crm
A Pipeline That Tells You the Truth
Most contractor pipelines forecast beautifully and predict nothing. The fix is not better percentages, it is stages defined by what the customer did.
A pipeline is supposed to answer one question: what is likely to happen next month. Most of them answer a different one, which is how optimistic the person who last touched each deal was feeling.
The failure is structural rather than personal, and it has a single cause. The stages are defined by what the seller did instead of by what the customer did.
Seller verbs and customer verbs
Look at a typical set of stages in a small contracting business. Contacted. Quoted. Following up. Negotiating. Closing.
Every one of those is something the seller did. None of them requires the customer to have done anything at all, which means a deal can advance through four stages while the customer remains entirely passive. "Following up" in particular is a stage that measures effort and reports it as progress. A deal can live there for a year.
Now the same pipeline written with customer verbs. Asked for a price. Opened the proposal. Asked a question about it. Agreed the scope. Signed. Paid the deposit.
Each of those is an observable event that either happened or did not. You cannot talk a deal into the next stage, and you cannot advance one by feeling good about it. The pipeline stops being a mood ring and starts being a record.
The two numbers that make it predictive
Once the stages are events, two simple numbers do most of the forecasting work, and neither requires a weighted percentage.
Stage conversion. Of the deals that reached "opened the proposal" in the last six months, what fraction reached "signed"? That is a measured rate, not a judgment, and it is specific to your business, your pricing and your market. It will not match the number in any article, including this one.
Time in stage. How long does a deal that eventually signs normally sit at each step? Once you know that, a deal sitting at twice the normal time is not "still warm". It is statistically a loss that has not been written down yet.
The second number is the one that changes behaviour, because it makes staleness visible without anybody having to make a judgment call about a customer they liked.
Aging beats percentages
The common alternative is to assign each stage a probability and multiply. Quoted is 40 percent, negotiating is 70 percent, and the forecast is the sum.
This fails in a specific and predictable way: the percentages are invented once and then never revised, so the forecast inherits the optimism of whoever set them up, permanently, and expresses it to two decimal places. Worse, it hides age. A deal quoted yesterday and a deal quoted in March are both 40 percent.
Aging does not have that problem. A list of open deals sorted by days since the last customer-side event is blunt, requires no model, and is almost always more informative than the weighted total. The oldest third of that list is usually dead, and seeing it written down is what gets it closed out.
Closing the dead ones is the point
Nobody enjoys marking a deal lost, and in a small business there is rarely any pressure to. The deal costs nothing to leave open, so it stays open, and the pipeline slowly fills with history dressed as opportunity.
The cost is real but indirect. A pipeline that is forty percent dead cannot be used to decide anything: not whether to hire, not whether to buy the second truck, not whether this month's slow start is a problem or a blip. Those are the decisions the number exists to support, and they are the ones a padded pipeline quietly breaks. General guidance on managing the cash side of exactly these decisions is collected in the SBA's manage your finances material, and all of it assumes the inputs are honest.
A lost deal is also better data than an open one. Lost with a reason, over a year, is the only report that tells you whether you are losing on price, on timing, on availability, or on something you could fix next week.
The weekly pass
The whole discipline fits in fifteen minutes a week, and it is worth more than any dashboard.
Open the list sorted by age. For each deal older than your normal time in stage, pick one of three answers. There is a next step with a date, in which case write it down. There is no next step, in which case close it lost with a reason. Or you genuinely do not know, in which case the action is to ask the customer, today, in one sentence.
Deals do not resent being asked. They resent being chased in a way that pretends not to be a chase. "Are you still planning to do this work this year, or should I close the file?" is a question most customers answer honestly and quickly, and either answer is better than the silence it replaces.
What to measure after a quarter
Three things, and none of them is pipeline value.
How many deals reached a customer-side event. How long the average signed deal took from first contact. What fraction of closed-lost deals had a reason recorded.
The third is the one that tells you whether the pipeline is being used or performed. A business where most losses have no reason attached is not tracking its pipeline. It is tidying it.