Selling a small online business

Consider what happens to a profit figure nobody can verify. A buyer cannot tell your honest $2,000 a month from someone else's optimistic $2,000 a month, so they price the category rather than the business — and the category price already has every unverifiable seller's claim baked into it. Evidence is not how you prove you are honest. It is how you stop being priced as though you might not be.

We see the same problem from the other side, and we have not solved it either. Our pipeline records what a source page states, and a stated profit stays a stated profit: it is not audited by the marketplace, by the buyer, or by us. That is why several of the medians we publish sit below one year of stated profit — some of those denominators were never real, and the sale price is the market's verdict on them. Your evidence pack is what stops your business being read as one of those rows.

Whatever you put in the listing, a serious buyer converts it into a multiple and compares it with sales they can point at. You may as well do that first, on the same data, so the first number in the conversation is yours and is defensible. Our calculator prices a description against sold comps and returns a range; the methodology says exactly how, including when it refuses to produce one.

Two things to take from the comp set before you take a number from it. First, the range is wide — comparable businesses genuinely sell at prices that are not close to each other, which is its own page. Second, the comps move: deals land weekly, so a range you pulled two months ago is not the range a buyer will pull today.

Six things, each with the same test: can a stranger confirm it without taking your word for anything, in an afternoon, before they have paid you? If the answer is no, assume it is discounted to zero.

1. Revenue, from the system that produced it. Processor and network dashboards — Stripe, PayPal, the ad network, the affiliate console, the app store — with at least twelve months visible, shown live in a screen share rather than as an image. A screenshot is an assertion about a dashboard; the dashboard is the evidence.
2. Costs, so that profit is profit. Hosting, tools, contractors, content, ads, the subscriptions you forgot about. Every cost you leave out is one a buyer will find and re-price the deal around, and finding one tends to cost you the benefit of the doubt on everything else.
3. The add-backs you are claiming, listed separately. If your profit figure is an SDE figure, say which owner expenses were added back and why. Some survive a buyer's scrutiny and some do not — which add-backs survive goes through them.
4. Traffic or users, at source, with history. Analytics access, not a screenshot, covering long enough to show whether the trend is a recovery or a decline. The same trailing-twelve-month number means very different things in those two cases.
5. Proof that it can actually be handed over. Domain registrar access, the accounts the revenue lands in, the ad or affiliate relationships that may not be transferable at all, anything that lives in your personal account. A business that cannot transfer cleanly is not worth its comps regardless of what it earns.
6. The operating record. What you actually do each week and how long it takes. A business needing forty hours a week is partly being sold as a job, and a buyer who discovers that after agreeing a price will re-trade it.

The quickest way to lose a credible buyer is to be vague about units, because from the outside it is indistinguishable from being deliberately vague about units. Three rules, applied to every number in the listing:

• Say whether a figure is monthly or annual, and use one of them throughout. Multiples get quoted both ways and the gap is a factor of twelve — see monthly vs annual multiples.
• Say whether it is revenue or profit, and if profit, profit after what. The two bases are not interchangeable and should never be compared against each other: revenue multiple vs profit multiple.
• Say what the headline price would include. Inventory, cash, a domain, an earnout, a transition period — a total hides its parts, and what a headline sale price includes is where deals quietly stop being comparable.

Price the ask against sales, not against other asks. Scanning live listings tells you what sellers hope for, and at this end of the market hope and outcome are not close: the gap between asking and sold prices runs in one direction, which is a page of its own. An ask built from other asks compounds everybody's optimism, and the market clears it by not clearing at all.

A rule of thumb is not a substitute either. The commonly quoted multiples describe screened, brokered, verified businesses, which may or may not be the sample your deal belongs to — we hold that up against our own sold prices in do the valuation rules of thumb hold up?

Whatever number you land on, be able to say where it came from in one sentence: these comps, this many of them, this basis, this size band. A seller who can do that is negotiating. A seller who cannot is waiting to be corrected.

We track sold outcomes, with a source URL each. We do not track how a business was prepared, how long it sat, how many buyers looked at it, or what it was asking first — so nothing on this page is a measured claim that preparation raises your price by any particular amount. We would rather publish the gap than fill it: see where sold-price data comes from.

What the data does support is narrower and still useful. A stated profit is only ever a stated profit to everyone downstream of you — the marketplace, the comp set, us. Every discount a buyer applies for that uncertainty is a discount your evidence pack is competing against, and that is a reason to build it that does not depend on us having measured its effect.