Insights · Brands, factories and margin

The £1 in every £2: how own-label quietly became half of British grocery

Jatinder Purewal · ACMA, CGMA · Executive MBA · 14th August 2026 · 17 min read

Brands, factories, margin: who really keeps the shopper's pound, from three seats at the same table.

More than half of what Britain now puts in its supermarket basket carries the retailer's name, not a manufacturer's. That crossing point is now true on every serious measure, and the measures disagree in an instructive way. Worldpanel by Numerator put own-label at 52.2% of grocery spend in January, the highest share it has recorded; Circana, measuring units across 230 packaged categories, put private label past half of everything sold for the first time, with a 44% value share on that narrower basis. The gap is not an error: a take-home panel that includes loose produce and in-store bakery reads higher than a packaged-categories measure, and anyone quoting one own-label number without saying which basis they mean is quoting a headline, not a measure. There was no single moment when this happened. It is the least discussed structural shift in British consumer business of the last two decades, and it rewrote who owns the brand, who owns the factory, and who keeps the margin. Each of those went to a different party, and that is the interesting part.

A note on scope: this piece is written from the consumer-products side, because groceries are where the shift is most visible and because I have sat around this particular table more than once: selling manufacturing capacity, buying it for eleven years, and inside the industry that ran the same play on medicine. The pattern the piece describes, margin migrating to the scarce asset, is much older than my involvement with it and travels well beyond food. The market facts here are public record, sourced at the foot; the finance reading of them is mine, and where something is my judgement rather than evidence, I say so.

The argument in 60 seconds:

Own-label is past half of British grocery on both spend and units, and the shift was driven by three forces in sequence: the discounters, the big four's tiered ranges, and the inflation years that made the habit stick.
The margin did not disappear with the brands. It moved to the two assets that became scarce: the shelf and the shopper data. The factories run on thin, utilisation-driven margins; the squeezed party is the mid-scale brand that owns none of the three.
The specialists' answer is consolidation: Greencore's £1.2bn purchase of Bakkavor in January created a £4bn own-label group, because the counter to buyer scale is supplier scale.
I have sat on three sides of this table: selling capacity at a then NASDAQ-listed clinical group, buying it for eleven years under my own trademark on the modern version of the shelf, and inside generics, which is own-label run on medicine.
The finance disciplines it rewards: price against realised utilisation, treat customer concentration as the shape of the business, mind the cash the contracts consume, and follow the margin with the balance sheet deliberately or not at all.

How half a market changes hands without anyone noticing

Three forces did the work, in sequence.

The discounters set the terms. Aldi and Lidl held around 10% of the British grocery market a decade ago; between them they now hold close to a fifth. Aldi displaced Morrisons from the traditional top four in 2022, and in May this year Lidl passed Morrisons too, leaving the grocer this article treats as the exception sixth in its own market. Their model is own-label at its purest: a short range, a handful of brands for signalling, and everything else under house names. A short range is not a cosmetic choice. It concentrates volume onto fewer lines, which fills a supplier's factory faster, which buys a lower unit cost, which funds the price gap the whole model advertises. Every point of share the discounters took forced the incumbents to respond in kind.

The big four answered with their own ranges. The price wars of the mid-2010s taught the traditional grocers that they could not out-discount the discounters on branded goods, and the reason is worth stating precisely. The list price of a branded case does not vary much between buyers. What varies is everything behind it: rebates, retrospective discounts, joint business plans, listing fees and, now, retail-media commitments. The discounters largely refuse that game, which is why the big four could not win a price war fought on someone else's terms sheet. What they could do was build tiered own-label architectures: entry ranges to defend the price perception, standard ranges to match the brands, premium ranges to trade shoppers up. Anyone who has run pricing will recognise this as price-pack architecture executed at national scale: the same shopper, segmented by occasion and willingness to pay, inside one aisle. The premium tier is the tell that this stopped being a recession behaviour: premium own-label took 7.5% of grocery sales over last Christmas, up from 7.1% a year earlier, and grew 9%. Shoppers did not just trade down to own-label. They traded up within it.

Then inflation locked it in. The 2022 to 2024 cost-of-living squeeze pushed another wave of shoppers across the aisle, and enough of them did not find the quality gap they expected. Behaviour that starts as a saving and survives the comparison becomes a habit. The volume share has kept climbing since.

Running under all of it is a regulatory backdrop worth naming. Since 2013 the Groceries Code Adjudicator has policed how the biggest retailers deal with their suppliers; responsibility for it moved to Defra on 1 July this year. Regulators are rarely created to referee balanced relationships, and the GCA's existence is a fair signpost to where the leverage in this industry sits. The sharper signal is more recent still: in August 2026 the CMA provisionally decided that Aldi and Lidl should be brought under the same Controlled Land Order that has bound the big supermarkets since 2010. The regulator has stopped treating the discounters as insurgents. It now treats them as incumbents.

The factories behind the label

Here is the question almost nobody asks in the aisle: if Tesco's name is on the box and Tesco owns no cereal plant, who made it?

The answer is an industry most shoppers have never heard of. Greencore, the world's largest maker of pre-packed sandwiches, completed a £1.2bn acquisition of Bakkavor in January, creating a group of roughly £4bn revenue, 28,000 people and 36 sites, and the UK market leader in five own-label categories; the CMA required the disposal of one Bristol plant as its price for clearance. A meal deal picked up almost anywhere in Britain has a fair chance of being theirs. Refresco, Dutch-headquartered with more than 80 production sites across three continents, is the world's largest independent bottler, filling retailer-label soft drinks alongside contract runs for the brands themselves; it acquired three US production sites from Coca-Cola and now runs them under contract for it, the same trade in a different costume. McBride, listed in London with revenue above £900m, is Europe's biggest maker of retailer own-brand household and cleaning products. These are serious industrial businesses built almost entirely on making things whose labels belong to someone else.

And the Greencore-Bakkavor merger is worth pausing on, because it is the specialists' answer to the arithmetic that runs through this piece. If the retailer's buying scale is what compresses your margin, the counter-move is to become unavoidable at your end of the chain. Consolidation among suppliers is what the second half of this story looks like.

Their economics are the mirror image of a brand's, and the contracts they live on make the mirror explicit. Own-label supply is typically won at tender and re-won at range review. In fresh, chilled and the commodity-linked categories the terms are often open-book or cost-indexed: the retailer sees the cost build-up and input-price movements pass through by formula, though labour, energy and packaging usually remain a negotiation rather than an automatic pass-through. Much of ambient own-label runs harder still, on fixed-price contracts and e-auctions. A brand owner would call those arrangements commercially unthinkable. A specialist calls it the business, because what it is really selling is not a product but capacity and reliability, and what it is really managing is utilisation: a factory is a fixed-cost machine looking for volume, and the margin is made by keeping the lines full. Scale, efficiency and multi-year agreements matter more than anything a marketing department does, with one caveat every supply-side finance director knows: a multi-year agreement is not the same as multi-year volume. Much of it is an agreement to supply against forecasts, with the orders themselves arriving days ahead.

The risk profile inverts too. A brand's nightmare is losing relevance slowly; a contract manufacturer's nightmare arrives in a single meeting, because one lost range review can knock a hole in a plant's utilisation overnight, and the fixed costs do not leave with the volume.

Who ends up with the margin, and why

Walk the pound through the chain and the shift stops being abstract.

In the branded world, the manufacturer's gross margin carries the whole weight of the model: the marketing that builds the brand, the innovation pipeline, the listing negotiations, and the profit. The retailer takes its margin on top, and then takes a second slice that never shows on a shelf ticket: commercial income, the promotional funding, retrospective rebates and listing fees that flow from supplier to retailer. Finance people treat that stream with respect for a reason. Tesco's 2014 accounting scandal, a £263m profit overstatement that ended in a £129m fine under a deferred prosecution agreement and an £85m investor compensation scheme, was at its heart a commercial-income recognition problem. Any account of grocery margin that ignores supplier income is a marketing story, not a finance one.

In the own-label world the premium does not disappear. It changes hands. The retailer captures the larger share of the economic rent without owning a plant, because it owns the two assets that became scarce: the shelf, which decides what exists commercially, and the shopper data, which tells it which category to enter next and at which price point. That second asset now has a P&L line of its own: retail media. UK retail media adspend is expected to pass £4.8bn this year, and Sainsbury's expects its Nectar360 platform to be delivering at least £100m of incremental profit by March 2027. The data asset is no longer inferred from strategy documents. It is visible, high-margin income. The specialist, meanwhile, keeps a thin, volume-dependent margin because efficient capacity at scale is genuinely hard to build, but rarely unique; there is usually another plant that could want the volume, once it is qualified. And the squeezed party is the mid-sized brand that owns neither the shelf, nor the data, nor the lowest-cost plant: strong enough to carry a marketing budget, not strong enough to be un-delistable.

Put numbers on the triangle and the migration is visible in public accounts. The specialists run mid-single-digit operating margins: Greencore reported 6.5% and McBride about 7% on an adjusted basis in their latest full years. Branded FMCG operating margins sit in the high teens to low twenties. The grocers themselves run at roughly 3% to 4.5%. Read naively, that says the retailer captured nothing. Read properly, it says the retailer's prize arrives as volume, data income and negotiating position rather than as percentage margin, and it points at the measure this whole industry actually runs on: return on capital, not margin. A specialist at a 6.5% operating margin can out-earn a brand at 18% if its capital turns hard enough, which is precisely why the specialists obsess over utilisation and why the brands sold their factories in the first place. Percentage margin tells you who marks up. Return on capital tells you who wins.

Margin sits where the scarce asset sits. It always did. What changed in grocery is which assets are scarce.

The exception that proves the rule

One British grocer took the opposite bet. Morrisons makes a large part of the fresh food it sells in its own factories, and describes itself as the UK's second-largest fresh-food manufacturer: its manufacturing arm, Myton Food Group, runs 19 sites on its own count and employs nearly 6,700 people. Its chief executive calls vertical integration part of the company's DNA, and in fresh categories, where provenance and speed to shelf are genuine advantages, it is a real point of difference.

But look at what Myton has started doing: courting supply deals with other retailers and hospitality customers. The group weighed an approach for the division, though it has said it is not in negotiations to sell it. The one grocer that owns factories is behaving more and more like the specialists, selling spare capacity to whoever brings volume. The exception is converging on the rule, and the rule is utilisation. Nobody else followed Morrisons into ownership for the same reason the brands retreated from it: factories are capital-hungry, and a factory you cannot fill is a liability with your name on it.

Even the most famous refusal in the business has been tested. Kellogg, as it then was, built decades of advertising on the promise that if you do not see its name on the box, it is not Kellogg's in the box. In February 2000, during a slump in profits, it agreed to make five own-label cereals for Aldi in Germany, reported at the time as the first such agreement in its history, and reported in the German trade press as following an Aldi threat to delist leading brands. It scrapped the arrangement within months. The reason it gave is the one that matters to a finance leader: once you make one retailer's label, the others ask for the same terms. The deal economics were never the whole calculation. The precedent was. Many brand owners today also produce retailer-label goods, which is publicly reported across dairy, cereal and household categories; others still refuse, judging that the brand's scarcity is worth more than the volume. Both can be right in their own category. The direction of travel, though, is not in doubt.

I have sat on three sides of this table

Most commentary on own-label is written from one seat, usually the brand's, which is why so much of it reads as mourning. I have sat in three of them, and the view is different from each.

Selling capacity. At Parexel, a then NASDAQ-listed clinical research organisation, I ran operations finance for the Early Phase unit: the five-site network as it then was, across three continents, whose economics were, structurally, a contract manufacturer's. The product was clinical trial delivery rather than sandwiches, but the machine was the same: heavy fixed costs, capacity that perishes if unsold, and revenue won contract by contract at competitive tender. The bids I worked on were built bottom-up, every person costed at a charge-out rate by grade and rolled into FTE-based pricing, and the discipline that mattered most was the one the own-label specialists live by: price against the utilisation you will actually achieve, not the theoretical capacity of the site. The sector was carrying early-phase capacity ahead of demand at the time, so every bid did double duty: winning the work, and earning its share of a fixed-cost base that was not going to be filled by hoping. We geared rates and bids to lift utilisation across the network rather than fill beds with under-priced work. Marginal work into genuinely idle capacity can still pay, and we took some of it deliberately; what does not pay is pricing the whole book as though the plant were full, because the under-priced work is what guarantees it never will be. Over that period the unit recorded the first positive operating income in its history. The unglamorous truth of capacity businesses, which every finance team in one learns early, is that an under-costed contract that wins is not a win. It is a loss with a signature on it.

Buying capacity. For the last eleven years I have been on the other side of the same negotiation, as the owner of a self-funded consumer-products business that designs and sells under its own registered trademark and buys its manufacturing from specialist factories in the Far East. And those years were lived on the modern version of the shelf. My brand sells through online marketplaces, and a marketplace is a supermarket with perfect memory: it owns the shelf, which online is the search page; it owns the shopper data at a depth no grocer of the checkout-scanner era could imagine; it charges for placement the way a grocer charges for a gondola end; and it runs its own label down the same aisle it rents to you. Every discipline this article attributes to grocery suppliers, I have paid at a platform's checkout: the shelf is rented, the data asymmetry is priced in, and the brand premium is re-earned every year or it is gone. So when I say the margin followed the shelf and the data, that is not analysis borrowed from the grocery trade press. It is my own P&L's line items.

The factory side of it taught the same lessons from the buyer's chair. The factory's question is always utilisation, so the volume I could commit shaped everything: unit cost, minimum order quantities, tooling terms, who carries which risk and when title passes. And when COVID collapsed my category by more than 70%, I watched the utilisation logic run in reverse, from both sides of the contract at once: my volumes fell away, my suppliers' lines emptied, and survival became a negotiation between two parties who each needed the other to stay solvent. We restructured costs, renegotiated supply, and traded through to the rebound. Nothing teaches you the fixed-cost arithmetic of a factory faster than being the volume it just lost.

The generics mirror. Before any of that, I worked inside the industry that ran this exact play on medicine. At Mayne Pharma I led group performance reporting for a global generics business, and generics are pharmaceutical own-label in all but name: the same active ingredient, demonstrated bioequivalent, with the premium stripped out and the price set at the level the scarce asset justifies once the patent no longer makes it scarce. The whole sector exists because a legal monopoly expires on a date, and margin rebases fast to whoever has the lowest-cost compliant plant and the distribution once effective exclusivity ends. And I had already seen the mirror image in the years just before, at Chiron, where the scarce asset was the customer relationship itself: across a 17-country distributor network, we flipped the larger markets from third-party distributors into direct country affiliates, taking back the route to market and the margin and the data that sat on it, and the markets business achieved record revenue and contribution growth of 20% on the prior year. That is the same strategic move as a retailer building an own-label tier or Morrisons buying factories: vertical integration toward whichever asset has become scarce, executed in the opposite direction. Watching both directions from the finance seat is what convinced me the asset-scarcity frame, and not any sentiment about brands, is what actually predicts where margin goes. That is judgement, not proof.

What a finance leader should take from it

Strip the groceries out and this is a capital-allocation story with six lessons.

First, locate the scarce asset before you price anything. Brand equity, efficient capacity, shelf access, customer data: margin migrates to whichever of these is genuinely hard to replicate in your market, and it migrates away from the ones that have become commodities. A finance function that still budgets as if the brand premium of 2010 were intact is planning for a market that no longer exists. The test is brutally practical: if you disappeared from the market tomorrow, what would your customers genuinely struggle to replace? That is the asset. Price and invest from there.

Second, if you sell capacity in any form, factory hours, clinical beds, warehouse space, professional time, price against realistic utilisation, never against theoretical capacity. Fixed overhead does not care what the site could do at full stretch; it has to be recovered across the volume that actually arrives. Every capacity business that quietly bleeds does it the same way: winning work at prices that only make sense if the plant were full, which the under-priced work itself guarantees it never profitably will be. The most familiar version of this business is a hotel. A room-night perishes at midnight, occupancy is utilisation, and the discounting decision is the same one: a cheap last-minute room that clears its marginal cost is worth having, but a rate card built on discounted rooms never covers the building. Revenue per available room, the metric the industry runs on, is just price multiplied by utilisation in one number, which is why every capacity business eventually reinvents it under its own name.

Third, treat dual-tracking as a utilisation decision with a brand consequence, not a philosophy. Filling spare capacity with someone else's label can be the difference between a plant that washes its face and one that does not; it can also finance your competitor's attack on your own range. The right answer is category-specific and it moves as relative scarcity moves. Model both sides of it honestly, including the version of the analysis where your brand's scarcity is worth more than the volume, which is the calculation behind every famous refusal, and the version Kellogg discovered in 2000: the cost of the first own-label deal is every customer who then asks for the same.

Fourth, if your customers are consolidating power, your volume visibility is your risk register. The specialists live or die on multi-year volume from a handful of buyers; so, increasingly, does any supplier to modern retail, and so does a contract manufacturer of any kind. The finance job is to know exactly how much of the plan rests on relationships the other side can re-tender at the next range review, and to price, invest and borrow accordingly. Customer concentration is not a note in the accounts. It is the shape of the business.

Fifth, mind the cash the contracts consume, because own-label supply is a working-capital model as much as a margin model. The terms run long, there is no brand-funded promotional income coming back, the inventory carries a retailer's specification that no one else will buy, and the tooling capex is committed to a single customer. A lost range review does not just take the volume; it strands the stock and the capacity the volume was carrying. I have run that arithmetic with my own capital on both sides of a 70% collapse, and it is why capacity businesses live on thirteen-week cash forecasts, not annual budgets.

Sixth, when the margin has moved, follow it with the balance sheet, deliberately or not at all. The half-committed position is the expensive one: a brand spending to defend a premium the market has stopped paying, or a retailer running sub-scale factories it cannot fill, or a manufacturer carrying capacity for volume it merely hopes will arrive. Morrisons owning factories at genuine scale is a strategy. The Chiron flips into direct markets were a strategy. What sits between, ownership without the scarce asset that justifies it, is where capital goes to underperform.

And because a frame is only useful if it changes what you examine first: walking into any business that sells or buys capacity, the first week's questions are always the same five. The contract maturity ladder and what share of revenue re-tenders inside eighteen months. The top five customers' share of contribution, next to their re-tender dates. Contribution per constrained line-hour, not per unit. The gap between planned and realised utilisation, line by line. And the working-capital swing if the single biggest range review goes the wrong way. None of that needs a strategy document. All of it decides whether the strategy document is fiction.

Half of British grocery changed hands in plain sight because each party optimised its own P&L one decision at a time. That is usually how structural shifts happen, and it is why the finance seat, which sees all three P&Ls meet, is where I would expect them to be spotted first.

The reason I write about this from three seats rather than one is that these are the problems I am asked into: cash and working-capital recovery where the terms have moved against the business, cost programmes that have to survive contact with a factory, customer-concentration risk that is not yet in anyone's risk register, and post-acquisition integration where two cost bases have to become one. The frame in this article is how I decide where to look first.

In brief

How much of UK grocery is own-label? Just over half, on both serious measures: 52.2% of take-home grocery spend in January 2026, the highest Worldpanel by Numerator has recorded, and past half of all units for the first time on Circana's 230-category packaged measure, where the value share is 44%. The bases differ, which is why quoted numbers differ. Premium own-label took 7.5% of sales over Christmas 2025 and grew 9%.

Who actually makes it? Mostly specialist manufacturers: Greencore, a roughly £4bn group after completing its £1.2bn acquisition of Bakkavor in January 2026; Refresco, the world's largest independent bottler with more than 80 sites; and McBride, Europe's biggest own-brand household-products maker. Contracts are typically tendered and re-won at range review, open-book or cost-indexed in fresh and commodity-linked categories, and it is utilisation, not marketing, that makes the margin.

Is any grocer still vertically integrated? Morrisons, which makes a large part of the fresh food it sells through Myton Food Group and describes itself as the UK's second-largest fresh-food manufacturer; it now also sells manufacturing capacity to other customers, and says it is not in negotiations to sell the division.

Why does it matter to a finance leader? Because the episode shows margin migrating to the scarce assets, the shelf and the shopper data, and away from mid-scale brands. The disciplines it rewards:

price capacity against realised utilisation, never theoretical;
treat customer concentration as the shape of the business, not a note in the accounts;
mind the working capital own-label contracts consume, on thirteen-week cash visibility;
follow the margin with the balance sheet deliberately, or not at all.

The same dynamic already runs online, where the marketplace owns the shelf and the shopper data, and charges for both.

Sources

Worldpanel by Numerator (via The Grocer, April 2026): own-label at 52.2% of UK take-home grocery spend, four weeks to 25 January 2026, its highest recorded share; grocery volume share past half for the first time.
Circana (April 2026): private label past 50% of units for the first time across the UK's ~230 packaged FMCG categories, with a 44% value share on that basis.
Worldpanel by Numerator / Kantar (January 2026): premium own-label at 7.5% of sales, up from 7.1%, growing 9%, four weeks to 28 December 2025.
Worldpanel by Numerator (12 weeks to 17 May 2026, via The Grocer): Lidl 8.6% vs Morrisons 8.3%, Lidl becoming the UK's fifth-largest grocer; Aldi's 2022 entry to the top four.
Greencore and Bakkavor announcements (January 2026): completion of the £1.2bn acquisition on 16 January 2026; combined group of roughly £4bn revenue, 28,000 people, 36 sites, market leadership in five own-label categories; CMA-required disposal of the Bristol soups and sauces site. Greencore FY25: adjusted operating margin 6.5%.
Refresco company information (2026): more than 80 production sites across three continents; completion of the SunOpta acquisition, May 2026; acquisition of three US production sites from The Coca-Cola Company with ongoing contract manufacturing.
McBride plc FY25 results: revenue £926.5m, adjusted operating margin about 7%, Europe's leading manufacturer of retailer own-brand household and cleaning products.
Wm Morrison Supermarkets accounts and Morrisons publications (2022-2025): Myton Food Group, 19 production sites on Myton's own count, ~6,700 employees; Morrisons' own descriptions of its fresh-food manufacturing share; January 2026 statement that it is not in negotiations to sell Myton.
The Irish Times (25 February 2000) and Marketing Week (February and July 2000): Kellogg's five-cereal own-label agreement with Aldi in Germany under the Gletscher Krone label, and its termination within the year, citing pressure from other retailers for equivalent terms.
Tesco PLC / Serious Fraud Office / FCA (2014-2017): the £263m commercial-income overstatement, the £129m deferred prosecution agreement and the £85m investor compensation scheme.
UK Government (2026): the Groceries Code Adjudicator's transfer to Defra, 1 July 2026; CMA provisional decision of August 2026 to designate Aldi and Lidl under the Groceries Market Investigation (Controlled Land) Order 2010.
Advertising Association / WARC and Sainsbury's statements: UK retail media adspend expected to exceed £4.8bn in 2026; Nectar360 targeted to deliver at least £100m incremental profit by March 2027.

The views here are my own. They do not represent the position of any current, former or future employer or client, and nothing here draws on confidential information.