One Metric That Matters: Picking the Number Your Team Actually Rallies Around

Most dashboards fail by being generous. They show forty numbers, and a team that is watching forty numbers is watching none of them. The One Metric That Matters, popularized by Alistair Croll and Benjamin Yoskovitz in Lean Analytics, is the discipline of choosing a single number that captures the most important thing happening in your business right now, and pointing everyone at it. This article covers how to choose an OMTM, why it changes over time, and how to avoid the trap of optimizing a number while the business quietly rots underneath it.

What OMTM actually is

The OMTM is the one number that, if it moves in the right direction, means you are winning at the thing that matters most this quarter. It is not your permanent north-star metric and it is not every KPI on your board. It is a deliberate, temporary focus that answers the question: given where the company is right now, what is the single most important thing to improve?

The word "temporary" is the part people miss. An OMTM has a shelf life. The number that matters when you are hunting for product-market fit is not the number that matters when you are scaling a proven model. Croll and Yoskovitz tie it to stages of growth: you find the OMTM for your current problem, obsess over it until you solve that problem, then deliberately pick a new one.

Why one number, not ten

Focus is the entire point. A single metric does three things a dashboard of forty cannot:

• It aligns the team. Everyone from engineering to sales can name the number and knows whether today's work moved it. Arguments about priorities shrink because there is a shared referee.
• It forces a hypothesis. Choosing one metric means saying out loud, "we believe this is the thing that matters most." That is a testable claim, not a vibe.
• It creates a comparison line. One number tracked consistently over weeks tells you whether your bets are working. Ten numbers all wiggling at once tell you nothing.

How to choose your OMTM

A good OMTM passes four tests.

1. It reflects your current stage. Pre-product-market-fit, you care about whether people come back — retention and activation. Post-fit, you care about efficient growth — revenue and acquisition cost. Do not measure scaling numbers before you have something worth scaling.
2. It is a rate or a ratio, not a vanity total. Cumulative totals only go up and always look good, which makes them useless. "Total signups" flatters you forever. "Weekly activation rate" tells the truth. Prefer numbers that can go down.
3. It is actionable. You must be able to draw a line from something your team does to the metric moving. If nobody can influence it, it is a scoreboard, not a lever.
4. It is comparative. You should be able to say "this week versus last week" or "this cohort versus that cohort." A number with no comparison is trivia.

Worked example: choosing between two candidate metrics

Imagine a subscription app with 10,000 signups. The founders are proud of that total. But watch what happens when you turn vanity into rate.

• Vanity metric: total signups = 10,000, and it only ever grows.
• Candidate OMTM: week-one activation rate — the percentage of new signups who complete the core action in their first seven days.

Suppose 10,000 signups produced 1,200 activated users, a 12% activation rate. That is the number that predicts whether growth compounds or leaks. If the team spends a quarter improving onboarding and pushes activation from 12% to 20%, then the same acquisition spend that used to yield 1,200 activated users now yields 2,000 — a 67% lift in the users who actually matter, with no extra marketing budget. The total-signups number could not have told you any of that. It would have kept climbing while the business bled out the bottom.

This is the whole argument for OMTM in one example: the right single number changes what your team does on Monday morning, and the wrong single number just makes everyone feel good.

Famous examples, and the caveat

You have heard the case studies: an early file-sharing product watching invites sent, a marketplace watching nights booked, a social network watching daily active users. They are useful illustrations, but treat them as prompts, not templates. Copying another company's OMTM is how teams end up optimizing a number that does not match their own business. The right metric comes from your product and your current problem, not from a keynote.

How the OMTM shifts across growth stages

Lean Analytics frames growth as a sequence of stages, and each one demands a different OMTM. Naming them makes the "temporary" idea concrete:

• Empathy — you are still learning whether a problem is real. The OMTM is qualitative: number of customer interviews that confirm the pain, or the percentage who describe the problem the same way.
• Stickiness — you have a product; does anyone keep using it? The OMTM is retention or repeat usage. Chasing acquisition here is premature; you would just pour new users into a leaky bucket.
• Virality — usage is sticky; can it spread? The OMTM becomes a referral or invite rate, the number of new users each existing user brings.
• Revenue — it spreads; does the math work? The OMTM shifts to something like revenue per customer or the LTV-to-CAC ratio.
• Scale — the model works; can you pour fuel on it? Now acquisition efficiency and channel economics take over.

The trap is skipping ahead. A team that fixates on acquisition (a scale-stage metric) while retention is still broken (a stickiness-stage problem) will burn money accelerating a car with no floor. The OMTM discipline forces you to solve the current stage before you buy a ticket to the next one, and it gives you an honest signal for when you have earned the right to move on.

How to track this in Inleads

An OMTM is only as good as the plumbing that keeps it visible. If your one number lives in a spreadsheet someone updates on Fridays, it is not really the metric that matters. Inleads is built around the AAARRR framework — acquisition, activation, retention, referral, revenue — which maps almost exactly onto how OMTM shifts as you grow, so the metric you care about this quarter already has a home.

The pirate-funnel analytics let you put your chosen number front and center and watch it as a rate over time rather than a growing total, and the product analytics tie that number back to the customer data platform so you can segment it — activation for last month's cohort versus this month's, revenue by acquisition channel — which is exactly the comparative view a real OMTM demands. When you are ready to change focus as your stage changes, you are re-pointing a dashboard, not rebuilding your reporting.

For teams that want the number where they already work, the MCP server lets you ask an AI assistant in Claude, ChatGPT, Cursor, or Windsurf about your live metrics in plain language, so checking whether your OMTM moved does not require opening a separate tool. The goal throughout is the same: make the one number impossible to ignore.

Keep it honest

The danger of OMTM is Goodhart's Law: when a measure becomes a target, it stops being a good measure. Pick daily active users and someone will find a way to inflate it with a pointless notification. Guard against this by keeping one or two "counter-metrics" in view — numbers that would get worse if you gamed the OMTM — and by revisiting the choice every quarter. The metric that matters is not a number you set and forget. It is a decision you keep making on purpose.