The Colour Is the One Thing on the Label Nobody Regulates

Every other product attribute in French wine is written into a cahier des charges. Grape varieties are listed, boundaries are mapped, minimum alcohol is fixed, yields are capped. The attribute that actually sells Provence rosé — the colour — appears in none of them.

That is not a small omission. Provence has spent thirty years converting itself into a single-colour region on the strength of a visual signal that no appellation can legally enforce, and the 2025 harvest shows how complete the conversion is.

Within Côtes de Provence, which is 902,583 hectolitres of the total, 814,611 is rosé. A region that historically made three colours now makes one and treats the other two as residue. The colour is the product, and the product is not defined in law.

You cannot blend your way to consistency

This is where rosé’s economics stop resembling Champagne’s and start resembling a fresh-food business.

The interprofessional council’s quality monitoring samples wines at the point of sale, not at the cellar door — the point is to test what a consumer actually receives. In 2025 it drew 211 samples of the three appellations across France and put the vintage composition on the record:

Eighty-two per cent of the rosé on sale in the sampling year came from the most recent vintage. Only 8 per cent carried no vintage at all. A Champagne house can correct a difficult year by pulling reserve wine into the blend; that is what its reserve is for, and it is why non-vintage Champagne is a manufacturing achievement rather than a marketing convenience. A Provence rosé cannot do it: the vintage is on the label for 92 per cent of the category, and the vintage is the year the colour has to be right.

The 2025 vintage gave the region 1,095,423 hectolitres, up 9.37 per cent on 2024’s 1,001,589. But the wine available to sell into 2026 was put at 1,160,036 hectolitres — up only 1.5 per cent. The gap is 64,613 hectolitres, 5.6 per cent of availability. There is no stock: the category is sold within the year it is made, so a colour failure is not smoothed over later. It is sold.

The cost is in the picking date, not the winery

Making a pale rosé is cheap in the cellar and expensive in the vineyard. Skin contact measured in hours rather than weeks releases less colour and less tannin, ties up a fraction of the capital a red of the same volume requires, and lets a producer switch a block into rosé inside a single season. That is why the category can grow quickly, and why it can also collapse quickly.

The consistency, though, comes from somewhere else: picking date and block selection. Paler means earlier, and earlier means lower potential alcohol, less extract and a narrower window in which the fruit is both ripe enough and pale enough. In the appellation model that cost cannot be recovered by planting more, because it is fixed by the surface the region has: 27,428 hectares producing 1,095,423 hectolitres of rosé is an average of 39.94 hectolitres per hectare across the whole vineyard.

The capital record says the same thing. LVMH’s plan at Château d’Esclans is to add 60 hectares of Côtes de Provence vines to the 74 already inside a 267-hectare estate — not a new winery, a farming decision. And the region is building the institution to match: the Vitipole Provence Rosé, a research centre with an experimental cellar, a laboratory and a tasting room, obtained its building permit and held its foundation ceremony on 19 December 2025 with the State, the Région Sud, the Var department, Crédit Agricole and Groupama among the funders. Rosé now has a research institute. It has had no protected definition of its own distinguishing feature.

What the defect rate actually tells you

In the same 2025 monitoring round, 84 per cent of samples were conforming to the appellation specifications and 32 wines were reported with major defects — 15.2 per cent of the 211 drawn. That is the interesting number, and it is not a scandal: it is what a category with a visual product specification and no legal definition of it should be expected to produce.

The sampling frame also happens to describe the commercial shape of the category. By outlet, 34 per cent of samples came from hypermarkets, 27 per cent from independent wine merchants, 23 per cent from supermarkets, 11 per cent elsewhere and 5 per cent from discount stores. Twenty of the 211 samples were bag-in-box.

A defect found in a hypermarket is a defect at scale, in front of the shopper who buys on sight. A defect in the discount channel is a defect in the one place where price, not colour, is the tie-breaker. Both matter, and Provence now measures both — which is more than most appellations do and rather less than a consumer would assume.

The mark is an attempt to regulate the unregulated

Provence’s response is a shared recognition symbol across Côtes de Provence, Coteaux d’Aix-en-Provence and Coteaux Varois en Provence, developed during 2025 and launching in 2026, together with a formal segmentation between AOP and IGP positioning. Read it for what it is: an interprofessional body acknowledging that the thing being copied is not the boundary or the grape, and reaching for the only lever — a mark — that touches colour, bottle and shelf language at once.

A mark is not a process. It cannot make a wine paler, and it cannot stop a warm region with a red press from producing something the same shade. What it can do is move the decision from the eye to the label, and that is a shift in the basis of competition that a buyer notices long before a consumer does.

Three things that would prove this reading wrong

If defect rates fall and price does not move once the mark is live, colour was never the purchase driver and this analysis is over-weighted on it.

If the vintage mix shifts materially towards non-vintage rosé, producers have found a way to blend for consistency, which changes the whole no-inventory argument above.

If a major appellation writes a colour parameter into its cahier des charges, the unregulated attribute becomes a regulated one, and the mark becomes redundant. Nothing in the 2025 record suggests that is being attempted.

Rosé is 8.5 per cent of world wine volume and grows 4.80 per cent a year in value on volume growth of 0.20 per cent. Almost all of that forecast is the price of a colour nobody has defined. That is a strong position while the colour reads as provenance — and a fragile one the moment it reads as a style any warm country can produce.