Two credible organisations published consolidation counts for the alternative protein sector in 2026. Green Queen’s tracker says at least 82 businesses have been acquired, merged, entered insolvency or shut down since September 2024, with 54 of those in the most recent twelve months. The Good Food Institute’s 2026 State of the Industry report says at least 19 plant-based companies were bought out or acquired during 2025, and gives no number at all for the ones that failed.
Those two figures are not in conflict. They are not even measuring the same thing. But an operator or investor reading either in isolation gets a materially wrong picture of the market, and neither publisher states its inclusion rule anywhere near its headline number.
When we align the two datasets on unit and window as closely as the published material allows, the gap between them shrinks from roughly 4× to something closer to 2× — and the residual is explained almost entirely by scope, not by disagreement about what happened.
Methodology note
Everything below comes from four published documents, all of which we fetched and read in full: Green Queen’s September 2025 tracker article, its July 2026 deep dive, its April 2026 write-up of GFI’s report series, and GFI’s own State of the Industry plant-based report PDF. Where a publisher’s summary text and its own itemised breakdown disagree, we use the breakdown and flag the discrepancy. That is house rule.
Neither publisher provides a downloadable event-level dataset. Green Queen presents its itemised list through embedded Infogram graphics, which we were unable to retrieve. All arithmetic attributed to us below is arithmetic we performed on figures the publishers state in prose, and we mark it as ours.
What Green Queen actually publishes
The tracker exists in at least three published versions, each with a different window and a different headline:
| Published | Stated window | Headline count | Embedded graphic title |
|---|---|---|---|
| 4 Sep 2025 | 1 Sep 2024 to 31 Aug 2025 | “more than 40” | “Alt Protein Sector M&A Activity” |
| 19 Dec 2025 | Not stated in body text | No count in body text | “Alt Protein Sector M&A Activity – Last 14 Months” |
| 21 Apr 2026 | “since September 2024” | “more than 70” | “Alternative Protein Sector M&A Activity” |
| 9 Jul 2026 | “since September 2024” | “at least 82” | “Alternative Protein Sector M&A Activity” |
The September 2025 article is the only one that publishes full breakdowns, and this is where the brief we started from turns out to be wrong in an instructive way. The regional split frequently quoted alongside the 54 figure — Europe 23, North America 16, Asia-Pacific 5 — does not belong to the 54 dataset at all. It belongs to the September 2025 article, covering September 2024 to August 2025.
And it sums to 44, not 40. So does the technology split in the same piece:
| Breakdown | Categories as published | Sum |
|---|---|---|
| Geography | Europe 23, North America 16, Asia-Pacific 5 | 44 |
| Technology | Plant-based 32, fermentation 7, cell cultivation 3, molecular farming 1, blended protein 1 | 44 |
| Event type | Acquisitions 24, closures 11, liquidations 4, bankruptcies 2, merger 1, plus “several notable asset sales” | 42 plus asset sales |
Two independent breakdowns agree at 44. The headline says “40+”. There is no missing region and no unallocated remainder — the article accounts for every event it counts. The headline is simply rounded down, and rounded down in a way that matters, because of what happens next.
The July 2026 piece states that the 54 events of the most recent twelve months represent “a 35% increase from the September 2024-25 period”. Our arithmetic: 54 divided by 1.35 is exactly 40. The growth claim is therefore calculated against the rounded headline, not against the 44 the earlier article’s own tables support. Against 44, the increase is 22.7%.
The cumulative figure does not reconcile either. The two windows overlap — September 2024 to August 2025 against “the last 12 months”, also described as “since July 2025” — so they cannot simply be added. But 44 plus 54 is 98, and the published cumulative total is 82. Sixteen events would have to fall inside the two-month overlap for those to be consistent. No reconciliation is published.
Practical consequence: the “35% increase” is the single most quotable number in the dataset and it is the one most sensitive to a rounding decision the publisher made a year earlier. If you are citing sector deterioration in an investment committee memo, cite 54 and 44 and let the reader do the division.
What GFI actually publishes
GFI’s plant-based State of the Industry report is explicit and narrow. Its exact words, which appear twice in the document:
At least 19 plant-based companies were bought out or acquired, reflecting a consolidation of product portfolios, technology stacks, and brand equity.
Immediately after, in the same passage: “multiple plant-based companies paused or ceased operations after struggling to secure follow-on financing.” Multiple. Not a number.
That asymmetry is the whole methodological story on GFI’s side. GFI publishes a count for acquisitions and a word for failures. Green Queen publishes a single total that sums both. A reader who treats 19 and 82 as rival estimates of the same quantity has misread both.
GFI’s named examples are also a different class of company from Green Queen’s: v2food’s acquisition of Daring Foods, and Vivera’s (JBS) acquisition of The Vegetarian Butcher from Unilever. These are trade sales between operating businesses, not distress.
One clarification we owe our own readers. LifeMate Tech previously reported that GFI named five consolidation deals alongside $881m of 2025 funding. That five came from GFI’s February 2026 funding commentary as relayed by a trade title, not from the State of the Industry report, and we could not re-fetch that page for this piece. What we can verify directly is the report itself, which publishes at least 19. Five was a list of illustrative deals in a funding release. It was never GFI’s count, and we should not have presented it adjacent to a total.
Practical consequence: if you need a defensible acquisition count for 2025 plant-based alternative protein, 19 is the number with a named publisher behind it. Five is a sample.
The comparison
| Dimension | Green Queen | Good Food Institute |
|---|---|---|
| Publisher type | Trade publication, editorial research | Non-profit think tank, annual report series |
| Latest headline | At least 82 since Sept 2024; 54 in last 12 months | At least 19 |
| Window | Rolling, restated at each republication; Sept 2024 onward | Calendar year 2025 |
| Unit counted | “Businesses” experiencing an “instance” | Companies “bought out or acquired” |
| Inclusion rule | Acquisition, buyout, merger, insolvency, closure, brand discontinuation, asset sale | Acquisition or buyout only |
| Exclusion rule | Liquidation and bankruptcy counts exclude businesses later acquired or shut, per the article’s own footnote | Not stated |
| Failures counted | Yes, summed into the headline | Not counted; described as “multiple” |
| Technology scope | Plant-based, fermentation, cultivated, molecular farming, blended | Plant-based report counts plant-based only |
| Geography | Global; Europe, North America and Asia-Pacific itemised for one window only | Global |
| Size floor | Not stated | Not stated |
| Methodology published | Partially: breakdowns for one window, footnote on double-counting | No inclusion criteria stated for the count |
| Itemised list available | Yes, via embedded graphics | No |
Now align them. Green Queen states that 65% of the 54 recent events were acquisitions. Our arithmetic: that is approximately 35 acquisitions, with roughly 19 insolvencies, closures or brand discontinuations. So the like-for-like comparison is not 82 against 19. It is approximately 35 acquisitions across all technologies over roughly July 2025 to July 2026 against 19 plant-based acquisitions in calendar 2025.
That residual gap is explained by two things and no mystery: Green Queen includes fermentation, cultivated and molecular farming companies, and its window runs seven months later. It is possible the two datasets are close to consistent. It is not possible to prove it without the event lists.
Practical consequence: a 4× discrepancy in the headlines becomes an unremarkable scope difference once the units match. Any market thesis built on “the failure rate quadrupled” is resting on the arithmetic of two different questions.
Three events that show why the unit matters
We verified three named events against sources outside both trackers.
Noumi. In July 2026 the Australian owner of Milklab agreed a buyout by Arrovest, the Perich family vehicle, headlined at A$737m ($518m). Read the structure and the headline dissolves. Arrovest already held about 52.5% of the shares. The transaction values Noumi’s equity at A$34.2m ($23.9m); the rest is a note redemption against debt obligations of around A$703m, including an A$610m mandatory cash redemption due the following May. Noumi guided to adjusted EBITDA of A$61–63m for FY2026, up year on year, and reported Milklab sales up 5.5%. This is a majority holder taking a profitable business private to refinance a bond. It is not a change of control, not distress, and not comparable to a startup liquidation — yet in a count of “instances” it scores one, exactly as a liquidation does.
SWAP Food. The Paris company formerly known as Umiami was placed into judicial liquidation by the Paris Economic Activities Court and ceased activities on 15 June 2026. It had raised more than €100m ($116m) in debt and equity since 2020, carried €16.5m ($19.1m) of debt at December 2024, and posted €4.5m ($5.2m) of 2024 revenue, most of it from what the accounts list as “ancillary activities” rather than finished product. Note that it is French, not Dutch, and that its US subsidiary was put through a separate asset sale process expected to conclude 1 July — which is potentially a second countable instance from one failure.
Meatable. The Dutch cultivated pork company wound down in December 2025 after failing to raise, having taken $95m to date. Agronomics wrote its £7.9m ($10.5m) investment, carried at £11.9m ($15.9m), to zero. Meatable is the cleanest illustration of the double-counting risk: during the same tracking period it acquired Uncommon Bio’s cultivated meat platform and staff. One company, plausibly two instances, pointing in opposite directions.
Practical consequence: a headline “consolidation event” number sums a debt refinancing at a profitable listed company, a court-ordered liquidation, and an acquisition made by a company that later collapsed. For diligence, the event type and the consideration structure carry all the information. The count carries almost none.
The denominator problem
A count of failures without a population is not a rate, and no rate claim in this sector currently has one.
GFI maintains an alternative protein company database, which is the obvious candidate. It is delivered as an embedded Airtable view; GFI states plainly that it “is not an exhaustive list”, and publishes no record count on the page. We could not extract a total.
The one population figure we found anywhere is for fermentation: GFI’s fermentation report counts more than 163 specialised companies active in 2025. Set Green Queen’s seven fermentation events for September 2024 to August 2025 against that and you get roughly 4% — which is indicative only, because the windows differ and the two counts were built for different purposes. It is nonetheless the only denominator-anchored figure available in this entire debate, and it does not look like a sector-wide collapse.
For plant-based, the segment carrying 72% of Green Queen’s recent events and all 19 of GFI’s acquisitions, no published population figure exists that we could find.
Practical consequence: until someone publishes a denominator, “54 companies failed” and “19 were acquired” are both absolute counts in a sector whose company formation rate is unmeasured. Neither supports a statement about the probability that a given portfolio company survives.
The counter-argument
The serious objection is that we are grading a trade publication against an academic standard it never claimed, and that a directional signal from a consistently applied imperfect count is more useful than no count at all.
That objection has real force, and on the central point Green Queen wins outright. It publishes its itemised list. Most consolidation trackers in this sector do not. It publishes a technology split, a regional split, an event-type split, and — notably — a footnote stating that its liquidation and bankruptcy figures exclude businesses later acquired or shut. That footnote is an explicit anti-double-counting rule, which is more methodological disclosure than GFI offers for its 19, where no inclusion criteria are stated at all. On disclosure, the trade publication is ahead of the think tank.
The direction is also almost certainly right. Every independent event we checked confirms genuine distress, and the funding backdrop GFI reports — cultivated meat and seafood at $73.9m in 2025 against $139m in 2024, fermentation at $357m against $651m — is consistent with a sector shedding companies.
Where we think the objection fails is at the specific claim that circulates most: the 35% increase. A directional signal is defensible. A precise growth rate calculated against a rounded prior-period headline, when the publisher’s own tables support a different base, is not directional — it is a number that will be quoted to two significant figures by people who never see the tables. The fix is trivial: state the base.
What we could not establish
- The itemised event lists for either dataset. Green Queen’s are published only inside embedded Infogram graphics, which we could not retrieve. We therefore could not verify any breakdown against the underlying records, and could not confirm whether the 82 total contains companies counted twice.
- How Green Queen reconciles 44 plus 54 to 82. The overlap between the two windows is the obvious explanation but no reconciliation is published, and 16 events would need to sit in a two-month overlap.
- Whether the 82 figure includes the September 2024 to August 2025 events on the same basis. The earlier article counted “major” ventures; the later ones count “businesses” and “instances”. Whether the size threshold moved is not stated.
- Any population denominator for plant-based companies. GFI’s company database publishes no record count and disclaims exhaustiveness.
- GFI’s inclusion criteria for its 19. The report gives the figure and two examples. It does not define “bought out or acquired”, state a size floor, or say whether asset-only purchases from insolvent estates qualify.
- The five-deal figure in our own earlier coverage. We could not re-fetch the trade report it came from, and have flagged above that it was a list, not a count.
- Green Queen’s technology percentages for the recent 54 sum to 98% (plant-based 72%, cultivated 15%, fermentation 11%), leaving roughly one event unaccounted. We could not determine what it is.
What to watch
- Whether Green Queen publishes a base figure alongside its next growth claim. If the September 2026 update states the prior-period count explicitly — 40 or 44 — the most-quoted number in this dataset becomes checkable. This is a testable claim with a date attached.
- Whether GFI’s 2027 State of the Industry publishes a closure count. GFI counted acquisitions and described failures as “multiple”. A number on the failure side would make the two datasets directly comparable for the first time.
- Whether the Noumi transaction completes in November as scheduled, and how it is classified. If a completed debt refinancing at an EBITDA-positive listed company appears in a consolidation total without qualification, that settles the question of whether “instance” is a useful unit.
- Whether SWAP Food’s US subsidiary sale generates a second countable event. One failure producing two tracker entries is the double-counting hypothesis in its clearest form.
- Whether anyone publishes a denominator. The fermentation figure of 163-plus active companies is the template. A comparable plant-based population count would convert every number in this piece from an anecdote into a rate.
Our related coverage: the peer-acquirer pattern in 2025 deals, the supply-layer shutdowns in cultivated meat, and how funding structure narrowed underneath all of it.