The Dead Zone

Four pieces in, and it’s the same problem each time, wearing a different function’s name. A function optimised on its own. Visibility split across teams. A number that looked fine and wasn’t. A number that showed up too late. This time it’s segmentation.

A new logo signs. Someone assigns a coverage tier, usually based on contract value, sometimes on contract potential, and moves on. That single number, decided once, at the point of least information about the account, quietly determines how much attention it gets for years.

The real problem isn’t which method gets used to set it. It’s that whichever one does, it’s built on a partial story, whatever was known and true the day the account signed, and nobody owns keeping it current after that. Some companies try to fix this with more granular segmentation, more tiers meant to capture more nuance. It rarely holds up any better, a more granular model just has more categories nobody’s kept accurate, not fewer stale ones. Other companies go the other way, a small number of broad tiers, and miss real differences between accounts that happen to land in the same bucket. Too granular or not granular enough turns out to be the wrong question. The real question is whether anyone revisits it at all.

One of many scenarios: when the tier is set on potential rather than current value, that judgement often sits with Sales, folded into net new logo value or some other catch-all growth metric. The account’s long-term owner may never see it or have a say in it. The number that shaped how the account gets covered had nothing to do with coverage. It came from how someone else’s number got built.

Two different leaks hide behind that same decision, and most companies are only watching for one of them. The first is the obvious one, a large contract that starts slipping, retention dropping on an account big enough that someone notices. The second is quieter and more common, a large portfolio of accounts that never expands, none of them churning, all of them flat, individually too small to trigger any alarm, collectively a meaningful share of the base going nowhere. Low churn gets read as health. It’s often just the first leak’s absence, not evidence the second one isn’t happening.

Why it happens

Rule of 40 is worth bringing in here, less as a target than as a warning about what a blended number can hide. Boston Consulting Group’s (BCG) 2025 benchmark of more than 100 private B2B SaaS companies found that businesses retaining current customers while still capturing new ones were the ones most likely to clear the bar, with gross revenue retention specifically named as one of the two metrics sitting at the intersection of top performance. That’s real and useful, and it’s also easy to satisfy on paper without fixing anything underneath it. Growth plus margin adding to a healthy score says nothing about where the growth actually came from, new logos, existing accounts expanding, or existing accounts simply not leaving yet.

A company can hit the number for years on new-logo growth alone while a large share of its existing base sits flat. The score won’t flag it. Flat isn’t churn, and churn is the thing segmentation-by-size tends to make visible, not the only thing worth watching for.

BCG’s newer research names Rule of 60, the same formula at a higher bar, as what AI-transformed companies should be aiming for next. The blending problem doesn’t disappear at the higher number, it just carries forward, same formula, higher threshold. Whichever version a company is chasing, the real question stays the same: growth from where, and margin built on what.

What leaders get wrong

Leadership checks segmentation the way they check most things, does the model exist, are tiers assigned, is coverage staffed against it. What gets missed is that the model was only ever accurate for the moment it was built, and nobody owns revisiting it as accounts drift from the profile they were assigned on day one. Segmentation gets treated as settled fact. It’s a hypothesis that was true once and has never been checked again.

The account owner who inherits a tier didn’t set it and can’t change it. Whoever owns the segmentation model, if anyone clearly does, isn’t watching individual accounts closely enough to catch a large, quiet portfolio going nowhere, because nothing about it is loud enough to demand attention on its own.

What it costs to leave alone

A company can hold a healthy growth and profitability score for years while a meaningful share of its base never expands, and the score won’t warn anyone, it was built to catch shrinking, not stalling. By the time it shows up, it isn’t one account anymore, it’s a whole cohort, all landing at the same conclusion at once, for reasons that were visible individually for years and simply never got added up.

The model doesn’t need to be perfect at signing. It needs an owner who keeps checking whether it’s still true.