What actually counts as a leading indicator in a service business?
Anything that moves before revenue and that you can influence this week. Booked backlog for the next seven days, unsold estimates aging, call answer rate, capacity fill and memberships due for renewal are leading. Revenue, average ticket and completed jobs are lagging, reporting decisions made days or weeks earlier. Put leading indicators at the top of the page and lagging ones underneath.
Test whether your indicator actually leads
People assign the label by intuition and are often wrong. There is a simple check: line up the candidate metric against revenue and compute the correlation at several lags, shifting the candidate forward by one day, three days, a week, two weeks. If the strongest relationship sits at lag zero, the metric is coincident, not leading, and watching it buys you nothing.
Do this once per metric and you will drop half of what is currently called leading. You will also discover the true lead time, which tells you how far ahead a change can be seen and therefore how much room you have to react.
One caution on the method: correlation at a lag can be an artifact of both series sharing a seasonal pattern. Remove the seasonal component from both before comparing, or every metric in the business will appear to lead every other one.
Leading indicators by horizon
- Hours. Unanswered and abandoned calls, unassigned jobs, technicians idle against a full board, same-day cancellations.
- Days. Booked backlog for the next week, capacity fill percentage, estimates awaiting follow-up, jobs sitting in a pending status.
- Weeks. Total value of unsold estimates and its age distribution, memberships due for renewal, lead volume by source, review flow.
- The overlooked one. Handling quality on inbound calls. It moves booking rate before booking rate moves revenue, which is why call analysis functions as an early warning system rather than a QA archive.
The trap: leading indicators that are mostly noise
The earlier a metric sits in the chain, the more volatile it is. Daily lead count leads revenue and also swings wildly, so watching it raw produces constant false alarms. Smooth it with a trailing window and compare against the same weekday, or you will spend your attention on variance.
The rule of thumb: the further ahead an indicator sees, the more smoothing it needs and the wider its normal band should be.
Every leading indicator needs an owner and a move
A leading indicator with no available action is trivia. Unsold estimate value implies a follow-up motion. Capacity fill implies a scheduling or marketing lever. Answer rate implies staffing. Write the action next to the metric.
That pairing is what turns revenue reporting from a rear-view mirror into something the team uses on Monday morning.
Order the page accordingly. Leading indicators with actions at the top, lagging results underneath as confirmation. Most dashboards are built in the reverse order, which is why they read as a report card rather than a control panel.
Topics: leading indicators · KPI selection · backlog · forecasting
Have a version of this question about your own business?
The useful answer usually depends on which systems you run and how they're connected. That's a conversation, not a blog post.