growth Is Price, Because the Pass Threshold Is the Only Food-Safety Standard Nobody Publishes"
    - id: MIMR-node_dce9ff4
      type: Dataset
      url: "https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/market-report/global-dehydrated-culture-media-and-microbiology-reference-materials-market/"
      name: "Global Dehydrated Culture Media & Microbiology Reference Materials Market Report, Size, Share and Forecast 2026–2033: USD 1.40 Billion to USD 2.30 Billion at 6.38 Per Cent — a Third of the Growth Is Price, and the Two Best-Informed Owners in the Category Have Just Reached Opposite Conclusions About It"
    - id: MIMR-node_6a178ef
      type: Dataset
      url: "https://www.pheonixresearch.com/food-beverages/oils-seeds-sweeteners/market-report/global-olive-oil-market/"
      name: Global Olive Oil Market size and share Analysis 2026-2033
    - id: MIMR-node_a8f6c16
      type: Dataset
      url: "https://www.pheonixresearch.com/food-beverages/spices-seasonings/market-report/global-saffron-market/"
      name: Global Saffron Market size and share Analysis 2026-2033
    - id: MIMR-node_d3a37cf
      type: Dataset
      url: "https://www.pheonixresearch.com/food-beverages/halal-meat/market-report/global-halal-market/"
      name: "Global Halal Market Report, Size, Share and Forecast 2026–2033"
    - id: MIMR-node_3e8d4da
      type: Dataset
      url: "https://www.pheonixresearch.com/food-beverages/beverages/market-report/global-ethanol-in-beverage-industry-market/"
      name: "Ethanol in Beverage Industry Market Size, Share and Forecast to 2033"
    - id: MIMR-node_079291f
      type: Dataset
      url: "https://www.pheonixresearch.com/food-beverages/oils-seeds-sweeteners/market-report/global-industrial-sugar-market/"
      name: Global Industrial Sugar Market size and share Analysis 2026-2033
  about:
    - "@type": Thing
      name: Market Intelligence
      mi_id: "MIMR-node_b7410a0#intelligence"
    - "https://www.wikidata.org/wiki/Q1901028"
  base_year: 2025
  forecast_year: 2033
  value_base_year: 1650.0
  value_forecast_year: 2980.0
  value_cagr: 7.67
  value_currency: USD
  value_unit_scale: Million

acf:
  base_year: 2025
  forecast_year: 2033
  value_base_year: 1650.0
  value_forecast_year: 2980.0
  value_cagr: 7.67
  value_currency: USD
  value_unit_scale: Million
  coverage_type: Global
  coverage_name: Global
  competitive_intensity_level: high
  market_structure_type: fragmented
  tier1_player_count: 6
  investment_trend_direction: rising
  capital_intensity_level: medium
  recent_mna_activity: yes
  overall_market_risk_level: moderate
  substitution_risk_level: high
  regulatory_complexity_level: high
  approval_pathway_structure: standardized_commercial
  innovation_intensity_level: high
  technology_maturity_stage: growth
  patent_activity_level: moderate
  supply_chain_complexity_level: high
  distribution_structure: hybrid
  primary_operational_model: hybrid
  executive_summary: "Food authenticity and adulteration testing is a USD 1,650.0 million market in 2025 reaching USD 2,980.0 million in 2033 at 7.67 per cent — the fastest growth in the food safety testing cluster and the lowest price capture in it. Volume runs 8.4 million to 14.6 million samples; realised price per sample moves USD 196.43 to USD 204.11, or 0.48 per cent a year. Of the USD 1,330.0 million added, USD 1,217.9 million is more samples, USD 157.7 million is each sample answering more declarable claims, and USD 45.5 million is lost to falling assay prices. The market has no statutory limit to anchor a price, a reference database rather than a published method as its real asset, and buyers on both sides of the fraud: the European Commission's own forensic investigation found operators using accredited laboratories to adapt honey and sugar blends so as to evade detection before import."
  toc: "<h2>Table of contents</h2>
<ol>
<li><a href=\"#section-main\">Executive summary — what the market is and what the number is</a></li>
<li><a href=\"#section-main\">How the value base was built, and what it excludes</a></li>
<li><a href=\"#section-main\">Volume, price and claim density: the three-way decomposition</a></li>
<li><a href=\"#section-main\">Why this is the fastest-growing segment in the cluster</a></li>
<li><a href=\"#section-vc\">The value chain: from composite sample to defensible verdict</a></li>
<li><a href=\"#section-vc\">Where the margin sits — and where it does not</a></li>
<li><a href=\"#section-vc\">The reference database is the asset, not the assay</a></li>
<li><a href=\"#section-cl\">Competitive landscape: the laboratory layer</a></li>
<li><a href=\"#section-cl\">Competitive landscape: the instrument, reagent and reference-material layer</a></li>
<li><a href=\"#section-cl\">The fraudster is a customer</a></li>
<li><a href=\"#section-ia\">Ownership, deal activity and what the disclosed anchors actually say</a></li>
<li><a href=\"#section-ti\">Technology: from targeted assay to fingerprint, and the deflation that comes with it</a></li>
<li><a href=\"#section-ti\">Three things that could break the database moat</a></li>
<li><a href=\"#section-mr\">Market risk: why the enforcement statistics point the other way</a></li>
<li><a href=\"#section-rl\">Regulation: the Breakfast Directive, FSMA 204 and the delegated acts of 2029</a></li>
</ol>"
  competitive_landscape_content: "<h2>Competitive landscape</h2>

<p>This market is fragmented in its service layer and concentrated in its capability layer, and the two are not the same firms. Six networks hold tier-one scale in food testing. Four instrument and reagent suppliers hold most of the analytical capability. One public body holds the reference materials. And one participant — the offender — is a customer.</p>

<h3>The laboratory layer</h3>

<p><a href=\"https://www.wikidata.org/wiki/Q324846\" target=\"_blank\" rel=\"noopener\">Eurofins Scientific</a> is the largest single presence. It reported revenue of EUR 7,296 million for 2025, up 5.0 per cent as reported and 3.7 per cent organically on its own reconciliation, with acquisitions contributing EUR 174 million and the fourth quarter the strongest of the year at 4.7 per cent organic. It is also the most explicit about authenticity as a discrete offer, running SNIF-NMR and oligosaccharide profiling for syrup extension, DNA-based species confirmation for meat and fish, isotope work for wine chaptalisation and geographical origin, and varietal and origin checks for coffee, tea and cereals. Its scale in authenticity is not disclosed as a line, which is the central evidentiary problem of this report.</p>

<p><a href=\"https://www.wikidata.org/wiki/Q669620\" target=\"_blank\" rel=\"noopener\">SGS</a> reported sales of CHF 6,945 million for 2025, up 5.6 per cent organically, with Testing & Inspection at CHF 6,165 million and 5.7 per cent organic growth. Its Health & Nutrition line reached CHF 882 million at 7.3 per cent organic with an adjusted operating margin of 14.1 per cent, and the disclosure inside it is the most useful single sentence in the public record: food delivered double-digit organic growth, driven by demand for contaminants, safety testing <em>and advisory services</em>. Advisory is named. In a market whose product is a defensible verdict rather than a pass or a fail, that is the tell.</p>

<p><a href=\"https://www.wikidata.org/wiki/Q1974503\" target=\"_blank\" rel=\"noopener\">Bureau Veritas</a> and <a href=\"https://www.wikidata.org/wiki/Q1136085\" target=\"_blank\" rel=\"noopener\">Intertek</a> compete on the same basis: broad food testing scope, global laboratory footprints, and authenticity sold as one application among many rather than as a flagship. <a href=\"https://www.wikidata.org/wiki/Q5028003\" target=\"_blank\" rel=\"noopener\">ALS Limited</a> brings a strong position in agricultural and food commodity testing, where authenticity questions arrive as contractual specifications rather than as consumer claims. <a href=\"https://www.wikidata.org/wiki/Q21015817\" target=\"_blank\" rel=\"noopener\">Mérieux NutriSciences</a> competes on the technical depth of its food science advisory alongside its analytical capacity, which is the position the market rewards.</p>

<p>None of the six breaks out authenticity revenue. That is not evasion; the segment is too small to be a reporting line inside a EUR 7 billion group, and it is sold inside broader food testing contracts. It does mean that the conventional market-share table cannot be built honestly from disclosure, and this report does not build one.</p>

<h3>The instrument, reagent and reference-material layer</h3>

<p><a href=\"https://www.wikidata.org/wiki/Q643674\" target=\"_blank\" rel=\"noopener\">Thermo Fisher Scientific</a> supplies both ends of the confirmatory stack — isotope-ratio mass spectrometry for origin and syrup detection, and real-time PCR for species — which makes it the single largest beneficiary of any rule that multiplies attributes per sample. <a href=\"https://www.wikidata.org/wiki/Q393762\" target=\"_blank\" rel=\"noopener\">Agilent Technologies</a> covers the chromatographic and mass-spectrometric determinations that underpin sterol, triglyceride and volatile profiling. <a href=\"https://www.wikidata.org/wiki/Q315287\" target=\"_blank\" rel=\"noopener\">QIAGEN</a> sits in the sample-preparation and PCR/NGS chemistry layer, where the reagent is the recurring revenue and the instrument is the lock-in.</p>

<p><a href=\"https://www.wikidata.org/wiki/Q431454\" target=\"_blank\" rel=\"noopener\">Bruker</a> is the most interesting position in the market and the clearest illustration of where the moat actually is. Its FoodScreener platform is a 400 MHz NMR system running push-button, fully automated analysis against matrix-specific databases, with profiling methods for honey, wine and juice. The honey database alone holds <strong>roughly 28,500 reference samples</strong> spanning more than 50 countries, 100 monofloral honey types, multiple seasons and production years, blends and production methods. That is not an instrument business. It is a data business that happens to sell the reader. Bruker's own framing — turning authenticity from after-the-fact detection into brand protection, one platform replacing many tests, catching even the fraud designed to pass conventional targeted tests — describes a product whose value is the reference set, and whose competitive defence is that no competitor can license 28,500 characterised samples quickly.</p>

<p>The public layer matters more than its commercial weight. The Joint Research Centre's certified reference materials for pollack and saithe, released in November 2024, are the state building the same asset in the public interest. If public reference materials proliferate across species and matrices, the private database premium compresses; if they do not, the database owners keep setting the price. That is the central technology risk in this market and it is set out in the technology chapter.</p>

<p>The referral layer completes the picture: brokers that aggregate accredited laboratories and resell determinations on a published rate card with a fixed service fee per order. They add no analytical capability and exist because the supply side is fragmented. Their growth is evidence of the fragmentation rather than a cause of it.</p>

<h3>The fraudster is a customer</h3>

<p>No competitor set in food safety contains this. The European Commission's forensic work on the honey cases — on-site inspection, sampling, and examination of computers and phone records, across 44 operators investigated and seven sanctioned — documented the use of accredited laboratories to adapt honey and sugar blends so as to elude detection by clients and official authorities before import operations. The offender bought analytical services to calibrate the fraud against the method.</p>

<p>Two consequences follow, and both are commercially significant. First, the addressable demand is larger than the compliant market: every enforcement advance creates a matching demand from the other side, and the only limit on it is the offender's risk calculus. Second, capability diffuses faster here than in any safety category, because the method is not merely published — it is sold. A supplier that launches a new screening method is simultaneously arming the customer who wants to pass it. That is a structurally different competitive dynamic from <a href=\"https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/market-report/global-food-allergen-testing-and-control-market/\">allergen testing</a>, where the absence of a reference method suppresses price, and from <a href=\"https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/market-report/global-dehydrated-culture-media-and-microbiology-reference-materials-market/\">culture media</a>, where a defined strain supports it.</p>

<p>The most exposed categories are the ones with the highest brand premium and the most complex supply chain. <a href=\"https://www.pheonixresearch.com/food-beverages/spices-seasonings/market-report/global-saffron-market/\">Saffron</a> is the textbook case — an 11 per cent suspicion rate in the EU coordinated plan despite being the most expensive spice by weight, because the gap between authentic and adulterated is large enough to fund sophisticated substitution. <a href=\"https://www.pheonixresearch.com/food-beverages/halal-meat/market-report/global-halal-market/\">Halal</a> certification adds a DNA-verifiable claim, porcine species absence, that sits entirely within this market's scope and is verified at the point of certification rather than at the point of import. <a href=\"https://www.pheonixresearch.com/food-beverages/beverages/market-report/global-ethanol-in-beverage-industry-market/\">Ethanol in beverages</a> carries chaptalisation and added-glycerol questions answerable only by isotope. And <a href=\"https://www.pheonixresearch.com/food-beverages/beverages/market-report/global-herbal-tea-market/\">herbal tea</a> is a botanical-identity problem before it is a safety one, where the declaration is species-level and the supply chain is long.</p>

<p>Where competitive intensity is genuinely high is not in winning contracts but in retaining the reference capability behind them. A laboratory that loses its isotope capability does not lose a service line; it loses the ability to adjudicate, and it becomes a screen. That is the position most of the mid-market occupies, and it is why the six tier-one networks keep investing in confirmatory capacity they cannot fully monetise.</p>"
  value_chain_content: "<h2>The value chain: from composite sample to defensible verdict</h2>

<p>The chain in this market is short and the margin is not where the sample is handled. Five commercial roles sit between a consignment and a decision, and only one of them is properly compensated.</p>

<p><strong>The first role is the consolidator or trader</strong>, which buys from producers across many origins and blends before the goods are labelled. This is where the fraud economically originates, because blending is the operation in which a cheap substitute becomes indistinguishable from an expensive original. The Commission's honey investigation found that more than half the operators involved — 57 per cent — had exported consignments suspicious of adulteration with extraneous sugars, and more than 60 per cent, 66 operators, had imported at least one such consignment. Fraud here is a procurement decision before it is an analytical problem.</p>

<p><strong>The second role is the brand owner or private-label retailer</strong>, which carries the legal exposure for the claim. It is the party that buys most authenticity testing, because it is the party that cannot argue ignorance. Eurofins describes its own authenticity service in exactly those terms, framing the offer as protecting the legacies of the largest food companies and covering ingredient verification, adulteration detection and geographical origin across agave syrup, honey and maple syrup, basmati rice, cereals, coffee, tea and spices, fish, flavours, fruit juice, meat, olive oil, spirits and wines. That list is not a product catalogue; it is a map of what a modern label asserts.</p>

<p><strong>The third role is the contract laboratory network</strong>, which performs the determinations. <strong>The fourth is the instrument, reagent and database layer</strong>, which supplies the capability. <strong>The fifth is the authority</strong>, which in the EU is the only actor whose result carries immediate legal weight, and which therefore buys a different product: surveillance rather than compliance.</p>

<p>Inside the analytical work there are four steps, and they have completely different economics.</p>

<p><strong>Sampling design</strong> decides everything and is almost never sold as a line item. A composite sample from a 20-tonne consignment of honey bought at the 2024 average third-country import price of EUR 1.92 per kilogram represents about EUR 38,400 of goods; the cheapest end of the published rate card for a species determination is EUR 150, which is 0.4 per cent of that consignment, and the dearest is EUR 550, which is 1.4 per cent. On Chinese honey at the 2024 average of EUR 1.20 per kilogram the same consignment is worth EUR 24,000 and the test is 0.6 to 2.3 per cent of it. The test is trivially affordable relative to the goods. What is not affordable is representativeness: one sample of a 20-tonne lot is one sample, and the Commission's own conclusion is that a simple test is not sufficient and that proper follow-up of suspicions is necessary. The binding constraint on this market is not the price of the assay. It is how many samples it takes to make one assay mean something, and that number is set by the customer's risk appetite, not by the laboratory's cost base.</p>

<p><strong>Screening</strong> is where volume lives. Fourier-transform infrared and Raman spectroscopy are the techniques the UK Food Standards Agency's 2025 review of edible oil authenticity singles out as showing promise for rapid, low-cost testing, while noting that many approaches are still under development and lack thorough validation. Screening is priced per sample at a level that makes wide coverage possible, and it is deliberately non-specific: its job is to sort samples into clear and suspicious, not to convict.</p>

<p><strong>Confirmation</strong> is where price lives. A suspicious screen goes to isotope-ratio mass spectrometry, to proton NMR profiling, to real-time PCR, or to a chromatographic determination of sterols and triacylglycerols. This is expensive, slower, and requires a fully equipped laboratory and a validated method. In edible oils the FSA review notes the increasing sophistication of fraud within the supply chain, with businesses relying on proprietary protocols — which hampers standardisation precisely because the protocol is the competitive asset.</p>

<p><strong>Adjudication</strong> is where margin lives. A confirmed analytical result is not yet a decision. It has to be placed against a definition of the product, a tolerance, and a traceability record, and that work is done by technical consultants, second-party auditors and, where it escalates, by forensic accountants and lawyers. It is the only step in the chain that scales with legal exposure rather than with sample count, and it is the reason the advisory layer is growing faster than the analytical layer.</p>

<p>The comparison with the safety side of this cluster is instructive. In <a href=\"https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/market-report/global-food-safety-testing-market/\">food safety testing</a> the verdict is generated by the method: a pathogen result against a published reference method is a pass or a fail, and no adjudication is required. In <a href=\"https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/market-report/global-mycotoxin-testing-and-feed-binder-market/\">mycotoxin testing</a> a statutory limit converts a concentration into a legal status in one step. Authenticity has no such conversion. Every positive result is an argument, and the argument must be won.</p>

<p>That is also why the intermediate layer exists at all. Brokerage of laboratory services — a single purchase order routed to whichever accredited laboratory holds the right scope, with a published service fee per order on top of the analytical price — is a real and growing part of the chain, because a brand owner with fifty authenticity questions across twelve matrices does not want fifty supplier relationships. The broker adds no analytical capability whatsoever and captures value purely from the fragmentation of the supply side, which tells you how fragmented it is.</p>

<p>Where the money finally settles is the reference layer. To determine that a fish is pollack rather than a cheaper substitute, a laboratory needs something to compare against; the Joint Research Centre responded in November 2024 by producing two certified reference materials, for pollack and saithe, as freeze-dried powders milled from single fillets and certified to species by sequencing and database comparison. The FSA's oils review reaches the same conclusion from the other direction, recommending that authentic certified reference materials are essential to support quality control and to encourage proficiency-testing uptake. Whoever holds the reference set holds the standard, and the standard is the only thing here that cannot be competed away on price.</p>

<p>One consequence deserves to be stated plainly. Because the reference layer sits upstream of the laboratory, the laboratory is not the most powerful actor in its own value chain. <a href=\"https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/\">This segment</a> is populated by very large testing networks that are, in the authenticity niche, price-takers on capability and price-makers only on service. The <a href=\"https://www.pheonixresearch.com/food-beverages/oils-seeds-sweeteners/market-report/global-olive-oil-market/\">olive oil</a> and <a href=\"https://www.pheonixresearch.com/food-beverages/oils-seeds-sweeteners/market-report/global-industrial-sugar-market/\">industrial sugar</a> chains illustrate the asymmetry: the first because its authenticity standard is a set of chemical parameters that anybody can run but only a reference panel can interpret, the second because it is the substrate that every syrup-extension fraud is made from, and the party selling the substrate has no interest in the test.</p>

<p>The value chain, in one sentence: the sample is a commodity, the confirmation is a service, the reference is a moat, and the verdict is a legal product sold by somebody who is not a laboratory.</p>"
  investment_activity_content: "<h2>Investment activity</h2>

<p>Capital in this market is spent on capability, not on capacity, and the distinction explains both the deal record and its silences.</p>

<p>A pathogen laboratory is a room with incubators, media and a plate reader; it scales by adding rooms and people. An authenticity laboratory is a room with a 400 MHz nuclear magnetic resonance spectrometer, an isotope-ratio mass spectrometer and a real-time PCR platform, and it scales by adding reference data. The first is a volume business with a linear cost curve. The second is a fixed-cost business with a research programme attached. That is why the six tier-one networks hold confirmatory authenticity capability out of proportion to the revenue it generates directly: it is the capability that lets them adjudicate rather than merely screen, and adjudication is what a client with a mislabelled consignment actually needs to buy.</p>

<p>The disclosed deal record is consistent with that reading and unhelpfully thin in exactly the place one would want it. <a href=\"https://www.wikidata.org/wiki/Q669620\" target=\"_blank\" rel=\"noopener\">SGS</a> acquired Murray-Brown Laboratories in the United States during 2025 — a testing laboratory specialising in food, ingredients and nutraceutical analysis, expanding its food safety and quality assurance presence in North America. That is the shape of the deals happening: capability tuck-ins inside an existing food footprint, bought for scope and geography rather than for an authenticity franchise. <a href=\"https://www.wikidata.org/wiki/Q324846\" target=\"_blank\" rel=\"noopener\">Eurofins</a> reported acquisitions contributing EUR 174 million of 2025 revenue, and stated that had those businesses been owned from 1 January the contribution would have been EUR 112 million higher — a programme of many small purchases rather than any single transformation.</p>

<p>The instrument layer is reallocating capital rather than consolidating it. <a href=\"https://www.wikidata.org/wiki/Q643674\" target=\"_blank\" rel=\"noopener\">Thermo Fisher</a> agreed in April 2026 to sell the microbiology business that had disclosed USD 645 million of 2025 revenue for approximately USD 1.075 billion, which is a portfolio decision about where instrument and consumable economics are best rewarded, not a retreat from food. <a href=\"https://www.wikidata.org/wiki/Q431454\" target=\"_blank\" rel=\"noopener\">Bruker</a> continues to invest in the one asset in this market that behaves like a durable franchise: the reference database behind its FoodScreener profiling methods, at roughly 28,500 characterised honey samples and growing.</p>

<p>What is absent from the record is as informative as what is in it. <strong>No authenticity-specific transaction discloses separable revenue, so this report states no transaction multiple.</strong> The reason is structural rather than a gap in our research: authenticity testing is sold inside broader food testing contracts, has no standalone reporting line at any of the six networks, and is therefore not separable in a purchase price allocation. Any multiple quoted for this niche would be a multiple for food testing generally, relabelled. We decline to do that.</p>

<p>Where the genuinely underinvested opportunity sits is the reference layer, and two public bodies have now said so in print. The Joint Research Centre manufactured certified reference materials for pollack and saithe — freeze-dried, cryogenically milled powders from single fillets, certified to species by DNA barcode sequencing and database comparison — explicitly so that laboratories using them can be confident their tests are accurate. The UK Food Standards Agency's 2025 review of edible oil authenticity concluded that authentic certified reference materials are essential to support quality control and encourage proficiency-testing uptake, and separately that the sector needs investment in widespread testing and point-of-use methods and in developing confirmatory techniques. Two regulators, two matrices, the same finding: the binding constraint is not instruments, it is characterised reference material and validated methods.</p>

<p>That is a product opportunity rather than a market-entry opportunity, and it is worth being precise about why. A certified reference material requires a characterisation programme, a certification body, an uncertainty budget and a stable supply of authentic starting material. It has no intellectual-property protection worth speaking of and it is priced by the unit. Nobody in this market is strongly incentivised to build a broad library of them, because the beneficiary is the whole laboratory population and the builder captures only the sale of the vials. That is a textbook public-good underprovision, and it is being filled slowly, matrix by matrix, by the JRC and by national reference laboratories.</p>

<p>The commercial consequence is a two-tier investment case. In the near term, capital flows to whoever can assemble private reference sets — a database business dressed as an instrument business — because that is where switching costs and pricing power actually live. In the medium term, if public reference materials proliferate across the matrices that matter (oils, species, botanicals, juices), the private premium compresses and the value migrates back to throughput, logistics and turnaround time. Which tier wins is a regulatory question, not a technology question, and it turns on how fast the delegated acts under Directive (EU) 2024/1438 require methods that only certified reference materials can support.</p>

<p>One further capital signal is worth recording because it is easy to miss. Part of the cost base of this market is borne by the state. Coordinated control plans in which 21 or 15 Member States sample and a single central laboratory analyses are a public subsidy to the analytical base: the JRC carries the method development and the reference work, and the Member States carry the sampling. That structure keeps the marginal cost of a surveillance sample low for authorities and correspondingly caps what a commercial laboratory can charge an authority for the same determination. It is one reason the enforcement segment of this market is the smallest and the slowest-growing of the five buyer groups.</p>"
  technology_innovation_content: "<h2>Technology and innovation</h2>
<p>The technology story in this market begins with a failure, and the failure is better documented than any success.</p>
<p>For two decades the workhorse method for detecting sugar syrup added to honey was stable carbon isotope ratio analysis by elemental analyser isotope-ratio mass spectrometry, AOAC method 991.41. It works because maize and sugarcane are C4 plants and their syrups carry a carbon isotope signature that honey, built from C3 nectar, does not. The Joint Research Centre’s 2023 study put that method against 320 imported consignments and reported that it <strong>was not effective in detecting honey suspicious of being adulterated</strong> — and drew the obvious conclusion: such syrups are no longer used to extend honey and have been replaced by syrups made mostly from rice, wheat or sugar beet. All three are C3 plants. The adulterant changed its isotope chemistry to defeat the reference method, and the reference method had no way to notice.</p>
<p>What replaced it was not a better single method but a panel. The JRC ran EA-IRMS and liquid chromatography IRMS, high-performance anion-exchange chromatography with pulsed amperometric detection, and proton nuclear magnetic resonance together, and treated the combination as the basis for suspicion. That is the structural fact about this market’s technology: <strong>authenticity is a multi-method problem because the adversary is optimising against each method separately.</strong> No single determination is dispositive, and the market therefore sells more determinations per sample than any other food testing category — the EU herbs and spices plan ran nearly 10,000 analyses on 1,885 samples, or about 5.3 analyses per sample.</p>
<figure><img src=\"https://www.pheonixresearch.com/wp-content/uploads/2026/11/figure-3-authenticity-technique-mix-scaled.webp\" alt=\"Donut chart of the 2025 food authenticity testing market by analytical technique: molecular 31 per cent, spectroscopic 21 per cent, chromatographic 19 per cent, isotopic 18 per cent, and sensory, microscopy and other 11 per cent.\" /><figcaption><strong>Figure 3.</strong> Technique mix, 2025 (USD 1,650.0 million). Source: Pheonix Research model.</figcaption></figure>

<p>The most commercially consequential technology is proton NMR profiling. <a href=\"https://www.wikidata.org/wiki/Q431454\" target=\"_blank\" rel=\"noopener\">Bruker</a>‘s FoodScreener runs at 400 MHz with fully automated, push-button analysis and reports against matrix-specific databases; the honey method detects exogenous sugars, verifies country of origin and botanical variety, assesses composition and freshness, flags atypical samples, and performs a regulatory compliance check for HMF, sucrose and glucose and fructose against EU and Codex criteria. The wine and juice methods run on the same platform. The reference database behind honey profiling holds roughly 28,500 samples across more than 50 countries and 100 monofloral types, spanning seasons, production years, blends and production methods, and the stated reason for that breadth is that the statistical models are only as good as the representativeness of what they were trained on.</p>
<p>That platform has two opposite economic effects at once, and separating them is the analytical core of this report.</p>
<p><strong>It is deflationary per sample.</strong> Bruker’s own proposition is that one platform replaces many tests and lowers the total cost of quality — a single push-button run against a database substituting for a series of targeted assays. When one run replaces three, revenue per sample falls even as information per sample rises. This is the mechanism behind the minus USD 45.5 million pure assay deflation term in our decomposition, and it is the reason realised price per sample grows at only 0.48 per cent a year while input costs rise faster.</p>
<p><strong>It is inflationary per claim.</strong> The same run produces determinations for origin, variety, purity, freshness and compositional compliance simultaneously. A customer who previously bought one answer now receives five, and the label that must be defended asserts five things. That is the claim-density term, worth USD 157.7 million of growth, and it is why the mix contribution accelerates from 2026 rather than growing smoothly: a legislative definition can turn a single-attribute claim into a multi-attribute obligation in one instrument.</p>
<p>The screening layer is where the next shift is likely and where the evidence is least settled. The FSA’s edible oils review found that spectroscopic methods — Fourier-transform infrared and Raman — show promise for rapid, low-cost testing, but also that many approaches remain under development and lack thorough validation, and that businesses rely on proprietary protocols, which hampers standardisation. A validated, cheap, point-of-use screen would move volume out of the laboratory and into the factory, which is good for coverage and bad for revenue per test. It would also make the confirmatory step the only defensible product, pushing the laboratory layer further toward adjudication.</p>
<p>Molecular methods are the most standardised leg of the tripod and therefore the least defensible commercially. Real-time PCR species determination is quoted at EUR 150–550 per sample with a detection limit of 0.1 per cent, two weeks’ turnaround and accreditation — a specification several hundred laboratories worldwide can meet. DNA barcoding gives the reference sequence, and the JRC’s pollack and saithe materials give the physical comparator. Next-generation sequencing extends this from presence to relative composition, which matters for products where a declared percentage is itself the claim. None of it is scarce. All of it is priced as a service.</p>
<p>Chromatographic determination of sterols, triglycerides and volatiles is the oldest and most legally entrenched layer, and it is where <a href=\"https://www.pheonixresearch.com/food-beverages/oils-seeds-sweeteners/market-report/global-olive-oil-market/\">olive oil</a> authenticity actually lives: a set of chemical parameters that any equipped laboratory can run. Its weakness is the one the FSA identifies — interpretation depends on the reference frame, so results are only as good as the panel or database behind them. Its strength is that it is written into standards, which makes it a compliance product rather than an investigative one.</p>
<p>Which brings the comparison that makes the cluster coherent. In <a href=\"https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/market-report/global-dehydrated-culture-media-and-microbiology-reference-materials-market/\">culture media and microbiology reference materials</a> the technology is stable and the reference strain is defined by a pharmacopoeia, so value sits in the definition and the market price-captures half its growth. In <a href=\"https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/market-report/global-food-plant-hygiene-monitoring-and-atp-verification-market/\">ATP and hygiene monitoring</a> the technology is weak and the threshold is unpublished, so value sits in the administrative switching cost. Here the technology is strong and improving and the definition is absent, so value sits in the reference set — and the reference set is the one asset an instrument vendor can own and a laboratory cannot easily copy.</p>
"
  market_risk_content: "<h2>Market risk</h2>
<p>Four risks matter here and three of them are not the ones usually listed.</p>
<h3>1. The demand signal went the wrong way in the most recent full year</h3>
<p>The intuitive demand driver for this market is fraud, and the most recent complete official dataset contradicts the intuition. The EU Alert and Cooperation Network’s 2024 annual report records <strong>9,460 notifications across the network, up 8 per cent</strong> on 2023, with RASFF rising 12 per cent to 5,250. Within that total, <strong>agri-food fraud notifications declined by one third</strong> and stood at 546, while plant health notifications almost tripled. Germany was the largest notifying country at 1,907, ahead of the Netherlands at 1,155 and Italy at 965.</p>
<p>There are three readings and we cannot choose between them from the outside. It may be enforcement reallocation: competent authorities in several Member States spent 2024 on plant health and animal welfare, both newly networked, rather than on fraud. It may be reporting behaviour: fraud cases increasingly routed through OLAF and Europol channels rather than through iRASFF, which would move the count without changing the activity. Or it may be genuine decline, which would be surprising given that OPSON XIV, run in the same period, seized goods worth EUR 95 million and reported 631 individuals to judicial authorities. What we can say is that <strong>a forecast resting on ‘fraud is increasing’ is resting on a statistic that fell by a third last year</strong>, and this one does not rest on it. The growth case here rests on claim density and on surveillance sampling rates, both of which are observable independently of reported fraud incidence.</p>
<figure><img src=\"https://www.pheonixresearch.com/wp-content/uploads/2026/11/figure-4-eu-notification-divergence-scaled.webp\" alt=\"Bar chart of EU Alert and Cooperation Network notification changes in 2024 against 2023: all network notifications up 8 per cent, RASFF up 12 per cent, and agri-food fraud notifications down 33 per cent.\" /><figcaption><strong>Figure 4.</strong> Change in EU notification volume, 2024 against 2023 (per cent). Source: EU Alert and Cooperation Network, 2024 Annual Report.</figcaption></figure>

<h3>2. The fastest-growing fraud typology is analytically invisible</h3>
<p>This is the sharper risk. Europol’s OPSON XIV outcome, published in October 2025, identified as the year’s main trend organised crime groups infiltrating waste disposal companies to reach expired food awaiting destruction, then <strong>removing the original best-before or expiry dates with solvents and printing new, falsified dates</strong>, describing the scale of the practice as unprecedented. It also records olive oil and wines with a protected designation of origin as remaining strongly affected by counterfeiting and designation abuse.</p>
<p>No assay in this market detects a reprinted date. That fraud is defeated by custody records, waste-manifest reconciliation and retail audit — documentation controls, not analytical ones. So the fastest-growing identified typology sits entirely outside the product this market sells, while the typologies it does address — composition, species, origin, variety — are the older ones. A market whose growth story depends on fraud getting worse should be worried that the newest and largest fraud trend has no test attached to it.</p>
<h3>3. Suspicion is not proof, and the asymmetry is expensive</h3>
<p>The Commission’s finding that current methods are not developed enough for a simple test to be sufficient, and that proper follow-up of suspicions is necessary, is a statement about the commercial product. 46 per cent <em>suspicious</em> is not 46 per cent non-compliant. In the honey action, 44 operators were investigated on the back of that testing and seven were sanctioned — a conversion rate of roughly one in six. That ratio is the single most useful number for understanding the economics of the analytical layer: most of what it sells is an indication, and the value is realised only after investigation.</p>
<p>The resulting asymmetry hurts both sides of the trade. A laboratory reporting a false positive against a named supplier carries contractual and reputational exposure; a laboratory reporting a false negative lets a fraud through and may never learn of it. The rational response is conservative decision thresholds, which produce more false positives, which make the result less useful and depress what a customer will pay for it. That dynamic — not instrument cost — is why price per determination is falling.</p>
<h3>4. Concentration in a handful of commodities, and the price of the authentic good</h3>
<p>Demand concentrates in honey, olive oil, herbs and spices, fish, juices and wine. The last risk in this chapter follows directly: <strong>fraud incentives are proportional to the price gap between authentic and substitute, so when the authentic price falls the market’s own demand driver softens.</strong> Third-country honey entering the EU in 2024 averaged EUR 1.92 per kilogram, 11.5 per cent lower than 2023, on volumes up 6 per cent to 172,367 tonnes, with Chinese honey at EUR 1.20 per kilogram. A cheaper authentic product narrows the gain from extension, and the narrowest-margin products in the category are precisely the ones where a multi-attribute determination looks least affordable to the buyer.</p>
<p>Olive oil runs the other way, and it is the useful contrast: a high-value, varietal, geographically designated product with an established chemical standard, which is why it recurs in every enforcement action and why it supports confirmatory testing. A portfolio of this market’s demand is therefore a bet on which commodities stay expensive and stay designated.</p>
<h3>Model risk and sensitivity</h3>
<p>The value base is constructed because no vendor reports authenticity as a line item; the band we would defend around the 2025 point is USD 1.45–1.85 billion. The sensitivity is published rather than buried: hold the realised price per sample flat at USD 196.43 and the market reaches <strong>USD 2,867.9 million</strong> in 2033 at the volume rate of 7.15 per cent; hold sample volume flat at 8.4 million and it reaches <strong>USD 1,714.5 million</strong> at the price rate of 0.48 per cent. The eight-year spread between those two answers is larger than most forecast disagreements in this cluster, and it is entirely a statement about whether authenticity becomes a required determination or stays a discretionary one.</p>
<p>A fifth risk is worth naming for completeness. Substitution risk is high, and it is internal to the customer rather than competitive: if point-of-use spectroscopy becomes validated and cheap, the screening layer moves into the factory and the outsourced market is left with confirmation only. That is a smaller market per sample and a defensible one per determination, and it is the most likely shape of 2033.</p>
"
  regulatory_landscape_content: "<h2>Regulatory landscape</h2>

<p>This is the only segment in the food safety cluster where regulation creates the obligation without creating the limit, and understanding that gap is the whole of the price story.</p>

<h3>The Breakfast Directive</h3>

<p>Directive (EU) 2024/1438 of 14 May 2024 amends four marketing directives at once — honey (2001/110/EC), fruit juices (2001/112/EC), jams, jellies and marmalades (2001/113/EC) and certain preserved milks (2001/114/EC) — which is why it is known as the Breakfast Directive. It was published in the Official Journal on 24 May 2024, had to be transposed into national law by <strong>14 December 2025</strong>, and its rules apply from <strong>14 June 2026</strong>. France transposed by Decree No. 2026-312 of 24 April 2026, in force on the same date.</p>

<p>For honey, the operative change is origin. The country or countries of origin where the honey was harvested must be indicated on the label, and where honey originates in more than one country the countries must be shown <strong>in the principal field of vision, in descending order of their share by weight, together with the percentage each represents, subject to a 5 per cent tolerance per share</strong>, calculated on the basis of the operator's traceability documentation. Where a blend draws on more than four countries and the four largest shares exceed 50 per cent, Member States may allow percentages for those four only, with the remainder listed in descending order and unquantified. Packages with a net quantity under 30 grammes may substitute two-letter ISO 3166-1 codes. The particulars become mandatory particulars under Regulation (EU) No 1169/2011, which means the general food labelling enforcement machinery attaches to them.</p>

<p>Two features of that provision drive demand. First, it converts origin from a category into a <strong>quantity</strong>: a label that once said 'blend of EU and non-EU honeys' now asserts a set of percentages. A percentage is falsifiable in a way a category is not, and proving a share to within five points is harder than proving a country, because it requires the determination to be quantitative across every component rather than merely consistent with the declared origin. Second, it moves the burden of proof onto the operator. For the first time in this category the producer, not the authority, has to be able to demonstrate something about origin on demand — and the reference point for that demonstration is the operator's own traceability documentation, which is precisely what the fraud cases showed being fabricated.</p>

<p>What the directive does <em>not</em> yet do is equally important. Article 4a empowers the Commission to adopt delegated acts on five things: the criterion of 'mainly' as regards floral or vegetable origin; composition criteria to ensure honey has not been heated or treated so as to destroy or significantly deactivate natural enzymes, by reference to the invertase index; criteria to ensure and verify that <strong>pollen is not removed</strong> from honey and that absolute pollen content and pollen spectrum are not modified; minimum pollen content in baker's honey; and the methods and criteria to determine where honey was harvested plus Union-wide traceability requirements. Those delegated acts are due by <strong>14 June 2029</strong>, preceded by feasibility studies, one of which must analyse digital solutions including a unique identifier code.</p>

<p>That is the second wave, and it is the event that would change the price structure of this market. A pollen-content criterion and an invertase index are <strong>numerical limits</strong>. A numerical limit is the one thing this market has never had, and the cluster evidence is unambiguous about what a numerical limit does to price capture: the categories where a threshold is published capture a third to a half of their growth in price, and the category where no threshold exists gives price back. If the 2029 delegated acts land as numbers, authenticity testing stops being an investigative service and starts being a compliance product — and the 2033 figure would be reached through price rather than volume.</p>

<h3>The United States chose documentation instead</h3>

<p>The contrast is stark and it explains the geography of this market. The FDA's Food Traceability Final Rule under Section 204(d) of the Food Safety Modernization Act requires persons manufacturing, processing, packing or holding foods on the Food Traceability List to maintain records of Key Data Elements at Critical Tracking Events and to produce them to the FDA within 24 hours. The original compliance date was 20 January 2026. The FDA proposed to extend it by 30 months to <strong>20 July 2028</strong>, and the Continuing Appropriations, Agriculture, Legislative Branch, Military Construction and Veterans Affairs, and Extensions Act of 2026 directed the agency not to enforce the rule before that date; the FDA has said it intends to comply with that directive. It also convened a public meeting on lot-level traceability on 6 November 2026.</p>

<p>FSMA 204 is a records rule, not an analytical one. It answers 'where did this lot come from and where did it go' — the question that defeats the relabelled-expiry fraud OPSON identified — and it does not answer 'is this what it says it is'. Both are legitimate regulatory strategies and they purchase different things. The EU is buying determinations; the United States is buying provenance. For the purposes of this report the consequence is that growth is EU-weighted, and we state that as a qualitative reading rather than modelling a regional split, because the model carries no geography.</p>

<h3>The enforcement architecture that generates the demand</h3>

<p>Three mechanisms matter. The EU Agri-Food Fraud Network sits inside the Administrative Assistance and Cooperation network and handles cross-border fraud cases through iRASFF. Coordinated control plans — the honey action across 15 Member States plus Switzerland and Norway, the herbs and spices action across 21 Member States plus Norway and Switzerland — are the mechanism by which the state buys analytical work at scale, with sampling done nationally and analysis done centrally at the Joint Research Centre. And OPSON, coordinated by Europol with OLAF and DG SANTE, ran its fourteenth edition across 31 countries in 2025 with 31,165 checks, 259,012 packages, 1.4 million litres of beverages and 11,566 tonnes of food seized, goods worth about EUR 95 million, 631 individuals reported to judicial authorities, 101 arrest warrants and 13 organised crime groups disrupted.</p>

<p>None of those numbers converts into a purchase order for an authenticity determination, which is the point. Enforcement activity is the market's marketing, not its revenue.</p>

<h3>Accreditation is the real barrier to entry</h3>

<p>The de facto licence to operate is accreditation to ISO/IEC 17025 for the specific method, and it is what separates a laboratory selling a determination from one selling an opinion. It is visible in the rate cards: the manuka pollen DNA method is quoted as accredited to ISO 17025, and the PCR species determination is sold as an accredited test method with a stated detection limit. That gives this market a standardised commercial pathway — there is no approval to seek, no licence to obtain, no product to register. The barrier is capability plus accreditation, which is why the market is fragmented among many accredited laboratories and concentrated only in the confirmatory layer where the capital cost bites.</p>

<p>The gap that no regulation fills is the one identified earlier in this report: there is no statutory reference method for authenticity anywhere, and no general legal definition of what makes a food authentic beyond not being misdescribed. Directive 2024/1438 narrows that gap for four breakfast categories from 2026 and for composition and pollen from 2029. Until then, every determination in this market is an expert judgement dressed as a measurement, which is why it is worth what it is worth.</p>"
  faq_question_1: "How large is the global food authenticity and adulteration testing market?"
  faq_answer_1: "USD 1,650.0 million in 2025, rising to USD 2,980.0 million in 2033 at 7.67 per cent compound. That covers laboratory analytical revenue for authenticity-specific determinations, instrument and consumable revenue attributable to authenticity workflows, and the data, surveillance and advisory layer. No company discloses authenticity testing as a standalone line, so the base is a constructed estimate; the band we would defend is USD 1.45–1.85 billion around the 2025 point."
  faq_question_2: "Why does this segment grow faster than food safety testing overall?"
  faq_answer_2: "The parent market grows 5.93 per cent to 2033; this one grows 7.67. The difference is what the unit of demand is tied to. Safety testing volume is tied to batches produced and to statutory limits, and it price-captures 25 per cent of its growth because a published reference method makes reagents comparable. Authenticity volume is tied to the number of declarable claims a product carries and to the number of consignments sampled, and neither has a ceiling set by regulation."
  faq_question_3: "Does the EU Honey Directive create new testing demand from 2026?"
  faq_answer_3: "Partly, and the timing is precise. Directive (EU) 2024/1438 had to be transposed by 14 December 2025 and its rules apply from 14 June 2026: the countries of origin of a blend must be declared in descending order of share, with the percentage each represents and a 5 per cent tolerance per share, calculated against the operator's traceability documentation. That shifts the burden of proof onto the operator for origin. It does not yet create a compositional limit: the delegated acts on composition criteria, pollen integrity and Union-wide traceability are not due until 14 June 2029."
  faq_question_4: "What is the biggest limitation of this forecast?"
  faq_answer_4: "Two things. The value base is constructed because no vendor reports authenticity as a line item, so the absolute level is arguable even where the growth rate is not. And the demand signal points the other way: EU agri-food fraud notifications fell by one third in 2024 to 546 while total network notifications rose 8 per cent, so a forecast resting on 'fraud is increasing' is resting on a statistic that declined in the most recent full year. The growth case here rests on claim density and surveillance sampling, not on fraud incidence."

global_schema:
  organization:
    "@id": MIWS-root#organization
    "@type": Organization
    url: "https://www.pheonixresearch.com/"
    name: Pheonix Research
  website:
    "@id": MIWS-root
    "@type": WebSite
    url: "https://www.pheonixresearch.com/"
    name: Pheonix Research
  api:
    "@id": MIWS-root#api
    "@type": WebAPI
    url: "https://www.pheonixresearch.com/"

graph:
  node_id: MIMR-node_b7410a0
  graph_node_endpoint: "https://graph.pheonixresearch.com/api/v1/query/live/node/MIMR-node_b7410a0"
  graph_snapshot: "https://graph.pheonixresearch.com/api/v1/read/live/graph"
  graph_exclusions: "https://graph.pheonixresearch.com/api/v1/read/live/exclusions"
  graph_meta: "https://graph.pheonixresearch.com/api/v1/read/live/meta"
  graph_bundle: "https://graph.pheonixresearch.com/api/v1/read/live/bundle"

discovery:
  discovery_json: "https://www.pheonixresearch.com/.well-known/pheonix-discovery.json"
  llms_txt: "https://www.pheonixresearch.com/llms.txt"
  sitemap: "https://www.pheonixresearch.com/sitemap.xml"
---
# Global Food Authenticity & Adulteration Testing Market Report, Size, Share and Forecast 2026–2033: USD 1.65 Billion to USD 2.98 Billion at 7.67 Per Cent — the Fraudster Buys the Test, and the Reference Database Sets the Price

- **Market Size (Base Year):** USD 1650.0 Million
- **Forecast Value:** USD 2980.0 Million
- **CAGR:** 7.67%
- **Forecast Period:** 2026 - 2033
- **Coverage:** Global - Global

## Executive Summary

Food authenticity and adulteration testing is a USD 1,650.0 million market in 2025 reaching USD 2,980.0 million in 2033 at 7.67 per cent — the fastest growth in the food safety testing cluster and the lowest price capture in it. Volume runs 8.4 million to 14.6 million samples; realised price per sample moves USD 196.43 to USD 204.11, or 0.48 per cent a year. Of the USD 1,330.0 million added, USD 1,217.9 million is more samples, USD 157.7 million is each sample answering more declarable claims, and USD 45.5 million is lost to falling assay prices. The market has no statutory limit to anchor a price, a reference database rather than a published method as its real asset, and buyers on both sides of the fraud: the European Commission's own forensic investigation found operators using accredited laboratories to adapt honey and sugar blends so as to evade detection before import.

## Market Report

## Executive summary

Food authenticity and adulteration testing is a **USD 1,650.0 million** market in 2025, measured as the analytical, instrument and advisory revenue attributable to verifying that a food is what it claims to be. It reaches **USD 2,980.0 million** in 2033 — **7.67 per cent compound**. Volume runs from **8.4 million samples to 14.6 million** at 7.15 per cent. The realised price per sample moves from **USD 196.43 to USD 204.11**, which is 0.48 per cent a year and, in a market where input costs and wages rise faster than that, is a price decline in real terms.

Of the **USD 1,330.0 million** the market adds, **USD 1,217.9 million (91.6 per cent) is more samples**, **USD 157.7 million (11.9 per cent) is each sample answering more declarable claims**, and **minus USD 45.5 million (−3.4 per cent)** is falling assay prices. That combination — almost all volume, a real but secondary mix contribution, and outright deflation underneath — makes this the least price-powered market in the Food Safety & Quality Testing cluster and the fastest-growing one in it.

**Figure 1.** Global food authenticity and adulteration testing market value, 2025–2033 (USD million). Source: Pheonix Research model.

 | Report |  | 2025 (USD bn) |  | 2033 (USD bn) |  | CAGR |  | Share of growth from price and mix |

 | [Food Safety Testing](https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/market-report/global-food-safety-testing-market/) |  | 25.80 |  | 40.90 |  | 5.93% |  | 25.1% |

 | [Food Allergen Testing & Control](https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/market-report/global-food-allergen-testing-and-control-market/) |  | 1.18 |  | 2.18 |  | 7.97% |  | −1.5% |

 | [Mycotoxin Testing & Feed Binders](https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/market-report/global-mycotoxin-testing-and-feed-binder-market/) |  | 3.28 |  | 5.80 |  | 7.39% |  | 28.3% |

 | [Plant Hygiene Monitoring & ATP](https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/market-report/global-food-plant-hygiene-monitoring-and-atp-verification-market/) |  | 1.15 |  | 1.91 |  | 6.51% |  | 33.1% |

 | [Dehydrated Culture Media](https://www.pheonixresearch.com/food-beverages/food-safety-quality-testing/market-report/global-dehydrated-culture-media-and-microbiology-reference-materials-market/) |  | 1.40 |  | 2.30 |  | 6.38% |  | 52.5% |

 | **Food Authenticity & Adulteration Testing** |  | **1.65** |  | **2.98** |  | **7.67%** |  | **8.5%** |

**Figure 2.** Market value by claim type, 2025 and 2033 (USD million). Varietal and botanical claims compound at 9.14 per cent, process claims at 5.93. Source: Pheonix Research model.

The order of that table is not an accident. Price capture in this cluster is high where a regulator publishes a number and low where it does not. Culture media price-captures 52.5 per cent of its growth because a pharmacopoeia defines the strain and a validation obligation per batch of medium makes switching expensive. Plant hygiene and ATP captures 33.1 per cent because the pass threshold is invented by the plant, frozen into its HACCP record, and never published by anyone. Allergen testing captures nothing — it actually gives back 1.5 per cent — because no statutory reference method exists anywhere for allergen detection, so reagents compete on price and nothing else.

Authenticity testing sits at the far end of that logic. There is no statutory limit for authenticity at all. A food is not required to contain a minimum quantity of anything in order to be itself; it is required only not to be misdescribed. That single difference means no vendor can sell compliance, no test result has automatic legal force, and the buyer is not purchasing a pass or a fail. The buyer is purchasing a *suspicion* — and then purchasing the work of turning that suspicion into something defensible.

The scale of the suspicion is documented. The European Commission's coordinated control plan on imported honey took 320 consignments originating in 20 exporting countries, sampled by 15 Member States plus Switzerland and Norway between October 2021 and February 2022, and the Joint Research Centre found **147 of them — 46 per cent — suspicious** of non-compliance with the Honey Directive's prohibition on added ingredients. Against the same kind of exercise in 2015–17, the rate was 14 per cent. The absolute count was highest for China at 66 of 89 consignments; the relative rate was highest for Turkey at 14 of 15, and for honey arriving from the United Kingdom at 10 of 10, which the Commission attributes to honey produced elsewhere and blended in the UK before re-export. The Commission's coordinated plan on culinary herbs and spices analysed 1,885 samples across 21 Member States plus Norway and Switzerland using nearly 10,000 analyses and found 17 per cent suspicious — but **48 per cent of oregano**, adulterated in most cases with olive leaves.

And then the caveat, in the Commission's own words, which is the single most important sentence in this market: the current analytical methods are *not yet developed enough at international level and a simple test is not sufficient to detect adulteration*; a proper follow-up of suspicions is necessary. That is the ceiling on the market and the reason for its shape. A test that produces a suspicion is worth less than a test that produces a verdict, the follow-up work is where the labour goes, and the price the market can charge is set by how far each method is from being conclusive.

Which brings the second fact, and it is the one that makes this market unlike any other in food safety. The Commission's forensic investigation into the honey cases — on-site inspection, sampling, and examination of computers and phone records across 44 EU operators, of whom seven were sanctioned — recorded as a modus operandi the **use of accredited laboratories to adapt honey and sugar blends so as to elude possible detection by clients and official authorities before import operations**. In no other food-safety category does the offender buy the test. Here the adversary is a customer, the method is the thing being defeated, and every advance in detection creates demand for the next blend. That is why the demand curve does not behave like a compliance market's.

The third fact is what sets the invoice. A determination is priced by how many attributes the claim requires, not by the cost of the assay. A species determination by real-time PCR is quoted at **EUR 150–550 per sample**, accredited, with a detection limit of 0.1 per cent and about two weeks' turnaround. Authentication of manuka honey under the New Zealand Ministry for Primary Industries definition requires five attributes — four chemical and one DNA marker from manuka pollen — and one United Kingdom laboratory prices that DNA marker alone at **GBP 246 per sample** in a five-sample pack. Same sample, same laboratory, one to five times the invoice, depending entirely on how many things the label asserts. That is the mechanism behind the 11.9 per cent claim-density term, and it is the reason the mix contribution is back-loaded: a legislative definition can multiply attributes per sample overnight, and one is about to.

On **14 June 2026** the rules of Directive (EU) 2024/1438 begin to apply. The directive had to be transposed by 14 December 2025, and it amends the honey, fruit juice, jam and preserved milk directives together. For honey it requires the countries of origin of a blend to be declared on the label in the principal field of vision, in **descending order of their share by weight, with the percentage each represents and a 5 per cent tolerance per share**, calculated against the operator's traceability documentation. Where a blend draws on more than four countries and the four largest exceed 50 per cent, Member States may permit percentages for those four only. Packages below 30 grammes may use two-letter ISO country codes. France transposed by Decree No. 2026-312 of 24 April 2026, in force on 14 June 2026. What that does is move the burden of evidential proof onto the operator for origin — not composition, not pollen, not traceability, all of which wait on delegated acts not due until **14 June 2029**.

How the market is sized. No company reports authenticity testing as a line item. The USD 1,650.0 million base is built from three layers: laboratory revenue for authenticity-specific determinations, derived from published rate cards and the installed base of accredited methods across the contract laboratory networks; instrument and consumable revenue attributable to authenticity workflows, being the fraction of NMR, isotope-ratio mass spectrometry, PCR and chromatographic placements and reagent consumption used for claim verification rather than safety; and the data, surveillance and advisory layer, which is the fastest-growing and least visible. The band we would defend around the 2025 point is **USD 1.45–1.85 billion**. The growth rate is firmer than the level: the volume line is anchored on consignments sampled and declarable claims carried, both of which are observable, and the price line is anchored on published rate cards rather than on revenue per test from companies that do not disclose it.

What the report refuses. It carries no regional split, because the model has no geography in it and the honest reading — that growth is EU-weighted because the EU is the only jurisdiction creating attribute-level analytical obligations — is a qualitative judgement, stated as one. It carries no market-share table, because no disclosed line maps onto this market's perimeter. It carries no transaction multiple, because no authenticity-specific deal discloses separable revenue. And it makes no claim about the incidence of fraud: the most recent full year of EU data shows agri-food fraud notifications *falling* by one third to 546 while total network notifications rose 8 per cent, and a forecast that leans on rising fraud is leaning on a statistic that went the wrong way.

## Table of Contents

Table of contents

Executive summary — what the market is and what the number is
How the value base was built, and what it excludes
Volume, price and claim density: the three-way decomposition
Why this is the fastest-growing segment in the cluster
The value chain: from composite sample to defensible verdict
Where the margin sits — and where it does not
The reference database is the asset, not the assay
Competitive landscape: the laboratory layer
Competitive landscape: the instrument, reagent and reference-material layer
The fraudster is a customer
Ownership, deal activity and what the disclosed anchors actually say
Technology: from targeted assay to fingerprint, and the deflation that comes with it
Three things that could break the database moat
Market risk: why the enforcement statistics point the other way
Regulation: the Breakfast Directive, FSMA 204 and the delegated acts of 2029

## Competitive Landscape

Competitive landscape
This market is fragmented in its service layer and concentrated in its capability layer, and the two are not the same firms. Six networks hold tier-one scale in food testing. Four instrument and reagent suppliers hold most of the analytical capability. One public body holds the reference materials. And one participant — the offender — is a customer.
The laboratory layer
Eurofins Scientific is the largest single presence. It reported revenue of EUR 7,296 million for 2025, up 5.0 per cent as reported and 3.7 per cent organically on its own reconciliation, with acquisitions contributing EUR 174 million and the fourth quarter the strongest of the year at 4.7 per cent organic. It is also the most explicit about authenticity as a discrete offer, running SNIF-NMR and oligosaccharide profiling for syrup extension, DNA-based species confirmation for meat and fish, isotope work for wine chaptalisation and geographical origin, and varietal and origin checks for coffee, tea and cereals. Its scale in authenticity is not disclosed as a line, which is the central evidentiary problem of this report.
SGS reported sales of CHF 6,945 million for 2025, up 5.6 per cent organically, with Testing & Inspection at CHF 6,165 million and 5.7 per cent organic growth. Its Health & Nutrition line reached CHF 882 million at 7.3 per cent organic with an adjusted operating margin of 14.1 per cent, and the disclosure inside it is the most useful single sentence in the public record: food delivered double-digit organic growth, driven by demand for contaminants, safety testing and advisory services. Advisory is named. In a market whose product is a defensible verdict rather than a pass or a fail, that is the tell.
Bureau Veritas and Intertek compete on the same basis: broad food testing scope, global laboratory footprints, and authenticity sold as one application among many rather than as a flagship. ALS Limited brings a strong position in agricultural and food commodity testing, where authenticity questions arrive as contractual specifications rather than as consumer claims. Mérieux NutriSciences competes on the technical depth of its food science advisory alongside its analytical capacity, which is the position the market rewards.
None of the six breaks out authenticity revenue. That is not evasion; the segment is too small to be a reporting line inside a EUR 7 billion group, and it is sold inside broader food testing contracts. It does mean that the conventional market-share table cannot be built honestly from disclosure, and this report does not build one.
The instrument, reagent and reference-material layer
Thermo Fisher Scientific supplies both ends of the confirmatory stack — isotope-ratio mass spectrometry for origin and syrup detection, and real-time PCR for species — which makes it the single largest beneficiary of any rule that multiplies attributes per sample. Agilent Technologies covers the chromatographic and mass-spectrometric determinations that underpin sterol, triglyceride and volatile profiling. QIAGEN sits in the sample-preparation and PCR/NGS chemistry layer, where the reagent is the recurring revenue and the instrument is the lock-in.
Bruker is the most interesting position in the market and the clearest illustration of where the moat actually is. Its FoodScreener platform is a 400 MHz NMR system running push-button, fully automated analysis against matrix-specific databases, with profiling methods for honey, wine and juice. The honey database alone holds roughly 28,500 reference samples spanning more than 50 countries, 100 monofloral honey types, multiple seasons and production years, blends and production methods. That is not an instrument business. It is a data business that happens to sell the reader. Bruker’s own framing — turning authenticity from after-the-fact detection into brand protection, one platform replacing many tests, catching even the fraud designed to pass conventional targeted tests — describes a product whose value is the reference set, and whose competitive defence is that no competitor can license 28,500 characterised samples quickly.
The public layer matters more than its commercial weight. The Joint Research Centre’s certified reference materials for pollack and saithe, released in November 2024, are the state building the same asset in the public interest. If public reference materials proliferate across species and matrices, the private database premium compresses; if they do not, the database owners keep setting the price. That is the central technology risk in this market and it is set out in the technology chapter.
The referral layer completes the picture: brokers that aggregate accredited laboratories and resell determinations on a published rate card with a fixed service fee per order. They add no analytical capability and exist because the supply side is fragmented. Their growth is evidence of the fragmentation rather than a cause of it.
The fraudster is a customer
No competitor set in food safety contains this. The European Commission’s forensic work on the honey cases — on-site inspection, sampling, and examination of computers and phone records, across 44 operators investigated and seven sanctioned — documented the use of accredited laboratories to adapt honey and sugar blends so as to elude detection by clients and official authorities before import operations. The offender bought analytical services to calibrate the fraud against the method.
Two consequences follow, and both are commercially significant. First, the addressable demand is larger than the compliant market: every enforcement advance creates a matching demand from the other side, and the only limit on it is the offender’s risk calculus. Second, capability diffuses faster here than in any safety category, because the method is not merely published — it is sold. A supplier that launches a new screening method is simultaneously arming the customer who wants to pass it. That is a structurally different competitive dynamic from allergen testing, where the absence of a reference method suppresses price, and from culture media, where a defined strain supports it.
The most exposed categories are the ones with the highest brand premium and the most complex supply chain. Saffron is the textbook case — an 11 per cent suspicion rate in the EU coordinated plan despite being the most expensive spice by weight, because the gap between authentic and adulterated is large enough to fund sophisticated substitution. Halal certification adds a DNA-verifiable claim, porcine species absence, that sits entirely within this market’s scope and is verified at the point of certification rather than at the point of import. Ethanol in beverages carries chaptalisation and added-glycerol questions answerable only by isotope. And herbal tea is a botanical-identity problem before it is a safety one, where the declaration is species-level and the supply chain is long.
Where competitive intensity is genuinely high is not in winning contracts but in retaining the reference capability behind them. A laboratory that loses its isotope capability does not lose a service line; it loses the ability to adjudicate, and it becomes a screen. That is the position most of the mid-market occupies, and it is why the six tier-one networks keep investing in confirmatory capacity they cannot fully monetise.

- **Structure:** fragmented
- **Tier 1 Players:** 6
- **Intensity:** high

## Value Chain

The value chain: from composite sample to defensible verdict
The chain in this market is short and the margin is not where the sample is handled. Five commercial roles sit between a consignment and a decision, and only one of them is properly compensated.
The first role is the consolidator or trader, which buys from producers across many origins and blends before the goods are labelled. This is where the fraud economically originates, because blending is the operation in which a cheap substitute becomes indistinguishable from an expensive original. The Commission’s honey investigation found that more than half the operators involved — 57 per cent — had exported consignments suspicious of adulteration with extraneous sugars, and more than 60 per cent, 66 operators, had imported at least one such consignment. Fraud here is a procurement decision before it is an analytical problem.
The second role is the brand owner or private-label retailer, which carries the legal exposure for the claim. It is the party that buys most authenticity testing, because it is the party that cannot argue ignorance. Eurofins describes its own authenticity service in exactly those terms, framing the offer as protecting the legacies of the largest food companies and covering ingredient verification, adulteration detection and geographical origin across agave syrup, honey and maple syrup, basmati rice, cereals, coffee, tea and spices, fish, flavours, fruit juice, meat, olive oil, spirits and wines. That list is not a product catalogue; it is a map of what a modern label asserts.
The third role is the contract laboratory network, which performs the determinations. The fourth is the instrument, reagent and database layer, which supplies the capability. The fifth is the authority, which in the EU is the only actor whose result carries immediate legal weight, and which therefore buys a different product: surveillance rather than compliance.
Inside the analytical work there are four steps, and they have completely different economics.
Sampling design decides everything and is almost never sold as a line item. A composite sample from a 20-tonne consignment of honey bought at the 2024 average third-country import price of EUR 1.92 per kilogram represents about EUR 38,400 of goods; the cheapest end of the published rate card for a species determination is EUR 150, which is 0.4 per cent of that consignment, and the dearest is EUR 550, which is 1.4 per cent. On Chinese honey at the 2024 average of EUR 1.20 per kilogram the same consignment is worth EUR 24,000 and the test is 0.6 to 2.3 per cent of it. The test is trivially affordable relative to the goods. What is not affordable is representativeness: one sample of a 20-tonne lot is one sample, and the Commission’s own conclusion is that a simple test is not sufficient and that proper follow-up of suspicions is necessary. The binding constraint on this market is not the price of the assay. It is how many samples it takes to make one assay mean something, and that number is set by the customer’s risk appetite, not by the laboratory’s cost base.
Screening is where volume lives. Fourier-transform infrared and Raman spectroscopy are the techniques the UK Food Standards Agency’s 2025 review of edible oil authenticity singles out as showing promise for rapid, low-cost testing, while noting that many approaches are still under development and lack thorough validation. Screening is priced per sample at a level that makes wide coverage possible, and it is deliberately non-specific: its job is to sort samples into clear and suspicious, not to convict.
Confirmation is where price lives. A suspicious screen goes to isotope-ratio mass spectrometry, to proton NMR profiling, to real-time PCR, or to a chromatographic determination of sterols and triacylglycerols. This is expensive, slower, and requires a fully equipped laboratory and a validated method. In edible oils the FSA review notes the increasing sophistication of fraud within the supply chain, with businesses relying on proprietary protocols — which hampers standardisation precisely because the protocol is the competitive asset.
Adjudication is where margin lives. A confirmed analytical result is not yet a decision. It has to be placed against a definition of the product, a tolerance, and a traceability record, and that work is done by technical consultants, second-party auditors and, where it escalates, by forensic accountants and lawyers. It is the only step in the chain that scales with legal exposure rather than with sample count, and it is the reason the advisory layer is growing faster than the analytical layer.
The comparison with the safety side of this cluster is instructive. In food safety testing the verdict is generated by the method: a pathogen result against a published reference method is a pass or a fail, and no adjudication is required. In mycotoxin testing a statutory limit converts a concentration into a legal status in one step. Authenticity has no such conversion. Every positive result is an argument, and the argument must be won.
That is also why the intermediate layer exists at all. Brokerage of laboratory services — a single purchase order routed to whichever accredited laboratory holds the right scope, with a published service fee per order on top of the analytical price — is a real and growing part of the chain, because a brand owner with fifty authenticity questions across twelve matrices does not want fifty supplier relationships. The broker adds no analytical capability whatsoever and captures value purely from the fragmentation of the supply side, which tells you how fragmented it is.
Where the money finally settles is the reference layer. To determine that a fish is pollack rather than a cheaper substitute, a laboratory needs something to compare against; the Joint Research Centre responded in November 2024 by producing two certified reference materials, for pollack and saithe, as freeze-dried powders milled from single fillets and certified to species by sequencing and database comparison. The FSA’s oils review reaches the same conclusion from the other direction, recommending that authentic certified reference materials are essential to support quality control and to encourage proficiency-testing uptake. Whoever holds the reference set holds the standard, and the standard is the only thing here that cannot be competed away on price.
One consequence deserves to be stated plainly. Because the reference layer sits upstream of the laboratory, the laboratory is not the most powerful actor in its own value chain. This segment is populated by very large testing networks that are, in the authenticity niche, price-takers on capability and price-makers only on service. The olive oil and industrial sugar chains illustrate the asymmetry: the first because its authenticity standard is a set of chemical parameters that anybody can run but only a reference panel can interpret, the second because it is the substrate that every syrup-extension fraud is made from, and the party selling the substrate has no interest in the test.
The value chain, in one sentence: the sample is a commodity, the confirmation is a service, the reference is a moat, and the verdict is a legal product sold by somebody who is not a laboratory.

- **Model:** hybrid
- **Distribution:** hybrid
- **Supply Complexity:** high

## Investment Activity

Investment activity
Capital in this market is spent on capability, not on capacity, and the distinction explains both the deal record and its silences.
A pathogen laboratory is a room with incubators, media and a plate reader; it scales by adding rooms and people. An authenticity laboratory is a room with a 400 MHz nuclear magnetic resonance spectrometer, an isotope-ratio mass spectrometer and a real-time PCR platform, and it scales by adding reference data. The first is a volume business with a linear cost curve. The second is a fixed-cost business with a research programme attached. That is why the six tier-one networks hold confirmatory authenticity capability out of proportion to the revenue it generates directly: it is the capability that lets them adjudicate rather than merely screen, and adjudication is what a client with a mislabelled consignment actually needs to buy.
The disclosed deal record is consistent with that reading and unhelpfully thin in exactly the place one would want it. SGS acquired Murray-Brown Laboratories in the United States during 2025 — a testing laboratory specialising in food, ingredients and nutraceutical analysis, expanding its food safety and quality assurance presence in North America. That is the shape of the deals happening: capability tuck-ins inside an existing food footprint, bought for scope and geography rather than for an authenticity franchise. Eurofins reported acquisitions contributing EUR 174 million of 2025 revenue, and stated that had those businesses been owned from 1 January the contribution would have been EUR 112 million higher — a programme of many small purchases rather than any single transformation.
The instrument layer is reallocating capital rather than consolidating it. Thermo Fisher agreed in April 2026 to sell the microbiology business that had disclosed USD 645 million of 2025 revenue for approximately USD 1.075 billion, which is a portfolio decision about where instrument and consumable economics are best rewarded, not a retreat from food. Bruker continues to invest in the one asset in this market that behaves like a durable franchise: the reference database behind its FoodScreener profiling methods, at roughly 28,500 characterised honey samples and growing.
What is absent from the record is as informative as what is in it. No authenticity-specific transaction discloses separable revenue, so this report states no transaction multiple. The reason is structural rather than a gap in our research: authenticity testing is sold inside broader food testing contracts, has no standalone reporting line at any of the six networks, and is therefore not separable in a purchase price allocation. Any multiple quoted for this niche would be a multiple for food testing generally, relabelled. We decline to do that.
Where the genuinely underinvested opportunity sits is the reference layer, and two public bodies have now said so in print. The Joint Research Centre manufactured certified reference materials for pollack and saithe — freeze-dried, cryogenically milled powders from single fillets, certified to species by DNA barcode sequencing and database comparison — explicitly so that laboratories using them can be confident their tests are accurate. The UK Food Standards Agency’s 2025 review of edible oil authenticity concluded that authentic certified reference materials are essential to support quality control and encourage proficiency-testing uptake, and separately that the sector needs investment in widespread testing and point-of-use methods and in developing confirmatory techniques. Two regulators, two matrices, the same finding: the binding constraint is not instruments, it is characterised reference material and validated methods.
That is a product opportunity rather than a market-entry opportunity, and it is worth being precise about why. A certified reference material requires a characterisation programme, a certification body, an uncertainty budget and a stable supply of authentic starting material. It has no intellectual-property protection worth speaking of and it is priced by the unit. Nobody in this market is strongly incentivised to build a broad library of them, because the beneficiary is the whole laboratory population and the builder captures only the sale of the vials. That is a textbook public-good underprovision, and it is being filled slowly, matrix by matrix, by the JRC and by national reference laboratories.
The commercial consequence is a two-tier investment case. In the near term, capital flows to whoever can assemble private reference sets — a database business dressed as an instrument business — because that is where switching costs and pricing power actually live. In the medium term, if public reference materials proliferate across the matrices that matter (oils, species, botanicals, juices), the private premium compresses and the value migrates back to throughput, logistics and turnaround time. Which tier wins is a regulatory question, not a technology question, and it turns on how fast the delegated acts under Directive (EU) 2024/1438 require methods that only certified reference materials can support.
One further capital signal is worth recording because it is easy to miss. Part of the cost base of this market is borne by the state. Coordinated control plans in which 21 or 15 Member States sample and a single central laboratory analyses are a public subsidy to the analytical base: the JRC carries the method development and the reference work, and the Member States carry the sampling. That structure keeps the marginal cost of a surveillance sample low for authorities and correspondingly caps what a commercial laboratory can charge an authority for the same determination. It is one reason the enforcement segment of this market is the smallest and the slowest-growing of the five buyer groups.

- **Trend:** rising
- **Capital Intensity:** medium
- **Recent M&A:** yes

## Technology & Innovation

Technology and innovation
The technology story in this market begins with a failure, and the failure is better documented than any success.
For two decades the workhorse method for detecting sugar syrup added to honey was stable carbon isotope ratio analysis by elemental analyser isotope-ratio mass spectrometry, AOAC method 991.41. It works because maize and sugarcane are C4 plants and their syrups carry a carbon isotope signature that honey, built from C3 nectar, does not. The Joint Research Centre’s 2023 study put that method against 320 imported consignments and reported that it was not effective in detecting honey suspicious of being adulterated — and drew the obvious conclusion: such syrups are no longer used to extend honey and have been replaced by syrups made mostly from rice, wheat or sugar beet. All three are C3 plants. The adulterant changed its isotope chemistry to defeat the reference method, and the reference method had no way to notice.
What replaced it was not a better single method but a panel. The JRC ran EA-IRMS and liquid chromatography IRMS, high-performance anion-exchange chromatography with pulsed amperometric detection, and proton nuclear magnetic resonance together, and treated the combination as the basis for suspicion. That is the structural fact about this market’s technology: authenticity is a multi-method problem because the adversary is optimising against each method separately. No single determination is dispositive, and the market therefore sells more determinations per sample than any other food testing category — the EU herbs and spices plan ran nearly 10,000 analyses on 1,885 samples, or about 5.3 analyses per sample.
Figure 3. Technique mix, 2025 (USD 1,650.0 million). Source: Pheonix Research model.
The most commercially consequential technology is proton NMR profiling. Bruker‘s FoodScreener runs at 400 MHz with fully automated, push-button analysis and reports against matrix-specific databases; the honey method detects exogenous sugars, verifies country of origin and botanical variety, assesses composition and freshness, flags atypical samples, and performs a regulatory compliance check for HMF, sucrose and glucose and fructose against EU and Codex criteria. The wine and juice methods run on the same platform. The reference database behind honey profiling holds roughly 28,500 samples across more than 50 countries and 100 monofloral types, spanning seasons, production years, blends and production methods, and the stated reason for that breadth is that the statistical models are only as good as the representativeness of what they were trained on.
That platform has two opposite economic effects at once, and separating them is the analytical core of this report.
It is deflationary per sample. Bruker’s own proposition is that one platform replaces many tests and lowers the total cost of quality — a single push-button run against a database substituting for a series of targeted assays. When one run replaces three, revenue per sample falls even as information per sample rises. This is the mechanism behind the minus USD 45.5 million pure assay deflation term in our decomposition, and it is the reason realised price per sample grows at only 0.48 per cent a year while input costs rise faster.
It is inflationary per claim. The same run produces determinations for origin, variety, purity, freshness and compositional compliance simultaneously. A customer who previously bought one answer now receives five, and the label that must be defended asserts five things. That is the claim-density term, worth USD 157.7 million of growth, and it is why the mix contribution accelerates from 2026 rather than growing smoothly: a legislative definition can turn a single-attribute claim into a multi-attribute obligation in one instrument.
The screening layer is where the next shift is likely and where the evidence is least settled. The FSA’s edible oils review found that spectroscopic methods — Fourier-transform infrared and Raman — show promise for rapid, low-cost testing, but also that many approaches remain under development and lack thorough validation, and that businesses rely on proprietary protocols, which hampers standardisation. A validated, cheap, point-of-use screen would move volume out of the laboratory and into the factory, which is good for coverage and bad for revenue per test. It would also make the confirmatory step the only defensible product, pushing the laboratory layer further toward adjudication.
Molecular methods are the most standardised leg of the tripod and therefore the least defensible commercially. Real-time PCR species determination is quoted at EUR 150–550 per sample with a detection limit of 0.1 per cent, two weeks’ turnaround and accreditation — a specification several hundred laboratories worldwide can meet. DNA barcoding gives the reference sequence, and the JRC’s pollack and saithe materials give the physical comparator. Next-generation sequencing extends this from presence to relative composition, which matters for products where a declared percentage is itself the claim. None of it is scarce. All of it is priced as a service.
Chromatographic determination of sterols, triglycerides and volatiles is the oldest and most legally entrenched layer, and it is where olive oil authenticity actually lives: a set of chemical parameters that any equipped laboratory can run. Its weakness is the one the FSA identifies — interpretation depends on the reference frame, so results are only as good as the panel or database behind them. Its strength is that it is written into standards, which makes it a compliance product rather than an investigative one.
Which brings the comparison that makes the cluster coherent. In culture media and microbiology reference materials the technology is stable and the reference strain is defined by a pharmacopoeia, so value sits in the definition and the market price-captures half its growth. In ATP and hygiene monitoring the technology is weak and the threshold is unpublished, so value sits in the administrative switching cost. Here the technology is strong and improving and the definition is absent, so value sits in the reference set — and the reference set is the one asset an instrument vendor can own and a laboratory cannot easily copy.

- **Innovation:** high
- **Patent Activity:** moderate
- **Maturity:** growth

## Market Risk

Market risk
Four risks matter here and three of them are not the ones usually listed.
1. The demand signal went the wrong way in the most recent full year
The intuitive demand driver for this market is fraud, and the most recent complete official dataset contradicts the intuition. The EU Alert and Cooperation Network’s 2024 annual report records 9,460 notifications across the network, up 8 per cent on 2023, with RASFF rising 12 per cent to 5,250. Within that total, agri-food fraud notifications declined by one third and stood at 546, while plant health notifications almost tripled. Germany was the largest notifying country at 1,907, ahead of the Netherlands at 1,155 and Italy at 965.
There are three readings and we cannot choose between them from the outside. It may be enforcement reallocation: competent authorities in several Member States spent 2024 on plant health and animal welfare, both newly networked, rather than on fraud. It may be reporting behaviour: fraud cases increasingly routed through OLAF and Europol channels rather than through iRASFF, which would move the count without changing the activity. Or it may be genuine decline, which would be surprising given that OPSON XIV, run in the same period, seized goods worth EUR 95 million and reported 631 individuals to judicial authorities. What we can say is that a forecast resting on ‘fraud is increasing’ is resting on a statistic that fell by a third last year, and this one does not rest on it. The growth case here rests on claim density and on surveillance sampling rates, both of which are observable independently of reported fraud incidence.
Figure 4. Change in EU notification volume, 2024 against 2023 (per cent). Source: EU Alert and Cooperation Network, 2024 Annual Report.
2. The fastest-growing fraud typology is analytically invisible
This is the sharper risk. Europol’s OPSON XIV outcome, published in October 2025, identified as the year’s main trend organised crime groups infiltrating waste disposal companies to reach expired food awaiting destruction, then removing the original best-before or expiry dates with solvents and printing new, falsified dates, describing the scale of the practice as unprecedented. It also records olive oil and wines with a protected designation of origin as remaining strongly affected by counterfeiting and designation abuse.
No assay in this market detects a reprinted date. That fraud is defeated by custody records, waste-manifest reconciliation and retail audit — documentation controls, not analytical ones. So the fastest-growing identified typology sits entirely outside the product this market sells, while the typologies it does address — composition, species, origin, variety — are the older ones. A market whose growth story depends on fraud getting worse should be worried that the newest and largest fraud trend has no test attached to it.
3. Suspicion is not proof, and the asymmetry is expensive
The Commission’s finding that current methods are not developed enough for a simple test to be sufficient, and that proper follow-up of suspicions is necessary, is a statement about the commercial product. 46 per cent suspicious is not 46 per cent non-compliant. In the honey action, 44 operators were investigated on the back of that testing and seven were sanctioned — a conversion rate of roughly one in six. That ratio is the single most useful number for understanding the economics of the analytical layer: most of what it sells is an indication, and the value is realised only after investigation.
The resulting asymmetry hurts both sides of the trade. A laboratory reporting a false positive against a named supplier carries contractual and reputational exposure; a laboratory reporting a false negative lets a fraud through and may never learn of it. The rational response is conservative decision thresholds, which produce more false positives, which make the result less useful and depress what a customer will pay for it. That dynamic — not instrument cost — is why price per determination is falling.
4. Concentration in a handful of commodities, and the price of the authentic good
Demand concentrates in honey, olive oil, herbs and spices, fish, juices and wine. The last risk in this chapter follows directly: fraud incentives are proportional to the price gap between authentic and substitute, so when the authentic price falls the market’s own demand driver softens. Third-country honey entering the EU in 2024 averaged EUR 1.92 per kilogram, 11.5 per cent lower than 2023, on volumes up 6 per cent to 172,367 tonnes, with Chinese honey at EUR 1.20 per kilogram. A cheaper authentic product narrows the gain from extension, and the narrowest-margin products in the category are precisely the ones where a multi-attribute determination looks least affordable to the buyer.
Olive oil runs the other way, and it is the useful contrast: a high-value, varietal, geographically designated product with an established chemical standard, which is why it recurs in every enforcement action and why it supports confirmatory testing. A portfolio of this market’s demand is therefore a bet on which commodities stay expensive and stay designated.
Model risk and sensitivity
The value base is constructed because no vendor reports authenticity as a line item; the band we would defend around the 2025 point is USD 1.45–1.85 billion. The sensitivity is published rather than buried: hold the realised price per sample flat at USD 196.43 and the market reaches USD 2,867.9 million in 2033 at the volume rate of 7.15 per cent; hold sample volume flat at 8.4 million and it reaches USD 1,714.5 million at the price rate of 0.48 per cent. The eight-year spread between those two answers is larger than most forecast disagreements in this cluster, and it is entirely a statement about whether authenticity becomes a required determination or stays a discretionary one.
A fifth risk is worth naming for completeness. Substitution risk is high, and it is internal to the customer rather than competitive: if point-of-use spectroscopy becomes validated and cheap, the screening layer moves into the factory and the outsourced market is left with confirmation only. That is a smaller market per sample and a defensible one per determination, and it is the most likely shape of 2033.

- **Overall Risk:** moderate
- **Substitution Risk:** high

## Regulatory Landscape

Regulatory landscape
This is the only segment in the food safety cluster where regulation creates the obligation without creating the limit, and understanding that gap is the whole of the price story.
The Breakfast Directive
Directive (EU) 2024/1438 of 14 May 2024 amends four marketing directives at once — honey (2001/110/EC), fruit juices (2001/112/EC), jams, jellies and marmalades (2001/113/EC) and certain preserved milks (2001/114/EC) — which is why it is known as the Breakfast Directive. It was published in the Official Journal on 24 May 2024, had to be transposed into national law by 14 December 2025, and its rules apply from 14 June 2026. France transposed by Decree No. 2026-312 of 24 April 2026, in force on the same date.
For honey, the operative change is origin. The country or countries of origin where the honey was harvested must be indicated on the label, and where honey originates in more than one country the countries must be shown in the principal field of vision, in descending order of their share by weight, together with the percentage each represents, subject to a 5 per cent tolerance per share, calculated on the basis of the operator’s traceability documentation. Where a blend draws on more than four countries and the four largest shares exceed 50 per cent, Member States may allow percentages for those four only, with the remainder listed in descending order and unquantified. Packages with a net quantity under 30 grammes may substitute two-letter ISO 3166-1 codes. The particulars become mandatory particulars under Regulation (EU) No 1169/2011, which means the general food labelling enforcement machinery attaches to them.
Two features of that provision drive demand. First, it converts origin from a category into a quantity: a label that once said ‘blend of EU and non-EU honeys’ now asserts a set of percentages. A percentage is falsifiable in a way a category is not, and proving a share to within five points is harder than proving a country, because it requires the determination to be quantitative across every component rather than merely consistent with the declared origin. Second, it moves the burden of proof onto the operator. For the first time in this category the producer, not the authority, has to be able to demonstrate something about origin on demand — and the reference point for that demonstration is the operator’s own traceability documentation, which is precisely what the fraud cases showed being fabricated.
What the directive does not yet do is equally important. Article 4a empowers the Commission to adopt delegated acts on five things: the criterion of ‘mainly’ as regards floral or vegetable origin; composition criteria to ensure honey has not been heated or treated so as to destroy or significantly deactivate natural enzymes, by reference to the invertase index; criteria to ensure and verify that pollen is not removed from honey and that absolute pollen content and pollen spectrum are not modified; minimum pollen content in baker’s honey; and the methods and criteria to determine where honey was harvested plus Union-wide traceability requirements. Those delegated acts are due by 14 June 2029, preceded by feasibility studies, one of which must analyse digital solutions including a unique identifier code.
That is the second wave, and it is the event that would change the price structure of this market. A pollen-content criterion and an invertase index are numerical limits. A numerical limit is the one thing this market has never had, and the cluster evidence is unambiguous about what a numerical limit does to price capture: the categories where a threshold is published capture a third to a half of their growth in price, and the category where no threshold exists gives price back. If the 2029 delegated acts land as numbers, authenticity testing stops being an investigative service and starts being a compliance product — and the 2033 figure would be reached through price rather than volume.
The United States chose documentation instead
The contrast is stark and it explains the geography of this market. The FDA’s Food Traceability Final Rule under Section 204(d) of the Food Safety Modernization Act requires persons manufacturing, processing, packing or holding foods on the Food Traceability List to maintain records of Key Data Elements at Critical Tracking Events and to produce them to the FDA within 24 hours. The original compliance date was 20 January 2026. The FDA proposed to extend it by 30 months to 20 July 2028, and the Continuing Appropriations, Agriculture, Legislative Branch, Military Construction and Veterans Affairs, and Extensions Act of 2026 directed the agency not to enforce the rule before that date; the FDA has said it intends to comply with that directive. It also convened a public meeting on lot-level traceability on 6 November 2026.
FSMA 204 is a records rule, not an analytical one. It answers ‘where did this lot come from and where did it go’ — the question that defeats the relabelled-expiry fraud OPSON identified — and it does not answer ‘is this what it says it is’. Both are legitimate regulatory strategies and they purchase different things. The EU is buying determinations; the United States is buying provenance. For the purposes of this report the consequence is that growth is EU-weighted, and we state that as a qualitative reading rather than modelling a regional split, because the model carries no geography.
The enforcement architecture that generates the demand
Three mechanisms matter. The EU Agri-Food Fraud Network sits inside the Administrative Assistance and Cooperation network and handles cross-border fraud cases through iRASFF. Coordinated control plans — the honey action across 15 Member States plus Switzerland and Norway, the herbs and spices action across 21 Member States plus Norway and Switzerland — are the mechanism by which the state buys analytical work at scale, with sampling done nationally and analysis done centrally at the Joint Research Centre. And OPSON, coordinated by Europol with OLAF and DG SANTE, ran its fourteenth edition across 31 countries in 2025 with 31,165 checks, 259,012 packages, 1.4 million litres of beverages and 11,566 tonnes of food seized, goods worth about EUR 95 million, 631 individuals reported to judicial authorities, 101 arrest warrants and 13 organised crime groups disrupted.
None of those numbers converts into a purchase order for an authenticity determination, which is the point. Enforcement activity is the market’s marketing, not its revenue.
Accreditation is the real barrier to entry
The de facto licence to operate is accreditation to ISO/IEC 17025 for the specific method, and it is what separates a laboratory selling a determination from one selling an opinion. It is visible in the rate cards: the manuka pollen DNA method is quoted as accredited to ISO 17025, and the PCR species determination is sold as an accredited test method with a stated detection limit. That gives this market a standardised commercial pathway — there is no approval to seek, no licence to obtain, no product to register. The barrier is capability plus accreditation, which is why the market is fragmented among many accredited laboratories and concentrated only in the confirmatory layer where the capital cost bites.
The gap that no regulation fills is the one identified earlier in this report: there is no statutory reference method for authenticity anywhere, and no general legal definition of what makes a food authentic beyond not being misdescribed. Directive 2024/1438 narrows that gap for four breakfast categories from 2026 and for composition and pollen from 2029. Until then, every determination in this market is an expert judgement dressed as a measurement, which is why it is worth what it is worth.

- **Complexity:** high
- **Approval Pathway:** standardized_commercial

## Frequently Asked Questions

**Q: How large is the global food authenticity and adulteration testing market?**

USD 1,650.0 million in 2025, rising to USD 2,980.0 million in 2033 at 7.67 per cent compound. That covers laboratory analytical revenue for authenticity-specific determinations, instrument and consumable revenue attributable to authenticity workflows, and the data, surveillance and advisory layer. No company discloses authenticity testing as a standalone line, so the base is a constructed estimate; the band we would defend is USD 1.45–1.85 billion around the 2025 point.

**Q: Why does this segment grow faster than food safety testing overall?**

The parent market grows 5.93 per cent to 2033; this one grows 7.67. The difference is what the unit of demand is tied to. Safety testing volume is tied to batches produced and to statutory limits, and it price-captures 25 per cent of its growth because a published reference method makes reagents comparable. Authenticity volume is tied to the number of declarable claims a product carries and to the number of consignments sampled, and neither has a ceiling set by regulation.

**Q: Does the EU Honey Directive create new testing demand from 2026?**

Partly, and the timing is precise. Directive (EU) 2024/1438 had to be transposed by 14 December 2025 and its rules apply from 14 June 2026: the countries of origin of a blend must be declared in descending order of share, with the percentage each represents and a 5 per cent tolerance per share, calculated against the operator's traceability documentation. That shifts the burden of proof onto the operator for origin. It does not yet create a compositional limit: the delegated acts on composition criteria, pollen integrity and Union-wide traceability are not due until 14 June 2029.

**Q: What is the biggest limitation of this forecast?**

Two things. The value base is constructed because no vendor reports authenticity as a line item, so the absolute level is arguable even where the growth rate is not. And the demand signal points the other way: EU agri-food fraud notifications fell by one third in 2024 to 546 while total network notifications rose 8 per cent, so a forecast resting on 'fraud is increasing' is resting on a statistic that declined in the most recent full year. The growth case here rests on claim density and surveillance sampling, not on fraud incidence.
