The Signal Read
Exit diligence used to be a financial review with a CRM export attached. It isn't any more. Buyers go into the system and read it directly — stage histories, contract terms, account-level concentration — and they know what they are looking for.
Three findings recur often enough to be predictable.
Pipeline staged late. Opportunities advanced near period close to flatter win rates. Contract mix presented generously, where month-to-month or annually-cancellable revenue is described as recurring. And customer concentration the seller cannot explain — top accounts carrying an outsized share of ARR with no articulated reason those accounts are safe.
Each of these compresses the multiple. Not because the business is bad, but because the buyer now has to price the uncertainty you left them.
The Decision Tree
Work in order of remediation clock, longest first. This is the opposite of the instinct, which is to fix the easiest thing.
Concentration — start immediately. It has the longest clock by far. You cannot diversify a customer base in two quarters; you can only start earlier. If the top three accounts exceed roughly a third of ARR, the work begins now or it does not happen.
Contract mix — one renewal cycle. Every month-to-month agreement that should be annual is a conversation you can have at the next renewal. Miss the cycle and you wait a year.
Pipeline hygiene — two to three quarters. Long enough that it cannot be done reactively, short enough that it is recoverable if you start before the process opens.
The narrative — continuous. The explanation of why the numbers look the way they do should be written while the numbers are being made, not reconstructed afterward from memory.
The Hidden Signal
You cannot retroactively clean a stage history, because the evidence is the shape of the data over time — and reshaping it late is itself visible.
This is the part most sellers miss. A buyer examining stage durations across eight quarters is not reading any individual opportunity. They are reading a distribution. If that distribution is irregular for six quarters and immaculate for the most recent two, the correction is more informative than the original problem: it says the seller knew, and cleaned up in preparation.
The asymmetry that follows is the entire argument for starting early. These findings take 18–24 months to remediate and zero weeks to discover. A buyer's analyst finds all three in an afternoon. There is no version of this where preparation beginning at the data room is preparation at all.
What This Looks Like With Real Data
Clean looks boring, and boring is the point.
Stage durations sit in a stable distribution across periods, with no clustering at quarter boundaries. Contract mix is stated plainly — a specific percentage of ARR under annual or multi-year terms, with the month-to-month portion named rather than blended away. Concentration is disclosed with its reasoning: why the top account is embedded, what the renewal history looks like, what would have to happen for it to leave.
The defensible position isn't a spotless record. It's a documented one, built against a signed baseline while the work was happening — so when the data room opens, the artifacts already exist and the story is the same one you have been telling for two years.