What is my open estimate pipeline actually worth?
Considerably less than its face value, and by an amount you can calculate. Group unsold estimates by age, multiply each group by the historical close rate for estimates of that age, and sum. That weighted figure is the realistic pipeline. Face value treats a ninety-day-old estimate as equal to one presented yesterday, which is why raw pipeline totals are almost always optimistic and almost never used for forecasting twice.
Face value is a sum of things that will not happen
An open estimate list is a list of possibilities, not assets. Yet most pipeline reports total the totals and present the sum as if it were pending revenue. The number is enormous, it never comes in, and after a couple of quarters people stop looking at pipeline reports entirely.
The fix is not to be more pessimistic. It is to weight each estimate by the probability that estimates like it actually sell.
Building the weighting from your own history
Take every estimate presented over the last year or two. For each, record whether it eventually sold and how many days elapsed between presentation and sale. Now bucket by age at the time of measurement and compute, for estimates that had reached that age unsold, the share that eventually closed anyway.
That produces a table: estimates zero to three days old close at one rate, four to fourteen days at a much lower rate, and so on down a long tail. Apply the table to your current open list and you get a weighted pipeline. Run the same calculation separately for emergency work and planned work, because their curves barely resemble each other.
What the weighted number is good for
- Forecasting. Weighted pipeline plus expected new demand is a far better revenue forecast than either alone.
- Follow-up triage. The weights tell you where callback effort is worth spending and where it is close to free money left on the table versus close to wasted time.
- Capacity planning. Weighted pipeline by job type tells you which crews are about to be busy.
- Source evaluation. A source producing large estimates that never close is not producing revenue, and unweighted pipeline hides that completely.
The honest limits
Two things break this model. First, small samples. If a job type produces only a handful of estimates a month, its close-rate table is noise and should be merged into a broader category. Second, structural change. Introducing financing, changing who presents estimates, or entering a new market invalidates the historical curve, and you should say so on the report rather than quietly keep using it.
The model is also a probability, not a promise. It is useful for planning, and it is not a commitment you should make to a lender or a bonus plan. Keeping that distinction visible is part of building business intelligence people continue to trust, and it is the difference between a forecast and a wish. See revenue intelligence for how pipeline sits alongside booked and completed revenue.
Topics: estimates · pipeline · forecasting · aging · close rate
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.