How do I report on estimates that haven't sold yet?
Report on them by age, not just by total. Open estimate value is meaningless if nothing is ever dismissed, because the pipeline only grows. Split estimates into presented, sold, dismissed and stale, then track close rate by technician and by age band. An estimate older than your normal decision window is not pipeline. It is a follow-up task that nobody performed.
Why the open estimate total is almost always wrong
Estimates get created constantly and dismissed rarely. Nobody is rewarded for marking an old estimate dead, and there is always a chance the customer calls back. The result is a pipeline number that only ever climbs and cannot be used for forecasting anything.
The fix is not discipline alone. It is an aging policy: anything past a defined window moves to a stale bucket automatically and is excluded from pipeline totals unless someone deliberately revives it. Pick the window from your own data — look at how long sold estimates actually took to close, and set the boundary past the long tail.
Four numbers worth watching
- Presentation rate. The share of qualifying jobs where an estimate was created at all. A low close rate with a low presentation rate is a different problem than a low close rate with high presentation.
- Close rate by age band. Same-visit closes, closes within the follow-up window, and closes after that. The shape tells you whether follow-up is doing anything.
- Options per estimate. Whether customers were given a choice, and what the distribution of chosen options looks like.
- Stale value entering each week. The amount of quoted work aging out. This is the number that makes follow-up feel urgent.
Close rate needs a denominator you can defend
Close rate calculated as sold estimates divided by all estimates punishes the technician who writes an estimate on every appropriate call and rewards the one who only writes them when the sale is already made. That is exactly backwards as an incentive.
A better denominator is estimates presented on jobs that were genuine opportunities, with pure maintenance and warranty visits excluded. It takes more work to define, and it produces a number people will actually accept.
The conversation matters more than the field
A dismissed estimate records that the customer said no. It does not record why, and the why is where the coachable material lives — whether options were offered, whether financing came up, whether the diagnosis was explained in terms the homeowner understood.
That detail exists only in the recording and the notes. Running structured evaluation across those interactions is what rubric-based evaluation does, and the output supports human coaching decisions rather than replacing them. Paired with pipeline data, it turns a low close rate into a specific thing a manager can work on. See coaching support for how that loop runs.
Topics: estimates · pipeline · close rate · follow-up
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.