Valuation guide
How to value a SaaS or app business with technical debt
A small software business can show steady profit while running on a framework two major versions out of date, with no tests, a deploy process that lives in the seller's head and a server only they can reach. None of that appears in the profit and loss. All of it is work the next owner will have to pay for.
Technical debt is a cost that has not been paid yet
The trailing profit of a software business is what it earned while the seller chose not to spend on upgrades, rewrites or hardening. Some of that is fine — not every old dependency needs replacing. But where the code needs work just to keep running safely, the profit figure is flattered by exactly that unspent cost, in the same way an unpaid bill flatters a bank balance.
So the useful question is not "how much technical debt is there?" but "what will I have to spend, and how soon, to keep this earning what it earns now?" That is a number you can estimate, and once you have it, it comes off the price rather than being folded into a vaguer sense that the business is "risky".
Separate the debt that must be paid from the debt you can live with
- Forced work. A runtime, framework or API version that is reaching end of life; a payment or app-store integration the provider is retiring; a dependency with a known security issue. This has a deadline set by someone else. Estimate it and deduct it.
- Fragility. No tests, no staging environment, manual deploys. Nothing has to change, but every change is slower and riskier. This mostly costs you time; price it if you plan to build on the product, discount it less if you plan to run it as it is.
- Knowledge held by one person. Undocumented setup, credentials only the seller has, a build that works only on their machine. This overlaps with owner dependence, and the remedy is the same: write it down and test it during the handover, before the money moves.
- Code you would simply have written differently. Unfashionable is not the same as broken. Do not deduct for a rewrite you would like to do but do not need.
How to find it before you make the offer
- Ask for read access to the repository under an NDA, or a screen-share walk-through if the seller will not grant it. A business that cannot show its code to a serious buyer is one you are pricing blind.
- List the runtime, framework and major dependency versions, and look up each one's support status yourself. This is the quickest way to find forced work.
- Watch a deploy. Ask the seller to ship a trivial change while you watch. How long it takes, and how many manual steps it needs, tells you more than a description of the process.
- Read the hosting and third-party bills. They show what the product actually runs on, and they are part of the costs that make up the profit figure. See how to verify a seller's revenue and profit.
- Get an outside estimate if you cannot read the code yourself. A few hours of a developer's time is small against a mispriced purchase. Ask them for a cost and a deadline for each item, not a grade.
The broader checklist is in due diligence for sub-$100k deals; what changes hands at closing is in transferring a business after the sale.
What our SaaS and app comps can and cannot tell you
Figures read from /api/stats when this page loads.
Our sold records hold the price, the stated revenue and profit, the business's age where the source gives it, the sale date and the source's own title. They do not record code quality, test coverage, framework versions or hosting setup. So the figures below are what SaaS and app businesses sold for, with whatever technical debt each happened to carry — a median over a mix of clean and neglected codebases that we cannot separate.
In practice that means: use the comps to price the business as if it were a typical sale of its type, then deduct your estimate of forced work from section 02 yourself. Do not expect a lower multiple in the data to have done that deduction for you.
| Asset type | Comps with stated profit | Median × profit | Middle half sold for |
|---|---|---|---|
| Loading live figures from /api/stats… | |||
The first two columns cover every non-demo sold deal of that type with a stated profit. The price range is measured over the paid report's window — the most recent 200 priced sales of that type — on the profit basis. A dash means we hold fewer than five comparables on that basis and publish nothing.
Getting a range instead of a number
Price the business against sold comps first, then subtract the cost of the work that cannot wait. The calculator compares a profit figure with the SaaS and app sales we hold and shows how many there were; the paid report adds public-web comparables, a risk register and questions to put to the seller.
Related: what a SaaS business is worth · what a mobile app is worth · platform dependence · methodology · all valuation guides