Franchise vs Independent in 2026: The Fee Stack, the Churn, and the Loan That Outlives the License

The article you have already read

There is a standard piece on this subject. You have seen it. It runs a table down the page with a column called "Franchise usually wins?" and fills it with Yes, Sometimes, Often no. Hotels: franchise. QSR: franchise. Car wash: independent. Self-storage: independent. Then it says something sensible about revenue not being profit, lists a dozen red flags, and stops.

Nothing in that table is wrong, exactly. It is just not an answer. It is a summary of what experienced operators already believe, restated with confidence and no arithmetic. Every dollar figure in those articles arrives as "FDD-derived sources show" or "public summaries report," with no document date, no state registry, no Item number. You cannot check any of it. You certainly cannot underwrite from it.

So we went and got the documents.

Over three research passes we pulled Franchise Disclosure Documents for 27 brands across 11 asset categories, extracted Items 5, 6, 7, 11, 12, 17, 19 and 20, graded the earnings disclosures, computed unit churn, read the current SBA rulebook against the current SOP, and went back through the peer-reviewed literature on whether franchising actually reduces failure risk. Some of what we found confirms the conventional table. A fair amount of it does not. And three things nobody writes about turn out to matter more than most of what everybody writes about.

Here they are up front.

One. The royalty is not the fee. The fee is the stack: royalty plus brand fund plus local marketing minimum plus technology plus reservation and loyalty charges. In hotels that stack reaches roughly 9.5 to 10 percent of gross room revenue before you allocate a cent of loyalty cost (1). In the sandwich and wing brands it lands between 10 and 11.5 percent. Against a category net margin in the mid to high teens, the brand is taking something close to half of an average operator's profit before rent and before debt service.

Two. Item 20 is the only part of the FDD that reports what happened to people who did what you are about to do. It is public, it is quantitative, and it is almost never quoted.

Three. The franchise agreement expires before the loan does. A 20-year hotel or QSR license against a 25-year SBA real estate amortization leaves five years of debt with no contractual right to the brand that generated the cash flow. Several of the largest systems in America grant no renewal right at all. That is the risk nobody prices, and it is the reason this article exists.

What the documents actually say: 18 numbers

1. Hampton by Hilton charges a 6 percent royalty plus a 4 percent program fee on room revenue (2).

2. Comfort Inn charges 6 percent plus a 3.5 percent system fee, and publishes no Item 19 at all (3).

3. Jersey Mike's runs 6.5 percent royalty plus 5 percent advertising, the heaviest straight stack in the QSR set (4).

4. McDonald's raised its US royalty to 5 percent for new units effective January 1, 2024, its first increase in nearly 30 years. Legacy units stay at 4 percent. Rent runs separately at roughly 8.5 to 15.75 percent of sales (5).

5. Wingstop's 2025 FDD reports median net sales of $2,001,753 across the 1,804 restaurants open for the entirety of fiscal 2024 (6).

6. Primrose Schools' 2026 FDD reports top-quartile EBITDA of $768,966. That figure comes from 89 franchised schools out of 558 open at year-end 2025 (7).

7. Chick-fil-A's Operator Program charges a $10,000 initial fee, a 15 percent service fee on gross sales, and a 50 percent split of pre-tax profit. Chick-fil-A owns or controls the restaurant (8).

8. 7-Eleven's model routes roughly half of store gross profit to the franchisor, which owns or leases the store (9).

9. Planet Fitness reported one termination and two other closures across 2,298 franchised units in fiscal 2024, alongside 270 transfers (10).

10. Wingstop reported zero terminations, zero non-renewals and zero reacquisitions in fiscal 2024 while growing domestic franchised units from 1,877 to 2,154 (10).

11. Four brands in our set publish no financial performance representation whatsoever: Servpro, Comfort Inn, TravelCenters of America's TA Express, and, on the most defensible reading, historically The UPS Store (11).

12. Dunkin franchise agreements require refurbishment at five years and again at fifteen years after opening, and impose a five-mile post-term non-compete (12).

13. Taco Bell's franchise agreement grants no territorial exclusivity and no contractual renewal right (13).

14. Hotel franchise agreements from both Hilton and Marriott are terminable at will by the franchisor on franchisee bankruptcy or insolvency (14).

15. The SBA Franchise Directory did not exist between August 1, 2023 and June 1, 2025. It was eliminated under SOP 50 10 7 and reinstated under SOP 50 10 8 (15).

16. Effective July 4, 2026, SBA Policy Notice 5000-879058 decouples the programs: a qualified borrower can carry up to $5 million of 7(a) and a separate $5 million of 504, for $10 million combined. The individual 7(a) cap stays at $5 million (16).

17. In the best microdata study on the question, roughly 62 percent of franchise startups survived against roughly 68 percent of independents, and average franchise profit was negative (17).

18. In March 2026 the FTC obtained a $17 million judgment against Xponential Fitness, which it called the largest consumer redress in agency history for an alleged Franchise Rule violation (18).

Part one: the fee stack

Ask a franchise development officer what the royalty is and you get one number. Ask what the total contractual fee load is and you get a pause.

Here is why the pause matters. Take a Jersey Mike's at the brand's reported 2025 average of roughly $1.367 million (4). The royalty is 6.5 percent, which is $88,855. The advertising contribution is 5 percent, which is another $68,350. That is $157,205 leaving the store before rent, before labor, before food cost, before the loan. On a shop that might clear $150,000 to $200,000 in restaurant-level cash flow in a good year, the brand's cut and the owner's cut are roughly the same size.

Now do it in hotels. Hampton by Hilton is 6 plus 4 (2). Comfort Inn is 6 plus 3.5 (3). On a select-service hotel doing $3 million in room revenue, that is $285,000 to $300,000 a year to the flag, and that is before you allocate loyalty program reimbursements, central reservation charges, and the marketing assessments that arrive under other names. Franchise cost in hotels is routinely underwritten at 9.5 to 10 percent and behaves like 12 to 14 once everything lands.

We built a screening metric for this and called it the Fee Load Index. It is simple. Total contractual recurring fees as a percent of gross revenue, divided by the category's typical operating margin. It answers one question: what share of an average operator's profit does the brand take before rent and debt?

The answer, in most brand-driven categories, is somewhere between a third and a half.

That is not automatically a bad deal. If the brand delivers a 40 percent revenue lift over what you could generate with your own sign on the building, a 10 percent fee load is the cheapest capital you will ever raise. The problem is that almost nobody tests the counterfactual. They compare the franchised pro forma to zero, not to the independent case.

The worst fee-load math in the set is not in hotels or QSR. It is in restoration. Servpro runs a sliding royalty from 3 to 10 percent depending on volume, plus roughly 3 percent national advertising, plus mandatory equipment sourcing (19). A franchise law firm modeled a startup Servpro at $350,000 in sales routing approximately 22 to 31 percent of gross to Servpro-related costs before a single hour of labor or a dollar of rent (19). Established franchises at $800,000 in sales drop to roughly 16 to 22 percent. That is the kind of number that decides whether the business works, and it appears in no category-verdict table anywhere.

The lightest load, unsurprisingly, is in the categories where the brand does the least work. Tommy's Express runs 4 percent royalty, a 1 to 3 percent brand fund and a $98 monthly technology fee (20). Against site-level EBITDA that can reach 45 to 63 percent in a well-located express tunnel (21), a 5 to 7 percent fee load is close to noise. Which is exactly why the conventional wisdom on car washes, that a good independent operator beats a franchisee, is directionally right but for the wrong reason. It is not that the royalty is punitive. It is that the brand contributes very little to a business where the customer is buying a location and a monthly membership price.

One structural note that keeps getting lost. Anytime Fitness charges a flat monthly royalty, most recently in the $699 to $842 range depending on FDD vintage (22). At $750,000 in revenue that is an effective royalty near 1.35 percent, which looks like the best deal in franchising. The FDD also reserves the franchisor's right to convert that flat fee to 8 percent of gross revenue on 30 days' notice. At the same revenue that is roughly $60,000 a year against roughly $10,000 today. Any model that does not stress-test the conversion is not a model.

Part two: Item 20, the number nobody quotes

Item 19 tells you what the good units did. Item 20 tells you what happened to everybody.

The table is mandatory. Every FDD carries three years of outlet flows: openings, terminations, non-renewals, reacquisitions by the franchisor, units that ceased operations for other reasons, and transfers between franchisees. From those six numbers you can compute two rates that matter more than any AUV.

Closure rate equals terminations plus non-renewals plus reacquisitions plus other cessations, divided by average outlets. Transfer rate equals transfers divided by average outlets.

Run those on the healthiest systems in our set and the numbers are genuinely reassuring. Planet Fitness reported one termination and two other closures across roughly 2,250 average franchised outlets in fiscal 2024 (10). That is a closure rate of about 0.13 percent. Wingstop reported no terminations, no non-renewals and no reacquisitions at all while adding 277 domestic franchised units, about 15 percent growth (10). If you are looking for evidence that a strong brand reduces failure risk, that is what it looks like.

Now read the transfers. Planet Fitness recorded 270 of them, a transfer rate near 12 percent of the system in a single year. Wingstop recorded 90, about 4.5 percent.

Transfers are the most misread line in the FDD. At a growing brand with multi-unit operators and private equity money in the franchisee base, a high transfer rate is portfolio trading, not distress. At a flat or shrinking brand, the same number means people are getting out. The rate alone tells you nothing. The rate plus net unit growth tells you almost everything.

The line that deserves the most suspicion is reacquisitions. When a franchisor buys a unit back, that unit does not appear anywhere as a closure. It moves quietly from the franchised column to the company-owned column, and the closure rate you compute stays clean. 7-Eleven's Item 20 is the clearest case in our set. It carries a separate abandonment category, running roughly 75 units a year, alongside franchisor buy-backs averaging more than 170 units a year over a decade (23). Neither of those is a termination. Both of them are somebody's business ending.

An honest note on our own work here. We could not reconcile the full three-year Item 20 tables to primary registry documents for any of the deep-dive brands inside the research budget. What we confirmed are the fiscal 2024 change counts and total unit counts, several of them through aggregators that reproduce the FDD rather than through the filed document itself. The arithmetic gate we set for ourselves, that start-of-year outlets plus openings minus all exits must equal end-of-year outlets, was not cleared. We are telling you that because it is the same gate the numbers in every other article on this subject would fail, silently.

Part three: Item 19 is optional, and four of these brands skip it

The FTC Franchise Rule requires 23 disclosure items. It does not require a franchisor to say anything at all about earnings. Item 19 is voluntary. If a franchisor makes a financial performance representation it must have a reasonable basis and must disclose the material assumptions, but it can simply decline.

Four brands in our 27 decline: Servpro, Comfort Inn, TravelCenters of America's TA Express program, and, on the reading we find most defensible, The UPS Store historically (11). Sources conflict on that last one and we have left the conflict on the page rather than resolving it in our own favor.

So we graded the disclosure instead of the number. Nine tests: does Item 19 exist, does it show any cost or profit data or only gross sales, median as well as mean, what share of the system is in the sample, is the sample restricted to flatter the result, is it company-owned data standing in for franchisee data, are rent and debt and owner compensation included, are the units restricted by age or format or geography, and is there a multi-year series or a single snapshot.

The grades that came out of that:

A minus. Primrose Schools and Planet Fitness. Primrose discloses quartile gross revenue and EBITDA, which almost nobody does. Planet Fitness segments membership revenue by thirds with a median.

B plus to B. Goddard School discloses EBITDAR, useful but rent-blind. Wingstop gives median and mean net sales across a defined, fully-seasoned unit base. Fairfield and Home Instead disclose revenue with reasonable framing.

C and below. Anytime Fitness, where the flat-fee structure obscures owner economics. Jersey Mike's, where Item 19 presence has been inconsistent across vintages. Taco Bell, which does not disclose AUV in Item 19 at all, so the roughly $2.24 million figure in wide circulation comes from Yum Brands corporate reporting rather than from the disclosure document.

F. The four with no Item 19.

Here is the important part, and it is a point about grading, not about numbers. Primrose earns the highest grade in the set and still has the defect that matters most. Its headline top-quartile EBITDA of $768,966 is drawn from 89 schools that reported rent expense, out of 357 that reported anything, out of 558 open at year-end 2025 (7). Top quartile of a self-selected rent-reporting subsample is not top quartile of the system. It is the best disclosure in our sample and you still cannot underwrite from the headline.

The practical consequence for a lender is direct. When Item 19 is absent, the entire earnings-diligence burden shifts to Item 20 and to franchisee validation calls. Do not close on a no-Item-19 brand without working the Item 20 contact list.

Part four: the loan outlives the license

This is the section that does not exist anywhere else, and it is the one I would read first if I were buying.

Franchise agreements have terms. Loans have amortizations. They are not the same length, and the mismatch runs the wrong way.

A 504 structure on a hotel or a childcare center amortizes over 25 years. A 7(a) real estate loan can run 25 years. A 7(a) business acquisition loan runs 10.

Against that, the terms in our set:

• Fairfield by Marriott: 15 to 20 years, with no built-in renewal, and renewal historically conditioned on completing a mandated property improvement plan (24).

• Hampton by Hilton: roughly 22 years on a new build (25).

• Taco Bell: 20 years, with no contractual renewal right. What exists is a successor policy at the franchisor's discretion, conditioned on an asset update (13).

• Dunkin: 20 years, no automatic renewal, one 20-year renewal term if conditions are met, refurbishment mandated at year 5 and again at year 15 (12).

• McDonald's: 20 years, and renewal is now a new-term process rather than a right.

• 7-Eleven: 15 years (26).

• Planet Fitness: 12 years. Anytime Fitness: 6. Servpro: 5. The UPS Store: 10. Jersey Mike's: 10 plus one (27).

Put a 20-year Taco Bell license against a 25-year real estate amortization and you have five years of secured debt outstanding on a building whose cash flow depends on a brand you have no contractual right to keep. Put a 6-year Anytime Fitness term against a 10-year 7(a) acquisition note and the license expires four years before the loan is retired.

Three things happen at renewal, and only the first one is obvious.

The fee. Anytime Fitness discloses a $7,500 renewal fee (22). Most brands charge something.

The capital call. Renewal is when the remodel or PIP lands. In hotels this is the single largest owner expense outside debt service and it arrives twice, at renewal and again on any change of ownership. Dunkin writes the refurbishment schedule directly into the agreement (12).

The worse contract. You renew onto the then-current form of agreement, not the one you signed. The UPS Store renews on the then-current version (27). When Arcos Dorados renewed its McDonald's master agreement, the royalty stepped from 6 percent to 6.25 percent to 6.5 percent (28). Renewal is where the franchisor repricing happens, and the borrower has no leverage because the loan is already in place and the alternative is deflagging a leveraged asset.

Now the part that should be in every credit memo and is in almost none.

Hotel franchise agreements from both Hilton and Marriott are terminable at will by the franchisor in certain circumstances including franchisee bankruptcy or insolvency (14). Marriott separately holds a right to purchase or lease the hotel on a proposed transfer to a competitor (24). Read those two together from a lender's chair. On default, the brand that generates the revenue can be pulled, and the franchisor holds a right that can block a going-concern sale. The collateral you underwrote was a flagged hotel. The collateral you can actually liquidate is a building.

That gap is precisely what the SBA addendum used to close. Form 2462 forced the franchisor to subordinate control terms for the life of the SBA loan and required that the addendum terminate only when the loan is paid or SBA no longer has an interest (29). Under SOP 50 10 8 the per-loan addendum requirement is being replaced by a one-time franchisor certification tied to Directory listing (30). Directionally that is a simplification for lenders. It also means the protection is now a brand-level attestation rather than a loan-level document, and if you are the lender you should know which one you are holding.

Then the exit. Dunkin's five-mile post-term non-compete means that when the term ends without renewal, the operator cannot simply take the sign down and keep selling coffee at the same corner (12). The "deflag and go independent" option that category-verdict articles treat as an always-available fallback is contractually barred at a meaningful number of brands.

Part five: what the lender sees in 2026

The financing chapter of the standard article is where it dates fastest, because the rulebook moved twice in three years and most published guidance never caught up.

The Franchise Directory is not a continuous fixture. It was created under SOP 50 10 5(J) effective January 1, 2018. SBA announced on May 11, 2023 that it would no longer support or maintain it, and it was formally eliminated with SOP 50 10 7 effective August 1, 2023. It came back under SOP 50 10 8 for loans approved on or after June 1, 2025 (15). Brands listed as of May 2023 had until July 31, 2025 to execute the new franchisor certification or come off the list. If you are reading an article that treats the Directory as a stable, always-available eligibility tool, that article was written about a system that did not exist for roughly 22 months.

Control is no longer an affiliation test. SOP 50 10 8 determines affiliation by ownership rather than by franchisor control (31). This is the quiet structural change that matters most for franchise eligibility, because franchisor control provisions were the historic reason franchises got caught in the affiliation net. If the brand is on the Directory, no further franchise documentation is generally required absent some other management agreement pointing at outside control (32).

The combined cap is real, and it is two loans. Policy Notice 5000-879058, dated May 18, 2026 and effective July 4, 2026, decouples the programs so a qualified borrower can access up to $5 million through 7(a) and up to $5 million through 504, for $10 million combined (16). The individual 7(a) maximum is unchanged at $5 million. The prior ceiling on total SBA exposure to one borrower and its affiliates was $3.75 million, so this is a material expansion for anyone stacking an operating loan on top of a real estate project. It is not a $10 million loan and describing it that way will confuse a borrower.

Equity injection has teeth. Ten percent of total project cost on start-ups and complete changes of ownership, counting all costs required to complete the transaction regardless of funding source (33). A seller note counts toward that injection only if it sits on full standby for the life of the SBA loan, roughly ten years with no principal and no interest, and only up to half the required injection. Most sellers find that commercially impractical, which is why the two-note structure keeps appearing: one note on full standby inside the equity calculation, a second subordinated note on limited standby outside it.

Pricing and coverage, July 2026. WSJ Prime sits at 6.75 percent. Variable 7(a) pricing generally runs in the 9 to 11.5 percent range depending on size and maturity, with maximum spreads set by SOP. A February 2026 rule added the 5-year Treasury, the 10-year Treasury and SOFR as permitted base rates, capped so the resulting rate cannot exceed Prime plus the allowed spread for that loan amount (34). On 504, effective debenture rates in July 2026 have been running in the low-to-mid 6s inclusive of CDC, SBA and central servicing fees. Lenders are underwriting to a 1.25x DSCR floor and generally want 1.50x, and in this rate environment the coverage test binds long before the loan ceiling does (35).

That last point deserves emphasis because the $10 million headline invites the wrong conclusion. Nobody's deal was constrained by the old $5 million cap. Deals are constrained by debt service coverage at 9 to 11 percent money. Raising the ceiling does not raise the DSCR.

Part six: does franchising actually reduce failure risk?

Now the uncomfortable part.

The claim that carries the entire franchise sales industry is that franchises fail less often than independents. The number usually attached to it is 95 percent success, or some cousin of it.

That number is not from a study. It traces to "Franchising in the Economy," a voluntary Department of Commerce survey from the 1980s in which roughly 2,000 franchisors chose which questions to answer. Some analysts read it as about 5 percent of units closing over five years. Salespeople inverted it into a 95 percent success rate. Both Commerce and SBA have said they never issued the statistic. Commerce killed the underlying series in 1987 (36).

What does the peer-reviewed literature say?

Timothy Bates, working with US Census Bureau microdata on roughly 20,000 young small businesses started between 1984 and 1987 and tracked into 1991, found that franchise startups showed both higher discontinuance and lower mean profitability than independent startups. Roughly 62 percent of franchise firms survived against roughly 68 percent of independents, and average profit for the franchise group was negative. Controlling for owner and firm characteristics in a logistic regression, being a franchise was negatively related to survival (17).

Scott Shane, analyzing SBA loan performance, found franchise default rates higher than those of comparable independent businesses (37).

Lafontaine and Shaw documented very high franchisor exit rates, and found that royalty rate, advertising fee, franchise fee and capital required have little power to explain survival. The main predictor of a franchisor lasting is years in business before it started franchising (38).

Holmberg and Morgan, using longitudinal franchisee failure data, found that prior estimates ranged from the old Commerce 4 to 5 percent all the way to 25 to 35 percent depending entirely on who was defining failure (39).

I want to be careful about how much weight this carries. The Bates data is a 1980s cohort. Franchising has professionalized enormously since, brand concentration has increased, and the multi-unit operator has largely replaced the single-store buyer in the strongest systems. Somebody should redo that study on modern Census microdata. Nobody has, at least not that we could find through mid-2026.

But the direction of the evidence is clear, and it is the opposite of the marketing. There is no credible peer-reviewed finding that franchising, by itself, improves survival once you control for who is buying and how well capitalized they are.

The modern loan data says something more useful than either camp: it depends entirely on the category.

On a resolved-loan basis, franchised limited-service restaurants default less than independents in the same sector, roughly 17.3 percent against 18.9 percent (40). In ambulatory health care, which contains home care, franchises again come out ahead, 7.5 against 8.2. In administrative support services, franchises do better by a wide margin, 13.7 against 17.3.

And then repair and maintenance, the sector containing car washes, where franchised businesses default at 20 percent against 14.4 percent for all businesses in the sector (40). Independents win, and not narrowly.

That single reversal is the whole argument. Franchising is not a risk-reduction technology. It is a trade, and whether it pays depends on whether the brand does real work in that specific category.

One methodological warning, because these numbers get quoted carelessly. Charge-off rates move by roughly a factor of ten depending on the denominator. On a dollar-weighted basis across 1995 to 2024, hotels run about 1.8 percent and limited-service restaurants about 6.6 percent (41). On a resolved-loan basis in 2026, limited-service restaurants run 18.9 percent (40). Both are defensible. They are answering different questions. Never put them in the same chart, and always ask which denominator a quoted default rate used.

Two more findings worth carrying into any credit decision. The seasoned-cohort work on 2008 to 2012 SBA vintages found roughly 7.5 percent charged off on a count basis, with a sharp term effect: loans of 10 years or less defaulted near 20 percent, against about 4 percent for loans with terms between 10 and 25 years (42). Shorter amortization is itself a risk factor, which loops back to the term mismatch problem above. And GAO found that among franchise loans it reviewed, 55 of 88 loans (63 percent) at four high-volume lenders defaulted, against 19 of 82 (23 percent) at the other 50 lenders (43). Who wrote the paper mattered more than which brand was on the sign.

Part seven: two structures that are not franchises

Two of the most cited names in any franchise article do not build the thing everybody assumes they are buying.

Chick-fil-A runs an Operator Program, not a franchise in the equity sense. The initial fee is $10,000, of which $5,000 is a working capital deposit. The operator pays a 15 percent Base Operating Service Fee on gross sales plus a 50 percent split of pre-tax profit, rents equipment at $750 to $5,000 a month, and contributes up to 3.25 percent to advertising. Chick-fil-A owns or controls the restaurant and the equipment (8). Operators typically net something in the range of 5 to 7 percent of gross, which on a $4 million unit is real money, roughly $200,000 to $240,000. But the operator cannot sell the business, cannot bequeath it, and does not hold an asset. Selection runs at roughly 80 to 100 new operators a year out of tens of thousands of applicants.

It is one of the best income-to-cash-invested opportunities in American food service. It is also not comparable to any other line in a franchise comparison table, and putting it in the same table as Taco Bell is how readers get confused about what they are buying.

7-Eleven splits roughly half of store gross profit with the franchisor, which owns or leases the store and finances inventory (9). The traditional franchise fee can be nominal or zero. That is a managed license with an income stream, not an equity build.

For a lender the distinction is not academic. In both structures there is no sellable business asset securing the loan and no resale equity to underwrite an exit against. Bucket them separately or your collateral analysis is fiction.

Part eight: category by category, with the documents behind each call

Here is where we agree and disagree with the conventional table.

Hotels. Franchise, with the worst term risk in the set. The flag earns its keep on reservation contribution, loyalty, corporate travel and lender familiarity, and the fee stack of roughly 9.5 to 10 percent is defensible against that. But hotels carry the deepest term mismatch (15 to 22 year agreements against 25-year amortization), the largest renewal capital calls, and the strongest franchisor exit-blocking rights, including at-will termination on insolvency and, at Marriott, a purchase or lease right on transfer to a competitor (14, 24). Underwrite the PIP twice, at renewal and at sale, and read the termination clause before the pro forma.

QSR. Franchise, if and only if the AUV-to-investment ratio works. The document evidence supports the conventional call. Wingstop's median of $2,001,753 against a total investment range starting near $300,000 is the best ratio in our set (6). The risk is not the brand, it is the operating math: prime cost near 60 to 65 percent, occupancy near 5.2 percent of sales, third-party delivery running an all-in 30 to 40 percent of the orders that come through it, and restaurant-level EBITDA landing in the mid to high teens (44). California operators carry an additional structural cost from the $20 fast food wage under AB 1228, effective April 1, 2024, which independent research estimates reduced California fast food employment by about 3.6 percent, roughly 18,000 jobs (45). Model the delivery mix explicitly. It is the line that quietly turns a good store into a break-even one.

Gas and convenience. The brand is optional, the foodservice is not. NACS 2025 data puts total industry sales at $817.5 billion, with in-store sales of $341.2 billion rising for the 23rd consecutive year. Fuel is 65 percent of sales dollars and 38.8 percent of gross profit. Foodservice is 28.5 percent of in-store sales and 38.9 percent of in-store gross profit dollars (46). That is the entire argument in two numbers. The winning operators are foodservice operators who happen to sell fuel. A c-store franchise can help a first-timer, but 7-Eleven's roughly 50 percent gross profit split is a very expensive way to buy systems if you already know how to run a kitchen.

Truck stops. There is barely a franchise decision to make. Love's, Pilot Flying J and Buc-ee's do not franchise. TravelCenters of America does, and even there the count was roughly 14 franchised against about 167 company-operated, with a TA Express structure carrying a $125,000 franchise fee, a graduated royalty, and no Item 19 (47). Treat travel centers as an owner-operator or corporate asset class. The value is land, diesel volume, parking and amenities, not a logo.

Car wash. Independent, and the loan data agrees. Tommy's Express is a competent system with an unusually honest cohort-based Item 19 (20). But the fee load is small precisely because the brand contributes little, and the sector-level SBA data shows franchised repair and maintenance businesses defaulting at 20 percent against 14.4 percent for all businesses in the sector (40). Meanwhile the operating model is extraordinary when it works: Mister Car Wash reported fiscal 2025 revenue of $1,051.7 million with unlimited membership at 79 percent of wash sales and adjusted EBITDA of $345.4 million (21). Single-site EBITDA of 45 to 63 percent is achievable. Underwrite membership churn, not car counts. Below roughly 8 percent monthly churn the revenue behaves like a subscription and the multiple follows.

Childcare. Franchise, if you can fund it. Primrose has the best disclosure in our set and the economics are strong where demographics support the tuition, but the capital requirement runs to $6 million or more with real estate, the fee load is 11 percent, and the enrollment ramp is long. Read the Item 19 sample construction before you believe the top-quartile EBITDA (7).

Home care. Franchise is a reasonable trade. Low capital, roughly 7 percent fee load, and franchised operators in the parent sector default less than independents (7.5 against 8.2) (40). The business is recruiting and scheduling, not brand. Watch the payer environment: the 2025 reconciliation law is estimated to reduce federal Medicaid spending by $911 billion over the decade, and median Medicaid payment to personal care agencies sits near $26 an hour against a bill rate for private pay closer to $34 to $35 (48).

Restoration. The one place I would push back hardest on the conventional table. Everybody says franchise, because insurance relationships and national accounts matter. That is true. But Servpro combines the heaviest effective fee load in our set, 22 to 31 percent of gross at startup volumes once mandatory equipment sourcing is counted, with no Item 19 at all (19). No earnings disclosure plus a mandatory supplier structure is the single worst pairing we found. Franchise here if you want, but do the validation calls first and model the equipment premium as a separate line.

Fitness. Be stricter than the table suggests. Anytime Fitness's convertible royalty is an unpriced option written against the franchisee (22). Planet Fitness has excellent unit stability and a good Item 19, but the fee structure includes a local advertising requirement of the greater of $60,000 or 7 percent, which is a hard floor that does not flex when a location underperforms. And the FTC's $17 million Xponential judgment in March 2026, followed by a $3,971,250 New York Attorney General settlement in June, both centered on how long studios actually took to open versus what the FDD said, is a live warning about buildout timeline representations in this category specifically (18, 49).

Pack and ship. Stable, limited upside, and one exogenous risk. The UPS Store is 5 percent royalty plus 3.5 percent marketing with essentially no company-owned units (27). The forward risk is at the parent: UPS has stated it will cut Amazon volume by more than half by the second half of 2026, and Amazon represented 11.8 percent of UPS's 2024 revenue (50). Less prepaid label and returns volume flows straight through store traffic.

Weddings, glamping, self-storage. Not a franchise question. Self-storage is a REIT third-party management market, not a franchise market. Extra Space managed 2,263 third-party and joint-venture stores at year-end 2025 and CubeSmart 902 (51). Those platforms scaled past any franchise system precisely because storage branding carries almost no consumer pull. In weddings and glamping the franchisors that exist are small and not registered in every state. The absence of a franchise market in a category is itself information: it means the customer is buying the site, and no logo is going to change that.

The decision rule, rebuilt

The conventional rule says franchise when the brand increases sales and stay independent when it does not. That is true and useless, because it is unfalsifiable at the moment of decision.

Here is a testable version.

Franchise when all four hold:

1. The fee stack, fully loaded, is less than a third of the category's typical operating margin, or the brand can demonstrate an AUV premium over independents large enough to cover it. Compute the Fee Load Index. Do not accept the royalty as the fee.

2. Item 19 discloses cost or profit data, states its sample size against total system units, and the sample is not restricted to the top quartile of a self-selected subgroup.

3. Item 20 shows a closure rate under 5 percent with net unit growth, and any high transfer rate is accompanied by growth rather than contraction. Watch reacquisitions separately.

4. The franchise term is at least as long as the loan amortization, or the renewal is a right rather than a discretion, or you have priced the renewal fee plus the mandated remodel into year-one underwriting.

Stay independent when any of these hold:

1. The value driver is the parcel. Traffic count, ingress and egress, corner position, highway visibility, truck parking. No brand improves a bad site and a great site does not need one.

2. The category shows franchised businesses defaulting at or above independents in the SBA data.

3. The brand cannot or will not disclose earnings, and the validation calls do not compensate.

4. Your own operating expertise in the category exceeds what the system provides, and you can access the same equipment, technology and vendor pricing independently.

And regardless of which side you land on: borrow the discipline. Standardized offerings, a real CRM, documented operating procedures, tiered pricing, review management, contract templates and monthly financial controls. Most independent operators who lose to franchises do not lose on brand. They lose on systems, which are free to copy.

Red flags, with the document reference attached

The standard red-flag list is a list of adjectives. Here is one you can actually check.

What to look for	Where it lives	Why it matters
No Item 19 at all	Item 19	Legal and common. Shifts all earnings diligence to validation calls
Item 19 shows only gross sales	Item 19	You still do not know owner profit
Sample is a subgroup of a subgroup	Item 19 footnotes	Top quartile of self-selected reporters is not top quartile of the system
Closed units excluded from the Item 19 base	Item 19 footnotes	Survivorship. Anytime Fitness excluded 29 permanently closed centers
Rising reacquisitions	Item 20	Buy-backs do not appear as closures
Transfer rate above 10 percent with flat or negative growth	Item 20	Franchisee distress, not portfolio trading
Fee stack far above the headline royalty	Item 6	Tech, brand fund, local minimum, reservation, loyalty
Franchisor-retained supplier rebates	Item 8 and Item 11	Your COGS is a revenue line for someone else
Royalty convertible at franchisor discretion	Item 6	Anytime Fitness reserves conversion from flat fee to 8 percent
Term shorter than the loan	Item 17	The loan outlives the license
Renewal at franchisor discretion	Item 17	There may be no renewal right at all
Renewal onto then-current agreement	Item 17	Repricing with no leverage
Right of first refusal on transfer	Item 17	Can block a going-concern sale in liquidation
At-will termination on insolvency	Franchise agreement	The collateral stops being a flagged asset at the worst moment
Post-term non-compete	Item 17	Kills the deflag-and-continue exit
No territorial exclusivity	Item 12	Taco Bell grants none. Dunkin reserves competing licenses
Alternative channel reservations	Item 12	Delivery, ecommerce and non-traditional venues carved out

Outlook to 2030

Four things are moving.

Regulatory attention is rising and it is bipartisan in effect if not in framing. The FTC's July 2024 actions on undisclosed fees and on contract provisions restricting franchisee communications, followed by the Xponential judgment in March 2026 and the New York settlement in June, establish that FDD representations about time-to-open and about fees are now enforcement targets (18, 49, 52). Franchisors will get more conservative in Item 19 and Item 11 language, which is good for buyers and bad for anyone hoping the disclosures get more informative.

Joint employer risk fell. The NLRB's 2023 broad standard was vacated in March 2024, and a final rule effective February 27, 2026 returned to the 2020 substantial-direct-control test (53). That is a genuine reduction in franchisor operating risk and one of the few structural changes in this period that favors the model.

The capital stack got looser and the coverage test did not. The July 4, 2026 decoupling to $10 million combined is real, but at 9 to 11 percent 7(a) money the binding constraint is DSCR, not the ceiling (16, 35). Expect more 504-heavy structures on real-estate-anchored assets and more scrutiny of the 7(a) piece.

The multi-unit operator keeps winning. Nothing in the Item 20 data suggests single-unit buyers are having a good decade. The systems with near-zero closures are the ones whose franchisee bases have consolidated into professional operators with portfolio-level overhead. If you are buying one unit of anything, you are competing against people who buy ten.

Frequently asked questions

Is it true that 95 percent of franchises succeed? No. That figure traces to a voluntary 1980s Department of Commerce survey that was discontinued in 1987, and both Commerce and SBA have said they never issued it as a success rate (36). The peer-reviewed microdata points the other way: roughly 62 percent franchise survival against roughly 68 percent for independents in the Bates Census study (17).

Are franchises safer than independent businesses? It depends entirely on the category. On resolved-loan SBA data, franchised limited-service restaurants and franchised home care businesses default less than independents in the same sector, while franchised repair and maintenance businesses, the sector containing car washes, default at 20 percent against 14.4 percent for all businesses (40). There is no general answer.

What is a normal franchise royalty? Around 6 percent of gross sales is the middle of the market. But the royalty is not the fee. Total contractual load, including brand fund, local marketing minimums, technology and reservation charges, commonly runs 9 to 11.5 percent in brand-driven categories (1, 2, 3, 4).

Can I still get an SBA loan for a franchise? Yes, and it got simpler. The Franchise Directory was reinstated effective June 1, 2025 under SOP 50 10 8, and control is no longer an affiliation test, so brand listing generally resolves the franchise question (15, 31). Since July 4, 2026 a qualified borrower can combine up to $5 million of 7(a) with up to $5 million of 504 (16).

What is Item 19 and why do some brands not have one? Item 19 is the financial performance representation. The FTC Franchise Rule requires 23 disclosure items but does not require an earnings claim, so a franchisor may lawfully say nothing. Four brands in our 27 do exactly that.

What happens to my franchise if I default on the loan? Frequently the franchise terminates. Hilton and Marriott agreements are terminable at will by the franchisor in circumstances including franchisee bankruptcy or insolvency, and Marriott holds a purchase or lease right on a proposed transfer to a competitor (14, 24). This is why the SBA addendum, now a one-time franchisor certification under SOP 50 10 8, exists.

Can I drop the brand and keep operating? Often not. Dunkin imposes a five-mile post-term non-compete, and Anytime Fitness's own FDD warns that the agreement may bar operating a similar business after the franchise ends (12, 22). Check Item 17 before you assume deflagging is an available exit.

Methodology, and what we could not do

We pulled Franchise Disclosure Documents for 27 brands across 11 categories and extracted Items 5, 6, 7, 11, 12, 17, 19 and 20. Every brand-level figure in this article carries an FDD issuance date or fiscal year where we have one. SBA rules were verified against the SOP, Policy Notices and Federal Register text. Operating benchmarks come from NACS State of the Industry, BLS, SEC filings and industry sources named in the source list. Survival evidence comes from the peer-reviewed literature and from BLS Business Employment Dynamics.

Four limitations, stated plainly.

The Item 20 reconciliation gate failed. We set ourselves an arithmetic test, that start-of-year outlets plus openings minus all exits must equal end-of-year outlets for each brand-year across three years. We could not retrieve the full three-year tables to primary registry documents inside budget for any deep-dive brand. What is reported here are fiscal 2024 change counts and total unit counts, several through aggregators that reproduce the FDD rather than through the filed document.

Roughly a third of our target brands are category context, not verified line items. Kiddie Academy, Lightbridge, Right at Home, BrightStar, Visiting Angels, Comfort Keepers, PuroClean, Paul Davis, Rainbow Restoration, Orangetheory, Crunch, PostNet, Dunkin, Domino's, IHG, WoodSpring, Microtel, Take 5, Zips, Circle K and ampm were not individually verified.

Some sources conflict and we left the conflicts visible. The UPS Store's Item 19 existence, Anytime Fitness's flat royalty across vintages ($699, $799 and $842 all appear), Servpro's franchise fee and investment range, and the exact 7(a) maximum spread table are all cases where published sources disagree. We have not resolved them in our own favor.

BLS does not publish establishment survival at 6-digit NAICS. Survival data for hotels, limited-service restaurants, car washes, childcare and the rest exists only at sector level. Anyone quoting a 5-year survival rate for a specific 6-digit industry from BED is quoting something that does not exist.

We are publishing the gaps because the gaps are the point. An article that presents 27 brands of tidy numbers with no failed gates is an article that did not set any.

Sources

(1) Loan Analytics Invest database and analysis, July 2026. Fee Load Index construction and category margin benchmarks. (2) Hampton by Hilton franchise terms, FY2024 SEC CMBS and franchisee filings. (3) Comfort Inn FDD, April 2025, via franchise disclosure summaries. (4) Jersey Mike's FDD 2025/2026, franchise disclosure summaries. (5) McDonald's USA royalty change announced September 22, 2023, effective January 1, 2024; McDonald's SEC Form 10-Q, FY2025. (6) Wingstop FDD issued March 2025, Item 19. (7) Primrose Schools FDD issued April 24, 2026, Item 19. (8) Chick-fil-A Operator Program Franchise Disclosure Document, Items 5 and 6. (9) 7-Eleven FDD 2025 and franchise structure summaries. (10) FDD Item 20 unit tables, fiscal 2024, Planet Fitness and Wingstop, via FDD-reproducing aggregators. (11) MMCG Item 19 disclosure grading, 27-brand sample, July 2026. (12) Dunkin franchise agreement, SEC-filed exhibit; DDIFO commentary on section 10.2. (13) Taco Bell franchise agreement, SEC-filed exhibit. (14) Chatham Lodging Trust SEC filings describing Hilton and Marriott franchise agreement termination provisions. (15) SBA Information Notice 5000-866746, April 22, 2025; SBA Franchise Directory document page, last version May 11, 2023; SOP 50 10 7 effective August 1, 2023. (16) SBA Policy Notice 5000-879058, dated May 18, 2026, effective July 4, 2026; SBA announcement, May 18, 2026; NAGGL summary. (17) Bates, T., "A Comparison of Franchise and Independent Small Business Survival Rates," Small Business Economics 7(5), October 1995; Bates, "Survival patterns among newcomers to franchising," Journal of Business Venturing 13(2), March 1998. US Census Characteristics of Business Owners data. (18) FTC press release and order, Xponential Fitness, March 18, 2026. (19) Servpro FDD 2024/2025; Lopes Law LLC fee-burden model. (20) Tommy's Express FDD 2025. (21) Mister Car Wash (Nasdaq: MCW) FY2025 results release, February 18, 2026; Car Wash Advisory site-level benchmarks. (22) Anytime Fitness FDD issued March 31, 2025. (23) 7-Eleven Item 20 structure; NCASEF and franchise data summaries. (24) Fairfield by Marriott FDD dated March 31, 2024; UBS Commercial Mortgage Trust 424B2 describing PIP-conditioned renewal. (25) Hampton by Hilton franchise term, FranchiseDirect summary. (26) 7-Eleven franchise term, FranchiseDirect summary. (27) The UPS Store FDD issued May 7, 2025. (28) Arcos Dorados master franchise agreement renewal royalty schedule, Nation's Restaurant News. (29) SBA Form 2462, Addendum to Franchise Agreement. (30) SOP 50 10 8 franchisor certification requirement; Starfield & Smith, June 2025. (31) SOP 50 10 8 affiliation provisions; Taft/Lexology analysis, 2025. (32) McDonald Hopkins, SBA Franchise Directory reinstatement guidance, 2025. (33) SOP 50 10 8 equity injection provisions; Starfield & Smith. (34) Federal Register, 7(a) Alternative Base Rate Options, February 10, 2026. (35) 2026 SBA lender underwriting standards, multiple lender sources; WSJ Prime 6.75 percent, July 2026. (36) "Franchising in the Economy," US Department of Commerce, discontinued 1987; Lafontaine and Shaw (1998) on the data series; NBC News and FranBest reporting on the statistic's origin. (37) Shane, S., research on SBA loan default rates for franchised versus independent businesses. (38) Lafontaine, F. and Shaw, K., "Franchising growth and franchisor entry and exit in the U.S. market: Myth and reality," Journal of Business Venturing 13(2), 1998. (39) Holmberg, S. and Morgan, K., "Franchise turnover and failure: New research and perspectives," Journal of Business Venturing 18(3), 2003. (40) PeerSense resolved-loan SBA default analysis, July 2026, sector level. (41) SBA 7(a) dollar-weighted charge-off analysis 1995-2024, SBALenders.com, March 2025. Reconfirmation flagged in methodology. (42) SBA7a.loans / CapTec seasoned-cohort analysis of 2008-2012 approvals, published August 24, 2023. (43) US Government Accountability Office, GAO-13-759, 2013. (44) QSR operating benchmarks: Toast and Restroworks 2025 industry data; BLS Occupational Employment and Wage Statistics, May 2024; delivery platform commission schedules, 2026. (45) Clemens, Edwards and Meer, NBER Working Paper No. 34033, July 2025, on California AB 1228. (46) NACS State of the Industry 2025 data, released April 15, 2026. (47) TravelCenters of America TA Express FDD 2024; company-operated versus franchised unit counts, 2025. (48) CMS CY2025 home health payment update; KFF analysis of 2025 reconciliation law Medicaid provisions; CareScout 2026 cost of care data. (49) New York Attorney General settlement with Xponential Fitness, June 9, 2026. (50) UPS Q4 2024 earnings call, January 31, 2025, on Amazon volume reduction. (51) Extra Space Storage and CubeSmart third-party management platform disclosures, year-end 2025. (52) FTC Policy Statement and Staff Guidance on franchise contract provisions and undisclosed fees, July 12, 2024; FTC Issue Spotlight, "Risks to Small Business Success in Franchising," July 12, 2024. (53) Chamber of Commerce v. NLRB, E.D. Tex., March 8, 2024; NLRB final rule effective February 27, 2026.

This article is general commentary for owners, operators and lenders. It is not investment, legal or tax advice, and it is not a substitute for reading the current Franchise Disclosure Document for any brand you are considering. Every figure here should be re-verified against the current FDD pulled from a state franchise registry before any capital commitment. MMCG Invest prepares independent feasibility studies for SBA 7(a) and 504 financing, and franchise underwriting is one of the questions those studies are built to answer.