a partner at a growth fund once told a founder, in a first meeting, that the deck was unanswerable as written. "your slide says 8% net churn. i can't tell whether you have a $3m business that's recoverable or a $3m business that's dying. those are the same number and they aren't the same company." the meeting ended 12 minutes later.
the deck had collapsed two metrics into one. churn — the customer who cancels and stops paying. contraction — the customer who stays but downgrades. they show up in the same revenue waterfall but they're produced by different mechanisms and fixed on different timelines. lumping them costs you the next round because the people writing the check have seen both shapes and they want to know which one you are.
what looks like a single 8% problem is almost always a 2%-and-6% problem, or a 6%-and-2% problem. one of those is a quarter of work. the other is a year. the distinction decides the term sheet.
the four-tile view that separates the two
a $3m arr saas business loses revenue every quarter. how it loses revenue is the question. logo churn — accounts that fully cancel — measures whether the product is worth keeping at all. dollar contraction — accounts that downgrade — measures whether your packaging or pricing is right for who's left. expansion measures whether the customers who do stay are growing inside the product.
put the four numbers next to each other and the story tells itself.
the company above has a packaging problem, not a product problem. only 3 of 152 customers cancelled. they're downgrading. the value is real but the price ladder is wrong — probably an expensive tier customers are dropping out of, or a usage-based meter pricing above what they actually consume. that's fixable in a quarter with a packaging revision and a renewal motion that catches downgrades before they happen. bessemer's state of cloud breaks contraction and logo churn out as separate lines for exactly this reason.
invert the numbers and the story inverts with it. 6% logo churn with 2% contraction means customers are leaving entirely, not staying and complaining. the product isn't sticky. no packaging fix rescues this — it's a roadmap problem, a year of work, and the next round will price it accordingly. same headline, different company.
why founders collapse the two
the collapse usually happens for an honest reason: the data is in two systems. stripe shows the dollar movement. the crm shows the logo movement. nobody owns the join. so the cfo pulls a single revenue waterfall, computes net churn against the prior period, and writes one number on the slide.
the slide is technically correct. it's just useless. a partner who has seen 40 saas businesses reads 8% net churn three different ways, and the version they assume is the worst one because that's what their job rewards. by the time you're explaining the composition in q&a, the meeting has decided what it thinks of you.
the second reason is that contraction is psychologically heavier than churn. churn is clean — the customer is gone, the bookkeeping closes. contraction is the customer still in your seat, still telling you the product is great, while paying 40% less than they did six months ago. that conversation forces you to acknowledge a real problem in the value ladder. most founders prefer to roll it into churn and move on.
the two-quarter test
if logo churn is high and contraction is low, you have 12 to 18 months of product work before retention bends. the people in the cohort don't want what you're selling. there's no shortcut. the next round is a story about a roadmap, and the price reflects the risk.
if contraction is high and logo churn is low, you have one to two quarters of packaging and renewal work. the cohort still wants the product — they're staying — they just don't want the version they're paying for. that's a tractable problem: a tier audit, a renewal motion that lands 60 days before contract end, sometimes a meter recalibration. you walk in with two quarters of contraction data trending down and the term sheet writes itself.
the diagnostic takes one afternoon. pull every account that lost dollars in the trailing quarter. tag each as a full cancel or a downgrade. count the dollars in each pile. the ratio is your composition. you do not need a tool — you need the discipline to do it before the partner does.
what to put on the slide
split the line. "net retention dynamics: 2% logo churn, 6% dollar contraction, 14% expansion." the partner reads it in three seconds and forms the right opinion. better — show the four-tile view, label each, and let the composition do the talking. that's the slide that turns a first meeting into a second meeting.
the slide that says "net churn: 8%" is the slide that ends the meeting. the slide that breaks the 8% apart is the slide that opens the conversation about how you'll fix it.
how zift handles this
zift computes logo churn, dollar contraction, and expansion mrr as separate lines every fifteen minutes against the stripe + crm data it already ingests. on monday morning the briefing names the named accounts that contracted last week, the dollar amount, and the tier they dropped to — so the renewal conversation can start before the customer asks for the discount.
if you're a finance lead at a series b team running multi-product or multi-entity packaging where contraction shows up across skus, zift handles that too.
8% net churn is a number. the composition is the company.