What is a leading indicator?
Also known as
Leading and lagging indicators
A leading indicator is a metric that moves before the outcome it predicts. A lagging indicator moves after, confirming what already happened. Revenue is lagging. Qualified pipeline is leading. Both are useful, and they’re useful for different things: you steer by the leading one and you’re graded on the lagging one.
The distinction is about timing and controllability, not importance. Lagging indicators are usually the numbers that matter most to the business — revenue, churn, retention, profit — but by the time they move, the decisions that caused them are months old. Leading indicators are noisier and further from money, but you can still do something about them.
They come in pairs
The useful unit isn’t a single leading indicator, it’s a pair: a lagging outcome you’re accountable for and one or two leading measures you believe drive it. Some examples across functions:
- Sales — closed revenue is lagging; pipeline coverage, meetings booked, and sales cycle length lead it. Coverage below target today is next quarter’s miss.
- Support — CSAT is lagging; first response time and ticket volume lead it. Response time slipping this week shows up in satisfaction scores next month.
- Software delivery — incidents and outages are lagging; lead time for changes, deployment frequency, and change failure rate lead them.
- Customer success — renewal is lagging; product usage, support sentiment, and a composite health score lead it, often by a full contract cycle.
Choosing one that’s actually predictive
A leading indicator is a hypothesis: this moves, therefore that will move. Test it. Pull both as a time series, shift the leading one forward by the lag you expect, and see whether the relationship holds in your data. Plenty of dashboards carry “leading” metrics that correlate with nothing.
Two things to watch. First, leading indicators are easier to game precisely because they’re closer to the work — activity metrics like calls made or tickets touched go up on command without any outcome improving. Second, don’t drop the lagging metric once you have a leading one; the lagging number is how you find out your hypothesis stopped being true.
A practical setup: a small set of leading indicators reviewed weekly by the team that can move them, and the lagging outcomes reviewed monthly by whoever owns the result.
Related terms
Put it to work
- Pipeline coverage — Metric
- First response time — Metric
- Lead time for changes — Metric
- Customer health score — Metric
