US Franchising’s Missing 18,881 Units
By Sean Goldsmith
Franchising is still expanding. But a striking gap between successive industry reports raises questions about what its headline growth figures really measure
Look, the official story remains upbeat. In February 2025, the International Franchise Association and FRANdata said the United States was on course to finish the year with 851,402 franchise establishments, an increase of more than 20,000 units. Employment was expected to pass nine million, franchise output was forecast to reach $936.4 billion and the sector was once again presented as outperforming the wider economy.
Twelve months later, however, the 2026 outlook began from a very different place. Its estimate for the end of 2025 was 832,521 establishments.
That is 18,881 fewer establishments than the previous report had projected.
The industry did not necessarily lose 18,881 operating franchises, and it would be irresponsible to describe the gap as 18,881 failures. But when a much-trumpeted expansion forecast reappears one year later almost 19,000 units lighter, operators, lenders and prospective buyers are entitled to ask what changed, and why the reconciliation is not sitting beside the new growth headline.
Speculation here is that the answer may be less dramatic than a mass closure event, but more important than a simple forecasting miss. The two reports appear to use materially revised historical estimates and a different sector structure. The 2025 outlook divided franchising into eight broad business lines. FRANdata’s 2026 material describes 11, including newly separated categories such as automotive, child services, and health and wellness.
Earlier years were also restated. The 2025 report estimated 830,876 establishments for 2024, while the following report put 2024 at 821,837, a reduction of 9,039 before the contested 2025 forecast is even considered. Its 2023 figure also moved from 811,247 to 808,911.
One plausible explanation, therefore, is a substantial reclassification or rebasing of the dataset rather than thousands of units suddenly disappearing. That would be legitimate analytical work. But if definitions or coverage changed, the industry needs to say so plainly, because otherwise two apparently comparable totals are being used to tell two quite different growth stories.
Supporters would counter that forecasting is not bookkeeping. The 2025 number was always labelled a projection, prepared amid assumptions about lower interest rates, improving consumer confidence and easier financing that did not necessarily arrive as expected. The 2026 outlook itself describes 2025 as a year of macroeconomic turbulence and recalibration, with constrained consumer spending, difficult credit conditions and continued pressure on unit economics.
FRANdata also tracks roughly 9,000 active franchise brands using proprietary databases, external economic information and qualitative input from franchisors, franchisees, lenders and specialists. Estimates built from a changing commercial universe will inevitably be revised as better information becomes available. That does not make the work misleading, and it certainly does not prove that the headline growth narrative was fabricated.
But let’s be real. A revision this large should not be buried beneath the next optimistic forecast.
The missing number is not the only one that moved
The establishment gap is the eye-catching figure, but the same pattern appears elsewhere.
The 2025 outlook forecast 9,005,757 franchise jobs by year-end. The 2026 report estimated 8,788,313, roughly 217,000 fewer. Output was forecast at $936.4 billion but subsequently estimated at approximately $907.3 billion, a difference of about $29 billion. Franchise GDP moved from a projected $578 billion to an estimated $549.9 billion.
Those changes do not mean the sector shrank in absolute terms. The newer report still says establishments, employment, output and GDP grew during 2025. What they do show is that the scale and speed of the expansion were substantially weaker than the prior headline suggested or that the measurement framework changed enough to make the comparison unreliable.
Either interpretation deserves a proper explanation. This matters because these forecasts are not harmless industry cheerleading. They are reused in franchise sales presentations, lender materials, political lobbying, investment commentary and stories telling prospective owners that franchising is accelerating. One franchise-investment article has already turned the new 2026 projection into a claim that more than 12,000 new locations will open, describing those units as 12,000 new owners realising their entrepreneurial ambitions.
Yet the outlook reports a net change in establishments.
It does not, at least in the public headline material, tell us how many units opened, how many closed, how many changed hands, how many franchise agreements were terminated or how many sold territories never became operating businesses.
Treating net establishment growth as synonymous with successful new owners risks converting an economic estimate into recruitment copy. Anyone who has operated units knows why the distinction matters. A network can open 100 locations and close 90, then announce ten units of net growth. Another can open 20 and close five, producing 15 units of net growth.
The second system has the smaller development machine but potentially the healthier operating story. Transfers complicate the picture further. A location can remain in the establishment count even when the original franchisee has lost money, sold under pressure or handed the business back. A franchisor can also report hundreds of signed agreements while only a fraction reach opening day.
None of that is captured by a top-line net unit figure. That is the question behind the missing 18,881:
Are industry growth statistics measuring the creation of durable franchise businesses, or mainly the changing number of branded establishments believed to be operating at a particular moment?
Those are not the same thing.
FRANdata’s 2026 material offers one useful clue. Its proprietary Historical Unit Success Rate reportedly fell to approximately 94.2% in 2025, from 95.9% in 2021, with the firm attributing part of the decline to macroeconomic pressure and the deliberate pruning of underperforming units. That is precisely the sort of operational reality the headline establishment forecast should be reconciled against.
Pruning may strengthen a system. Closing weak stores can protect franchisees elsewhere, improve average performance and stop a franchisor throwing support at locations that no longer work. But “pruning” also means somebody’s unit shut, was terminated or ceased to qualify for inclusion. The industry cannot celebrate gross development ambition while treating the removal of weak operators as a technical footnote.
Growth is not the same as franchisee prosperity
The bigger problem is that establishment counts say remarkably little about whether the people funding those establishments are getting richer.
Franchise output can rise because prices increased. System sales can grow while labour, rent, food, insurance, technology charges, debt service and royalties consume the gain. A franchisor can expand its royalty base even as franchisee free cash flow deteriorates. An existing operator can acquire several struggling locations, producing a more concentrated network without creating a single new entrepreneurial success story.
FRANdata says 19.3% of franchisees now control 58.8% of franchise locations, which suggests that consolidation among operators is becoming central to the industry’s growth model.
That may reflect capable multi-unit owners deploying capital efficiently. Others may argue it also reflects a market in which scale is becoming necessary merely to survive. This is where the next IFA and FRANdata response needs to be more forensic than promotional.
They should be asked:
Whether the 2025 figures in the 2026 outlook are directly comparable with the previous year’s projections.
Whether definitions, classifications, brand coverage or data sources changed.
How much of the 18,881-unit difference represents forecast error and how much represents revision to the historical base.
How many gross openings, closures, transfers and terminations occurred during 2025.
Why the estimate for 2024 fell by more than 9,000 establishments between reports.
Whether the full 2026 analysis contains a reconciliation table.
Whether any reliable industry-wide dataset tracks franchise agreements sold but never opened.
Without those answers, the fairest conclusion is not that America’s franchise boom collapsed.
It is that the public numbers do not yet allow outsiders to determine exactly how much of the projected boom occurred, how much was revised away and how much may be an artefact of changed measurement.
For franchisors, this should be a warning against using macro growth forecasts as a substitute for disclosing system health. Candidates need openings, closures, transfers, terminations, resale performance and time-to-open data—not just a chart showing that franchising is a trillion-dollar economic force.
For operators, the lesson is blunter. A growing industry does not guarantee a growing bank balance. National establishment totals will not tell you whether your labour model works, whether local demand supports another unit or whether fellow franchisees are quietly exiting. For lenders and the wider industry, the credibility issue is simple.
Revisions are normal. Unexplained revisions create suspicion.
When the sector asks policymakers and investors to rely on its economic contribution, it should make the bridge between one year’s forecast and the next year’s estimate easy to inspect.
Franchising may indeed add another 12,000 establishments in 2026. The model remains active, adaptive and capable of producing strong businesses. But before the industry celebrates the next boom, it ought to explain what happened to the last one.
Operators who opened, closed, sold or took over units during 2025: did the market feel like a boom, a shakeout or something much messier? We want the unfiltered version.