The RevOps metrics that matter. Six numbers, not sixteen.

I'm Lauren Pearson, and when a founder asks me which RevOps metrics matter, they are usually holding a dashboard with thirty numbers on it and a genuine sense that most of them are noise. They are usually right. The RevOps metrics that matter are a short list, not a long one, and the short list only works once you know which numbers are the outcome and which ones are just feeding it.

The short answer. Six numbers, not sixteen.

Pipeline coverage tells you whether there is enough open pipeline, typically three to five times the quarter's target, to hit the number even after the usual share of deals falls away. Win rate tells you how much of that pipeline actually closes, and whether that rate is moving in the right direction stage by stage, not just at the final gate. Sales cycle length tells you how fast a deal moves once it enters the pipeline, and a lengthening cycle is often the earliest sign something upstream has changed before it shows up anywhere else. Forecast accuracy tells you whether the number your CRO is telling the board can be trusted, measured as the gap between what was called and what actually closed. Net revenue retention tells you whether the revenue you already won is expanding, holding, or quietly leaking out the back through churn and downgrades. A data quality or adoption score, field completion rates, duplicate rates, login frequency, tells you whether the other five numbers are measuring something real or measuring a CRM full of stale records.

Every other metric a RevOps function tracks, activity counts, email opens, dashboard logins, individual rep scorecards, exists to explain movement in one of these six. It should not sit on the leadership dashboard as a metric in its own right.

How it works in practice. Leading indicators feed the lagging ones.

The six above are mostly lagging: they tell you what already happened. What actually lets you act before the quarter is over is the leading indicator underneath each one. Speed to lead, how long a new lead waits before first contact, feeds stage conversion, which feeds win rate. Stage conversion at each individual gate, not just the overall close rate, feeds sales cycle length, because a deal that stalls at one specific stage repeatedly is a process problem, not a market problem. Data quality feeds everything downstream of it: a pipeline coverage number built on stale stage dates or duplicate records is a confident-looking wrong answer, and a wrong answer with confidence attached is more dangerous to a forecast than an honest "we don't know".

Part of why this gets confused in practice is tooling. RevOps Co-op's 2025 State of RevOps report found that 40% of enterprise companies now run more than sixteen tools in their revenue stack, and 20% run more than twenty. Each tool tends to arrive with its own dashboard and its own version of a metric that looks similar to one already tracked elsewhere but is not calculated the same way. A win rate pulled from the CRM and a win rate pulled from a separate forecasting tool can legitimately disagree, and most teams never reconcile which one is the source of truth until a board meeting forces the question. Pick one system of record for each of the six core metrics before adding a seventh tool that promises a better view of the same number.

A concrete version of the chain: speed to lead drifts from same-day contact to a three-day average, usually because a rep is out or a routing rule broke quietly. Stage-one conversion on those late-contacted leads drops within the month, because the buyer has often engaged a competitor by the time your rep calls. Win rate on that cohort falls six to eight weeks later, once those deals reach a close date. If the only number leadership is watching is win rate, the actual cause, a broken routing rule, is six to eight weeks in the past and forgotten by the time anyone notices the lagging number move. Watching the leading indicator catches the problem while it is still a two-day fix rather than a quarter's worth of lost pipeline.

Review cadence should match how fast each number can realistically move. Pipeline coverage and speed to lead are worth a weekly look, since they change fast and a broken routing rule left for a month compounds. Win rate and sales cycle length are more useful reviewed monthly, since a handful of deals closing in either direction can swing a weekly number without meaning anything. Net revenue retention and forecast accuracy are quarterly by nature; checking them weekly mostly adds noise, because neither one is designed to move on a weekly cycle.

What good looks like. A dashboard someone can actually read.

I ask every RevOps dashboard proposal one question: what decision changes if this number moves? If nobody in the room can answer that, the metric does not go on the dashboard, however easy it was to pull. A US-based SaaS team I worked with had a leadership dashboard running to roughly forty metrics, most inherited from tools added over several years rather than deliberately chosen, and nobody outside RevOps opened it more than once a quarter. Cutting it to the six core metrics, each with a named owner accountable for explaining any move in it, made the weekly pipeline review noticeably shorter and gave the CRO a number they were confident enough to repeat to the board without a caveat attached.

Good also means every metric has an owner. A number with no name attached to it drifts; it gets reported because a dashboard generates it automatically, not because anyone is actually watching it move and asking why. Assign each of the six to whoever is closest to the lever that moves it, sales cycle length to the sales manager, data quality to whoever owns CRM administration, rather than defaulting all six to RevOps by title.

Pitfalls to avoid. Where RevOps metrics programmes fail.

The first pitfall is chasing forecast accuracy directly rather than fixing what feeds it. A forecast built on loosely defined pipeline stages will not get more accurate because leadership asks for more accuracy; it gets more accurate once the stage definitions and the data quality underneath it are fixed. I cover the mechanics of getting the CRM itself to enforce that discipline in how the CRM functions as RevOps' system of record.

The second is copying another company's benchmark dashboard wholesale. A sales cycle length that looks alarming next to a published industry figure may simply reflect a longer, multi-stakeholder deal structure that has nothing wrong with it. Benchmark against your own trend over time before benchmarking against someone else's business model.

The third is treating activity metrics as proof of health. Calls made and emails sent measure effort, not outcome, and a team can hit every activity target while pipeline coverage quietly falls apart underneath it. Activity metrics are useful for coaching an individual rep. They do not belong next to the six numbers a CRO actually needs to answer "are we going to hit the number".

The fourth is adding a metric and never removing one. Dashboards accumulate; they rarely get pruned. Review the full list at least once a quarter and ask, honestly, whether each number still changes a decision, or whether it survived only because nobody got around to deleting it.

The fifth is reviewing leading and lagging indicators in the same meeting at the same level of urgency. A weekly pipeline review that spends equal time on speed to lead and net revenue retention teaches the room to treat a number that should be checked quarterly with the same reactive energy as one that needs a same-day fix, and that habit burns out a team faster than the actual workload does. Keep the fast-moving operational numbers in the weekly cadence and save the slower, structural ones for the review built for them.