A forecast built on what happened, not on how it feels

Every month the pipeline review produces a number. Every month the actual comes in below it. Nobody is lying and nothing is broken: the number was built by asking people how confident they felt, and confidence is wrong in the same direction every time.

It is not a character flaw in salespeople. A deal you have worked for three months feels closer than a deal you have worked for one, and it usually is. What it is not is 80% likely to close, which is what gets written next to it.

There is a version of this that works, and the only thing it requires is a count of what happened to your last two years of deals.

Take every deal from the last eighteen to twenty-four months. For each stage in your pipeline, count how many deals ever reached it, and how many of those eventually closed won. Divide one by the other and you have the conversion rate for that stage.

That number is not a matter of opinion. If deals that reached Proposal sent have closed 28% of the time, then a deal at Proposal sent is worth 28% of its value in the forecast, whatever anyone believes about this particular one.

It will be wrong about every individual deal, which is the point. Across twenty deals it is roughly right, and a forecast is a statement about twenty deals rather than about any one of them.

That last point confuses people the first time. The denominator falls as you move down the funnel while the numerator stays the same, which is exactly why the conversion rate rises.

Eighteen open deals worth ₹3.91 crore at face value. Weighted by stage, they come to ₹94.9 lakh, which is 24.3% of the pipeline.

That gap is the whole argument. A sales team reporting a ₹3.91 crore pipeline and a finance team planning on ₹94.9 lakh of it are both right, and the second number is the one you can hire against.

A low weighted share is not bad news. It usually means most of the pipeline is early, which is what a healthy pipeline looks like. A high share means most of your deals are nearly closed and there is nothing behind them, which is a problem two quarters out.

Read the variance column down rather than across. One month at minus 21% is a deal that slipped. Four months averaging minus 5% is a forecast you can plan on, and the improvement from April to July is what happens when the close dates stop being optimistic.

The Forecast sheet has two rows feeding the total, and keeping them apart matters more than it looks.

Committed and recurring is revenue you already have: subscriptions in force, contracted minimums, retainers that renew. It is not a probability and it must never be weighted. Applying a stage weight to your existing customer base is a mistake that quietly destroys the forecast.

Weighted new business is the pipeline. It is uncertain by nature and that is what the weights are for.

Weighting fixes whether a deal closes. It does nothing about when, and the when is what decides which month the revenue lands in.

• ·Deals slip later far more often than they pull forward. Almost nothing closes early.
• ·Procurement, legal review and purchase order processes are the usual cause, and they are largely outside your control.
• ·Approval cycles at larger customers are longer than at smaller ones, so moving up market lengthens the slip as well as the cycle.
• ·A deal that slips twice usually slips a third time, and the pattern is visible in the pipeline long before anyone admits it.

The practical correction: work out how much later your closed deals landed than first forecast, on average, and add that to every expected close date before you enter it. If the average was three weeks, add three weeks. It feels pessimistic and it is simply what happened.

The weights are only as good as the stages they are attached to, and stage definitions are where most pipelines quietly break.

The failure is always the same. A stage is defined by what your team did rather than by what the customer did. Proposal sent is an action you took; it says nothing about whether anyone read it. Demo completed means someone attended a call. Neither tells you anything about the deal.

The fix is to define every stage by an observable customer behaviour, and to write down the evidence required to enter it.

• ·Qualified: the customer has confirmed a budget, a need and who signs. Not that your salesperson believes those things.
• ·Discovery done: requirements are written down and the customer has agreed they are right.
• ·Proposal sent: the commercial proposal is with the customer and they have acknowledged receiving it.
• ·Verbal yes: a named person has said yes, subject to paperwork, and you could name them in a meeting.
• ·In procurement: legal review or a purchase order process has actually started, with someone in their procurement team involved.

Written this way, a stage is checkable by someone who was not on the call. That is what stops deals sitting two stages further along than they belong, which is the thing that makes weights meaningless however carefully they were calculated.

1. 01Update the stage of every open deal, and add anything new. Move deals backwards when they deserve it, which is the discipline most pipelines lack.
2. 02Remove deals that are dead. A pipeline with nine-month-old deals sitting at Proposal sent is not a pipeline, and they distort the weighting badly.
3. 03Update expected close months honestly, including the slips.
4. 04Enter last month's actual on the Forecast sheet.
5. 05Look at the accuracy block before the deal review, not after. It frames the conversation.
6. 06Rebuild the stage conversion rates every six months, because stage definitions drift and a move up market changes them.

Fifteen minutes a month once the pipeline list exists, and the accuracy block after two quarters is worth more than any individual deal discussion.

• ·It is not a CRM, and it is not a substitute for one beyond about forty open deals. It sits on top of whatever your CRM exports.
• ·It forecasts bookings, not cash. When the money arrives depends on payment terms and collections, which the 13-week cash flow template handles.
• ·It weights annual contract value. One-off implementation fees should be forecast separately if they are material, because they convert differently.
• ·It assumes your stages mean something consistent. If Proposal sent means four different things to four salespeople, fix the definitions before trusting the weights.
• ·It needs enough history to count. With twenty closed deals the rates are noisy, and the honest response is to use them anyway and replace them as the data builds.
• ·It says nothing about whether the pipeline is big enough. That is a coverage question: weighted pipeline against target, which is worth a glance every month.

What it does is replace a number nobody quite believes with one that has a method behind it, and then keep score. For most companies the second part is what changes behaviour.