Ask ten founders what their company is worth and nine will quote a multiple they heard at a conference. Ask ten buyers and you will get a different answer: it depends on what kind of asset you actually are. After 25 years and more than 200 closed transactions in payments, fintech, AI, and vertical SaaS, the most expensive mistake we see is a founder anchored to a generic multiple that was never calibrated to their asset class, their size, or their risk profile.
This piece publishes the actual bands we use, the same calibration that powers our valuation calculator. Not because the numbers alone will value your company, but because seeing how wide the bands run, and what moves a company inside them, changes how you prepare for an exit.
Merchant portfolios: the widest band in the market
A merchant portfolio or residual stream is valued as a multiple of net monthly residual, and in 2026 that band runs from the low teens to 65x. That is not a typo, and it is why a one-size multiple is useless here. A small static book with elevated attrition trades in the teens. A 10,000-account, low-attrition portfolio locked into a software integration can clear 60x monthly. Three factors create most of that spread.
Attrition is destiny. A book losing 5% of revenue a year and a book losing 18% are different assets, even at identical residual. Buyers price portfolios like a discounted cash flow: every point of churn compounds against the purchase price. Scale earns a structural premium because 10,000 aggregated accounts save an acquirer years of signing merchants one at a time. And integration is the new dividing line: when payments are embedded in software the merchant depends on daily, attrition falls, stickiness rises, and the multiple follows. High-risk portfolios are the exception that proves the band: they trade as revenue-stream purchases, typically 24x to 36x monthly, regardless of how good everything else looks.
Whole ISOs: legacy versus integrated is the whole story
An operating ISO is valued on adjusted EBITDA, and the 2026 market has split it into two businesses that happen to share an industry. A legacy ISO, built on phone calls and feet on the street with no proprietary technology, trades at roughly 6x to 8x adjusted EBITDA depending on scale. An ISO with meaningful integrated payments or software revenue trades at 10x to 18x. We closed a transaction at 18x for an ISO integrated into government payments. Same industry, more than double the multiple, because the buyer was purchasing durable, software-locked revenue instead of a sales force.
Size matters on a curve, not a line: the market pays up through roughly $20 million in EBITDA and flattens above it. Attrition, processor relationships, vertical concentration, and documentation quality each move the number by turns, not decimals.
Fintech: recurring revenue is the multiple
Fintech companies trade on revenue multiples that run from roughly 1x to 8x, and the single biggest separator is revenue quality. A predominantly recurring revenue base earns a premium; a transactional, one-time revenue mix can cut the multiple by more than half. Net revenue retention tells buyers whether the base is compounding or quietly shrinking, growth sets the slope, and size works asymptotically: small companies degrade fast, and the curve flattens past roughly $30 million in revenue. Owned licensing or a regulatory moat adds real turns, and monetized embedded payments, a take rate on the money moving through your platform, is one of the most reliable premium drivers we see.
Vertical SaaS: depth beats breadth at exit
Vertical SaaS companies trade on ARR multiples in a similar band to fintech, but the drivers rank differently. Growth and net revenue retention lead. The Rule of 40, growth rate plus profit margin, is the efficiency bar buyers use to separate durable compounders from companies buying growth with losses, and in 2026 it keeps paying above 40 rather than plateauing. Gross margin separates real software economics from services dressed as software: the spread between 80%-plus margins and sub-65% margins is worth more than a full turn of ARR. Vertical depth itself is why these companies outsell horizontal tools at exit: when your software runs a niche industry's operations, the buyer is acquiring a market position, not a feature.
AI: the moat is the multiple
AI companies trade in the widest premium band we track, roughly 2x to 16x revenue, and the spread has almost nothing to do with revenue. The first question every sophisticated buyer asks is whether there is a genuine proprietary asset: training data nobody else has, a model edge that survives scrutiny, a workflow position that compounds. The second is defensibility: what happens to this company when the next foundation model release ships? Real moat plus high defensibility earns software-premium multiples. A thin wrapper on someone else's model gets valued like a feature, regardless of growth. The technical team carries standalone value too; an elite, retainable AI team is worth multiple turns over a thin one. And net revenue retention matters more in AI than anywhere else, because it proves customers stay after the novelty wears off.
Why generic calculators get this wrong
Most online valuation tools apply a flat multiple to revenue and call it a day. The bands above show why that fails: the difference between the bottom and top of each band is not noise, it is the majority of the value. A real model has to anchor to size first, then move the multiple in turns for the factors buyers actually re-price in diligence: attrition, integration, retention, moat, margin, documentation. That is how we built the 733Park valuation calculator: calibrated on our own closed transactions, vertical by vertical, with every factor's impact shown in the report. It takes about a minute, it is confidential, and the full factor-by-factor breakdown arrives by email.
Frequently asked questions
What is an ISO worth in 2026?
A legacy ISO with no proprietary technology typically trades at 6x to 8x adjusted EBITDA. An ISO with meaningful integrated payments or software revenue trades at 10x to 18x adjusted EBITDA, with scale, attrition, and documentation quality moving the number inside those bands.
What multiple do merchant portfolios sell for?
Merchant portfolios trade on net monthly residual, from the low teens for small, high-attrition books to 55x-65x for large, low-attrition, software-integrated portfolios. High-risk books typically cap near 36x monthly.
What is a fintech company worth in 2026?
Most fintech companies trade between 1x and 8x revenue. Recurring revenue mix, net revenue retention, growth, and company size drive where a business lands, and monetized embedded payments add a measurable premium.
What multiple do vertical SaaS companies sell for?
Vertical SaaS companies trade on ARR multiples driven by growth, net revenue retention, Rule of 40 performance, and gross margin. Strong vertical SaaS businesses earn meaningfully higher multiples than horizontal tools of the same size.
How are AI companies valued?
AI companies trade in a premium band of roughly 2x to 16x revenue. Proprietary data or model moat, defensibility against foundation-model releases, team quality, and net revenue retention determine the multiple far more than revenue alone.
Lane Gordon is the founder of 733Park, a boutique M&A advisory firm for payments, fintech, AI, and vertical SaaS companies. Find out what your company is worth, or reach out for a confidential conversation. 733Park.com
Related insights
How merchant portfolios and residual streams are valued in a sale, what raises or lowers the multiple, and how ISOs benchmark before going to market.
How acquirers value payments, fintech, and SaaS companies, what raises or lowers your multiple, and how to benchmark your business before a sale.