Leading vs lagging indicators. Manage the inputs, not the outcomes.

Lagging tells you what happened. Leading is what you can change.

I find it easiest to think about this in an Atomic Habits way. The lagging indicator is the result you care about — your weight, in the personal example. But if all you ever look at is the result, you're powerless over it; by the time the scale moves, the behavior that moved it is weeks in the past. The leading indicators are the controllable inputs: the calories, the workouts, the drinks, the stress. Those you can actually do something about today.

In a SaaS business it maps cleanly. MRR, revenue, churn — those are the weight. They're the numbers the business ultimately needs to manage, but you can't reach in and change them directly this week. The leading indicators are the calories: leads, demos scheduled and completed, win rate, activation, product adoption. Those are the inputs you can put your hands on now, and they're the ones that will move the lagging numbers a month from now.

An all-lagging scorecard turns your meeting into a guessing game

Here's why a board of only lagging numbers is a real problem, beyond just being late. When MRR is down or churn is bad this week and the scorecard shows you nothing but the outcome, your weekly meeting degrades into speculation. The leadership team sits there generating hypotheses — maybe it's seasonality, maybe it's that one big account, maybe sales is distracted — with no data to tell them which guess is right.

The fix isn't to drop the lagging numbers — you need them; they're what the business is actually trying to move. The fix is to look at both. Put the lagging outcome next to the leading inputs that feed it, and a red number stops being a mystery to debate and becomes a chain to trace. You're not guessing why churn spiked; you're looking at the adoption and activation numbers that were already telling you it would.

The chains we actually watched

On the go-to-market and sales side, we traced the chains deliberately. Two we watched closely:

Demos Scheduled → Demos Completed → New MRR. The scheduled demos are the earliest signal — if that number softens, you know weeks ahead that new MRR is going to soften too, long before it shows up in the revenue line. And Pipeline Coverage → Quota Attainment → Revenue: if coverage thins out, attainment and then revenue follow on a predictable lag. Watching the front of each chain bought us time to act while there was still time.

I'll be honest about where we were weaker, because it's the more useful admission. We didn't watch the product-adoption chain nearly as hard. We looked at DAU/MAU, but that's only loosely connected to what actually drives an individual account to churn — the real signal lives in account-level and cohort-level adoption, and we didn't trace that chain to churn the way we traced demos to MRR. If I were building the board again today, that's the gap I'd close first: activation and adoption are the leading indicators of retention, and we leaned on a lagging churn number longer than we should have.

A leading indicator you might not expect: the bug count

Leading indicators aren't always the obvious funnel metrics. One of the most useful we had was the bug count. A spike in open bugs is a genuine leading indicator, because it points at several things at once: product quality slipping, a feature that launched badly, an engineering team underwater. And it's a quiet predictor of churn. A bug backlog that stays elevated for weeks isn't just an engineering problem — it's customers steadily accumulating reasons to leave, well before any of them actually cancels.

That's the mindset to develop: asking, of any operational number, "what lagging outcome does this quietly predict?" A bug count looks like an engineering hygiene metric until you realize it's an early-warning system for retention. The best leading indicators often hide in one function and predict an outcome in another.

How to find yours

You don't have to theorize your leading indicators into existence up front. In my experience they surface naturally, right in the meeting: the moment a lagging number goes red and the leadership team starts asking "why did sales drop?" or "why is churn so bad this month?", the answers people reach for are your leading indicators. Someone says "well, demos were down three weeks ago" — there's one. Someone says "the activation numbers on the last cohort were soft" — there's another.

So the practical method is almost the reverse of what you'd expect. Don't sit in a room trying to brainstorm the perfect leading indicators in the abstract. Run the cadence, let the lagging numbers prompt the "why," and promote the recurring answers onto the scorecard as leading indicators in their own right. The conversation does the work of finding them.

The rule: two leading for every lagging

You do need a mix, and the gravity is all in one direction. The temptation is to fill the board with lagging numbers, because those are the things the business is on the hook to manage — MRR, churn, revenue. So I use a simple counterweight.

Applied across the board, that ratio forces the scorecard to be something you can act on rather than just something you report. Every outcome you care about ends up sitting next to the handful of inputs that produce it. And that's the whole point of leading indicators: they're the only numbers on the board you can actually do something about this week. The lagging numbers tell you whether it worked. The leading ones are where the work happens.