A multiple is a conclusion, not an input
Sector ranges are useful for orientation. Vertical B2B SaaS typically trades above mature enterprise software in the middle market, and AI platforms with defensible data trade higher still.
What those ranges do not tell you is where inside them a specific asset lands, and the spread is the whole conversation. On a $30M-revenue business, every turn of revenue multiple is $30M of enterprise value. Sector explains almost none of that. Four other things explain most of it.
One: how durable the revenue is
Net revenue retention above 110% moves an asset toward the top of its band. Below 90% moves it toward the bottom, or out of the band entirely.
Buyers weight this heavily because it is the input their model is most sensitive to and the one a seller can least easily dress up. Gross retention matters alongside it, because net retention above 100% can hide serious churn masked by expansion in a handful of accounts. A buyer will ask for both, split by cohort, and will build the curve themselves from raw data.
Two: how concentrated it is
Concentration is priced directly and the curve is not linear. Below roughly 10% for the largest customer, buyers barely comment. Between 10% and 20%, it appears in the risk section. Above about 20%, it starts moving the number. Above 35%, several buyer types stop bidding, because the asset is no longer a software business in their model.
The reason is simple. Diversified revenue is a statistical pattern that will broadly repeat. Concentrated revenue is a decision that one procurement officer will make next year, and no model can underwrite that.
Three: what it costs to run alone
For a carve-out this is often the largest single adjustment, and it is the one sellers most often get wrong.
A division inside a group consumes finance, HR, legal, IT, security and procurement without seeing an invoice. Group accounting allocates a share of that cost by formula, usually as a percentage of revenue set years ago for management reporting. That number is almost never what the business would pay as an independent company.
Buyers build their own standalone model. Where their number is higher than yours, the difference comes out of EBITDA and then out of the price, multiplied by whatever multiple you had agreed. A seller who has done this work properly and can show the basis of each line removes the largest source of re-trading in the process.
Four: how cleanly it separates
Perimeter and separability affect the multiple through two channels. They change the cost of the transaction for the buyer, and they change how many buyers can bid at all.
An asset with its own codebase, its own infrastructure and its own team can be bought by a financial sponsor with no operating capability. An asset entangled in a shared platform can realistically only be bought by a strategic acquirer who can absorb it. That halves the field, and a smaller field means less competition, which shows up in the price more consistently than any of the operating metrics.
What this means for sector ranges
Use them to sanity-check an expectation, not to set one. If your adviser quotes a number at the top of the sector range, ask which of the four they are relying on and what the evidence is. If they quote it without reference to any of them, they are quoting the sector, which is the least informative thing you can know about an asset.
The one factor outside the asset
There is a fifth input, and it does not sit inside the business at all: how many credible bidders are in the room.
A prepared asset in front of a curated field clears materially higher than the same asset sold to the buyer who happened to call, and closes on better terms as well, because competitive pressure works on structure as much as on price. The gap between the highest and lowest credible bid in a middle-market process is commonly thirty percent or more.
That is worth restating, because it is the only lever that does not require changing the business. Everything else on this list takes two to four quarters. Running a real process takes four to eight months and is the last decision you make, not the first.
Divestitures.com Research · Published for orientation, not as advice on a specific transaction. Any figure cited is orientation, not a valuation. See market notes.