Pheonix Research - Editorial Index
Editorial analysis anchored to industries, segments, and the market intelligence behind them.
China banned Illumina sequencer imports from 4 March 2025 but did not restrict reagents, consumables or maintenance. Illumina's Greater China consumables fell 16.74 percent while its instrument funnel closed entirely - a decaying annuity, not a shrinking market. In H1 2025 MGI and Berry Genomics took 59.3 percent of Chinese sequencer procurement value.
10x Genomics sold 18.73 percent more consumable reactions for 2.79 percent more revenue: realised price per reaction fell 13.43 percent. Hold that decline and grow volume 12 percent and the library preparation market contracts 3.04 percent a year. The forecast rests entirely on that number narrowing toward 5 percent.
Medicare pays a new advanced diagnostic laboratory test at its own list price for roughly nine months, then at the median of actual claims. Shield launched at USD 1,495, Guardian360 at USD 5,000, Signatera at USD 3,500. And in the CY2025 fee schedule, one code moved 34.26 percent in ten days with no change to the product.
The colorectal screening determination effective 8 June 2026 requires either 90 percent sensitivity with 87 percent specificity, or 79 percent with 90 percent. Shield reports 83.1 and 89.6. Against the second route it is four-tenths of a point short — but the two specificity figures measure different denominators, and neither document available to a reader resolves which one applies.
Four of the six largest tumour profiling participants lost money in 2025. The one that made USD 66.4 million sells two tests. Decipher grew 27 percent to USD 310.7 million on about 102,000 tests — roughly USD 3,046 each — while a broad menu turned USD 727 million into a 5.91 percent EBITDA margin. Coverage, not the panel, is the product.
QIAcuity produced USD 89 million in 2025 against a USD 600 million digital PCR segment: 14.8 percent share. The 2028 target of USD 250 million, at 15 percent segment growth, requires 27.4 percent. A dollar target on a growing denominator is a share target in disguise, and the denominator is the part nobody states.
Guardant grew oncology revenue 26 percent on 34 percent more tests: realised revenue per test fell 6.04 percent. Veracyte fell 1.09 percent per test. Myriad's tumour profiling shrank while volume rose. In tumour profiling, growth is being bought with price, and the identity that reconciles volume to revenue is the most useful number in the category.
QIAGEN sizes the 2025 qPCR market at about USD 2.7 billion. Roche's Molecular Lab alone reported USD 3,048 million in the same year. Both are honest and they cannot describe the same market: one measures open chemistry, the other measures closed cartridge systems. Adopting a supplier's addressable market as the size of a category understates the level and imports the wrong growth rate.
Consumables in this market have been deflated roughly 75 percent while forecast value growth runs at 20.50 percent - and that is not a contradiction, it is what happens when the unit of sale moves from the device to the experiment. The cost per usable data point decomposes into three inputs, and only one of them is broken.
An FDA qualification licenses a use, not a product - and as of the end of 2024 no organ-on-chip model had completed the process. That single fact, not instrument performance, explains why a twenty-year-old technology still measures its market in hundreds of millions.
Transient transfection cannot be banked, licensed or scaled into commercial supply, which is why it will never become a manufacturing platform - and why its growth tracks the number of experiments started rather than plant capacity. The 2025 filings, and the multiples paid for transfection reagents, show where the value actually sits.
ICH Q5A(R2) made next-generation sequencing an acceptable alternative to classical adventitious-agent testing. That created a sequencing market that is not driven by discovery funding but by the number of manufacturing batches started - and whose real product is a qualified bioinformatics pipeline, not an instrument.
The FDA's laboratory-developed test rule was vacated, then replaced with something close to the status quo. That settled less than it appears. With the clinical laboratory fee schedule entering capped annual reductions in 2027 and the EU's IVDR dates now split by risk class, the growth rate of this market is a coverage forecast with a technology input.
Roche's diagnostics margin halved in 2025 while cost of sales rose five times faster than revenue. That is not a procurement failure - it is what happens to a business whose real product is a reagent sold through an installed base of instruments. A close read of the 2025 filings shows why placements are a cost and pull-through is the revenue.
Illumina's 2025 revenue was flat at USD 4.34 billion. Read as a headline that looks like a market stalling; read underneath, it shows growth has migrated to the high-throughput instrument, the clinical customer and a published reimbursement code. The sequencer is no longer where the money is.
Twenty-three of the FDA's 46 novel approvals in 2025 carried orphan designation — half the year's approvals on 19.2 per cent of its pivotal evidence. The median orphan programme enrolled 114 patients, nine ran on a hundred or fewer, and the whole cohort's evidence base is smaller than three non-orphan programmes enrolled in the same year.
Ontario's Auditor General found that nine of the twenty AI documentation systems it tested fabricated content or proposed referrals and blood tests the physician never ordered — twelve transcribed a different drug than prescribed. Five weeks earlier, an independent evaluation of the same technology class reported 1.11 hours a week saved and same-day note completion up 19 points. Both documents are true, and the gap between them is where the next cycle of the Canadian market is decided.
Nine of the 46 methods Brazil recognises as alternatives are animal tests, and the register is deep at the skin and empty inside the body. That gap, not the ban, is what decides which ingredient classes can be served in vitro — and why the growth curve is a step in 2027, a plateau, and a second step in 2031.
The optics industry earns USD 8.55 a year for every person living with vision impairment, against a USD 411 billion annual productivity loss it captures 4.57 per cent of. That mismatch — not demand, not pricing — is what actually sets the growth path of a USD 18.8 billion layer heading to USD 30.5 billion by 2033.
A pill is eligible for Medicare price negotiation at seven years, a biologic at eleven. With the negotiated price taking effect two years later, that is nine protected years against thirteen — a four-year subsidy for injectables written into a price-control statute, and the reason a 10 per cent market can grow on flat approval counts.
