# Buying a SaaS Business

At searcher size, SaaS sells on profit rather than on ARR headlines. Why churn is the thesis, and why the founder is usually the entire engineering team.

Source: https://searchspheresource.com/guides/buying-a-saas-business
Last checked: 2026-10-03

## Forget the ARR Headlines

Venture-scale SaaS trades on [ARR](https://searchspheresource.com/glossary/arr) multiples; the SaaS a [searcher](https://searchspheresource.com/glossary/searcher) buys does not. BizBuySell's sold software and app listings put half the market between 2.38x and 4x [SDE](https://searchspheresource.com/glossary/sde) across 2021 to 2025, with a 3.12x median. The publisher reads 4x as the top of that middle range rather than the floor of a premium tier: a business over $1M of sales may approach it, while one under $200k trades just above 2x. ARR multiples only enter the conversation for growing, team-run companies. When a listing quotes an ARR multiple on a founder-run product, translate it back to SDE before comparing anything.

## Churn Is the Thesis

[Recurring revenue](https://searchspheresource.com/glossary/mrr) is only as good as its retention, and analysis of SaaS pricing consistently shows [churn](https://searchspheresource.com/glossary/churn) differences translating into multiple-turn [valuation](https://searchspheresource.com/glossary/valuation) gaps. Pull cohort retention, not the blended rate: logo churn and revenue churn separately, expansion revenue, and how the oldest cohorts behave. A product whose customers leave at 8% a year is a fundamentally different asset from one at 3%, whatever the identical MRR charts suggest.

## The Founder Is Usually the Engineering Team

Most searcher-sized SaaS was built and is maintained by the selling founder personally, which puts the company's entire technical capacity on the departing party. Underwrite the handoff explicitly: documentation quality, code maintainability (an independent technical review is worth its fee), deployment and incident processes, and what happens on the first serious bug after the transition period ends. Your plan needs a named answer to 'who fixes production at 2 a.m.,' and that answer costs money that belongs in the model.

## Revenue Quality Beyond the Churn Rate

Map the acquisition channel too, since a product fed by the founder's audience or content often churns its growth engine at close. Two assignments decide whether an asset deal works at all: a change of registrant on the domain starts a 60-day lock on changing registrars unless the seller opts out first, and the payment account moves only with notice and a written assumption. A stock sale avoids both. Look inside the MRR:

## What to Verify in Diligence

Beyond cohorts and code: infrastructure and hosting costs at current pricing, third-party API dependencies and their terms, since a critical API repricing can erase the margin, and license compliance in the codebase. Then data-protection posture and breach history, support ticket volume and who answers it, intellectual-property assignment from every contractor who ever touched the code, and the payment stack's transferability. For anything AI-adjacent, unit economics per customer at real usage, since inference costs scale with success.

## Financeability Notes

Small SaaS deals often close with more structure and less bank debt than main-street businesses: earnings histories are short, collateral is nonexistent, and many lenders pass, though SBA loans do close on established, documented products. [Seller notes](https://searchspheresource.com/glossary/seller-note), marketplace escrow and, where no SBA loan is involved, [earnouts](https://searchspheresource.com/glossary/earnout) tied to retention are common, since the SBA prohibits seller earnouts. Model debt service, where debt exists, on retention-adjusted revenue with a real budget line for the development capacity you must replace.

## Terms in This Industry

- **Expansion revenue**: Extra money from existing customers, through more seats, more usage or a higher tier. It is what makes net retention exceed a hundred percent, and it is the difference between a business that grows without selling and one that has to replace every departure before it grows at all. It also tells you where the product's pricing actually bites. Ask for revenue from existing accounts split into upgrades and usage, because a business whose expansion is all usage grows with its customers' fortunes instead of with its own product.
- **Logo churn versus revenue churn**: Two measures of the same leak: one counts the customers lost, the other counts the dollars. The gap between the two is the story. A business can shed a great many small accounts while revenue grows from a handful of larger ones, which reads as healthy until you notice the base is thinning. Ask for both, by cohort, and ask which direction the average account size is moving. A seller quoting only the flattering measure has told you which one to check first.
- **Gross revenue retention**: What last year's customers still pay this year before any upgrades, capped at one hundred percent. Net retention can look healthy while the base is leaking, because a handful of expanding accounts hides everyone shrinking or leaving. Gross retention cannot, since it only counts losses. Ask for both and treat the gap between them as a measure of how concentrated the growth is. Under a bank loan the distinction stops being academic: debt service is paid out of the revenue that stays, not the revenue that might expand.
- **Seat-based versus usage pricing**: Whether customers pay per person or per unit of what they consume, which sets how revenue moves. Seat-based revenue is predictable and shrinks in a layoff; usage revenue tracks the customer's own volume and can fall without anyone cancelling. The distinction decides how much of a forecast a lender should believe and how quickly a bad quarter reaches you. Read a full year of monthly revenue by account rather than the annual total, and ask which contracts carry minimums, because a minimum is the difference between a floor and a hope.
- **Successor assignment**: The clause that lets a customer contract follow the business, and the conditions that void it. The usual worry is backwards here. Published software terms generally do let a customer assign the agreement to a successor by merger, acquisition or a sale of substantially all assets without asking, and the vendor reserves a free hand to assign its own side. The leak is in the conditions. One set requires prompt written notice, a written assumption of every obligation by the buyer, and compliance with the vendor's own documentation requirements, and says any attempt outside that is null and void. So the contracts do not travel by themselves. Build the notice and assumption list from the actual agreements during diligence, and treat an undefined word like prompt as a deadline somebody else gets to define.
- **Source code escrow**: A third party holding the code for customers, on release terms bankruptcy law narrows. Enterprise contracts often require one, and a small software company that signed the clause and never set the arrangement up has a breach sitting in its largest accounts. The clause almost always names a bankruptcy filing as the release event, and that is the one trigger that does not work: the bankruptcy code voids a contract term conditioned on the filing itself, on insolvency, or on a trustee taking over. What does work is the customer's own election once the trustee rejects the contract, which preserves its rights under the license and any supplementary agreement, including a right to obtain the code from another entity. That last phrase is what reaches the escrow agent. Two limits survive. The release turns on rejection and an election, not on the agent's reading of a clause, and the definition of intellectual property it runs on leaves trademarks out, so a customer gets the code and no right to keep selling under the name. Ask which contracts require it, whether the deposit is current, and when it was last verified.
- **Voided renewal**: A renewal Florida voids when the seller's notice fails, saved only by a documented error and a refund. Gross revenue retention counts what last year's customers still pay this year, and it cannot tell which of those dollars arrived through a renewal a statute can unwind. Florida's automatic renewal section says a violation renders the automatic renewal provision void and unenforceable, and the only escape it offers is a refund of the unearned portion as of the date the seller is notified of the error. The multiple is paid on renewed revenue, so read the actual signup flow and the renewal email instead of the terms page, and count the contracts on annual terms. The scope line comes first, though: that statute defines a consumer as an individual and excludes an individual acting for a business, so a pure seat book sits outside it and a freelancer book does not. Most books are mixed.
- **Implementation revenue**: One-time setup and onboarding fees, which are not recurring even when a summary counts them as ARR. Setup fees arrive with a new customer and never repeat, so counting them inside annual recurring revenue inflates both the growth rate and the multiple applied to it. The distortion is largest in the year a product wins several large customers, which is usually the year the business is offered for sale. Ask for revenue split into subscription and one-time, by month, and rebuild the recurring figure yourself. Two lines of a spreadsheet can move the price by a whole turn.

## What the Data Says

- BizBuySell's sold software and app listings put half the market between 2.38x and 4x SDE across 2021 to 2025, with a 3.12x median, on a $625,000 median sale price and $483,261 of median revenue. The publisher reads 4x as the top of that middle range rather than the floor of a premium tier: a business over $1M of sales may approach 4x while one under $200k trades just above 2x. (BizBuySell, software and SaaS sold-listing benchmarks (2021-2025): https://www.bizbuysell.com/learning-center/valuation-benchmarks/software-apps-saas/)
- SaaS Capital's 15th annual survey of more than 1,000 private B2B SaaS companies puts 2025 median growth at 22%, down from 25% in 2024. Only 7.3% of companies were flat or shrinking, with bootstrapped companies at 20% median against 25% for equity-backed. That is the growth backdrop any SaaS deal's projections get underwritten against. (SaaS Capital, 2026 private B2B SaaS growth rate benchmarks: https://www.saas-capital.com/research/private-saas-company-growth-rate-benchmarks/)
- The publisher names low owner involvement among the drivers that put a software business at or above the upper quartile, alongside consistent financials, growth potential, unique competitive advantages and a seller willing to finance. Full-time owner involvement, a small niche and many direct competitors push it to the bottom. Nobody at this bar quantifies what documentation is worth, so treat the effect as directional. (BizBuySell, software and SaaS sold-listing benchmarks (2021-2025): https://www.bizbuysell.com/learning-center/valuation-benchmarks/software-apps-saas/)

## Who Else Is Buying in This Industry

- Valsoft Corporation (Montreal, Canada): A Montreal buy-and-hold acquirer of vertical-market software, ranked the most active strategic SaaS buyer two years running, buying mission-critical software across thirty-plus niches and keeping it. Newest here: Mirus Information Technology Services, Inc. · 2026 · A Houston restaurant intelligence and data management software company, bought through the Edelweiss group. 24 more confirmed on its profile. (https://searchspheresource.com/buyers/valsoft)
- Banyan Software (Atlanta, Georgia): An Atlanta permanent-capital acquirer of vertical-market software that buys, holds, and grows enterprise B2B software businesses for life, preserving founder legacies across the US, Europe, and beyond. Newest here: HMM · 2026 · A German B2B eHealth platform for managing medical-aid approvals, billing, and payments for health insurers and providers. 7 more confirmed on its profile. (https://searchspheresource.com/buyers/banyan-software)
- saas.group (Distributed, with US bases in Seattle and Las Vegas and offices in Germany and France): Profitable, product-led SaaS companies at $1M to $10M ARR with five or more years of history, bought whole and run inside a portfolio of twenty-four brands rather than merged away. Newest here: Ayrshare · 2025 · API-first social media management platform, the group's twenty-fourth acquisition. (https://searchspheresource.com/buyers/saas-group)

## What It Costs to Replace the Owner

The multiples are quoted on SDE, which adds the owner's pay back into earnings, so they hold only if you do the owner's job. For this trade the replacement is usually the technology manager who owns delivery, paid a median of $175,140 a year nationally; at a 3x multiple that wage takes about $525,420 off what the business is worth to you. Computer and information systems managers, BLS Occupational Employment and Wage Statistics (2025), national, all industries, before payroll taxes and benefits (https://searchspheresource.com/data/manager-wages).

## The Numbers That Run This Business

- MRR movement (new, expansion, churn)
- Logo and revenue churn by cohort
- Support ticket load
- Infrastructure cost per customer
- Failed-payment recovery rate

Site index for machines: https://searchspheresource.com/llms.txt