A pipeline is the most confidently presented and least reliable artefact in commercial diligence. It is a forward-looking number assembled from optimistic judgements by people whose pay depends on those judgements, recorded in a system nobody audits. It can still be enormously informative — but only if you read it as a record of behaviour rather than a source of numbers.
Stop asking “is this pipeline big enough?” Ask “does this pipeline behave like one that converts?” Coverage ratios can be manufactured in an afternoon. Behaviour cannot.
Start with what the stages actually mean
Every CRM has stages. Very few have stage definitions that describe something observable. The difference determines whether any pipeline analysis is worth doing at all.
A weak definition describes the seller’s state of mind: Qualified, Interest confirmed, Verbal. A strong one describes something the buyer did: Economic buyer identified and met, Success criteria agreed in writing, Procurement engaged. The first kind cannot be falsified, so it drifts upward under pressure. The second can be checked.
Ask for the written stage definitions before you look at a single opportunity. Then test them against reality: pull twenty opportunities currently sitting at the penultimate stage and check whether the evidence the definition requires is actually attached. In businesses where the pipeline is a genuine management instrument you will find it. Where you do not, the stages are labels and the coverage ratio is arithmetic performed on a feeling.
This is not a reason to dismiss the pipeline. It is a reason to weight what follows differently, and it is itself one of the more actionable findings you can hand a new owner.
Six tells for a manufactured forecast
None of these is conclusive alone. Two or three together are a pattern, and the pattern is what matters.
- Close dates cluster on period ends. Real buying decisions do not respect your quarter. Heavy clustering on the last day of a month or quarter means dates are being set by the seller’s calendar, not the customer’s.
- Close dates roll rather than slip. Look at how often the close date on a single opportunity has been pushed. Three or more pushes without a value change usually means the deal died some time ago and nobody has said so.
- Value never moves. An opportunity created at exactly £100,000 that closes at exactly £100,000 has not been negotiated. Round numbers that survive to close mean the value was a placeholder that nobody revisited.
- Stage skipping. Opportunities that jump from early stage to commit without passing through the middle are being reclassified to make a number, not progressed.
- Creation spikes after a bad quarter. A surge in new opportunities immediately following a miss is often pipeline being generated to reassure rather than demand being generated to sell.
- The commit is always right and the pipeline is always wrong. If the committed number lands accurately every quarter while everything behind it converts poorly, the business is not forecasting. It is only counting the deals it had already won.
The four numbers worth calculating yourself
Derive these from the raw export rather than accepting them from a deck. Each takes minutes and each answers a question the summary will not.
| Measure | How to calculate | What it tells you |
|---|---|---|
| Win rate by entry stage | Won divided by all opportunities that ever reached a given stage, not by those currently there | Whether qualification is real. A high late-stage win rate with a poor early-stage one means the pipeline is filtered by attrition rather than judgement. |
| Cycle time distribution | Created to closed-won, plotted as a distribution rather than an average | Whether there is one repeatable motion or several. A bimodal distribution usually means two different businesses sharing a sales team. |
| Forecast accuracy | Each of the last eight commit submissions against actual | Whether the business can predict itself. Consistent bias in either direction is more workable than random error. |
| Concentration of the number | Share of last year’s revenue closed by the top two sellers | How much of the plan depends on individuals who may not stay through a transaction. |
The last one is routinely the most consequential and the least often asked. A business where two people close most of the revenue is not necessarily a bad business, but it is a different asset from the one the org chart implies, and the retention package should reflect that.
Why coverage ratios mislead
Three times coverage is the number everyone quotes and almost nobody interrogates. The ratio is only meaningful if the pipeline in the numerator has the same conversion characteristics as the pipeline that produced the historic rate — and under a sale process it usually does not.
Two distortions are common. Pipeline gets padded ahead of a transaction, which lifts coverage while lowering quality. And pipeline ages: an opportunity that has sat in stage three for nine months counts fully toward coverage while converting at a fraction of the rate of a fresh one.
The correction is straightforward. Recalculate coverage using only opportunities created in the last two typical sales cycles, and apply conversion rates derived from the same window. The gap between that figure and the headline is frequently the most useful single number in the whole exercise.
Take every open opportunity with a close date inside the next quarter. What proportion has had any recorded activity in the last thirty days? Below half, and the forecast is describing a pipeline nobody is working.
From findings to a plan
Pipeline findings are unusually actionable, which is why this is worth doing properly rather than treating it as a box to tick. Three of the most common conclusions and what each implies for the first hundred days:
- Stages are subjective. Rewrite them around buyer-verifiable evidence and enforce at review. Low cost, weeks not months, and it makes every subsequent number trustworthy.
- Qualification is weak. Sellers are spending time on deals that were never winnable. Fixing this raises win rate and cycle time simultaneously without adding headcount, which is the cheapest growth available in most mid-market businesses.
- The number depends on two people. This is a structural risk that no process change fixes quickly. It belongs in the retention plan and, arguably, in the price.
What ties them together is that none is visible from the financial statements, and all three change what a buyer should pay or plan to do.
Frequently asked questions
How much CRM history is needed for pipeline diligence?
Twenty-four to thirty-six months of opportunity-level data including closed-lost records, with stage history where the system captures it. Less than two years makes it difficult to separate seasonality from trend.
What is a healthy pipeline coverage ratio?
There is no universal figure, because it depends entirely on conversion rate and cycle length. A business converting one in three with a short cycle needs far less coverage than one converting one in ten. Recalculating coverage using recent opportunities and matching conversion rates is more informative than any benchmark.
Can pipeline be assessed if the CRM is poorly maintained?
Yes, and the poor maintenance is itself a finding. Where record-keeping will not support analysis, the conclusion is that forecasting is judgement-based and the improvement plan has to begin with basic instrumentation.
Who should run pipeline diligence?
Someone who has carried a number and managed a pipeline. The analysis is not technically difficult, but interpreting seller behaviour from CRM records requires knowing how sellers behave.
Ready to aim higher?
If you need a view on whether a forecast is credible before it underwrites a price, we assess pipeline and forecasting alongside the rest of the commercial engine and deliver a prioritised plan.