Most forecasts stop at the signature. Revenue does not. A usable forecast has to account for the buyer, the contract, delivery capacity, and the customer reaching first value. Until those four clocks agree, the number is not an operating plan. It is a sales opinion with a percentage attached.
From the archive
A forecast meeting usually begins with one number and four different definitions. Sales hears the value likely to sign. Finance hears revenue it can plan against. Operations hears work it may soon have to deliver. Customer Success hears customers it will be expected to retain. Everyone nods at the same figure, but they are not agreeing. They are translating it into four different futures, each with its own assumptions, dependencies, and timing. Then the meeting ends and the number is entered into the plan as if the company settled something. It did not. It only gave four incompatible meanings the same label.
Most revenue forecasts are built around a single event: the signature. The opportunity has a value, a probability, and an expected close date. If the rep is confident and the stage looks advanced, the number moves into commit. That describes the sales team’s intention. It says almost nothing about what the company can safely expect to keep.
A signed agreement still has to survive procurement, handover, onboarding, delivery, adoption, and the first moment the customer sees the value they were promised. Capacity has to exist. Information has to move. Someone has to own the next event. If any of those conditions fail, the company can book the deal and still lose the revenue economics underneath it.
This is the part the normal forecast cannot see. It treats the signature as proof that the system worked, even when the signature merely transfers the risk from Sales to another department. The CRM reports a win. Operations receives an exception. Customer Success receives an expectation nobody documented. Finance receives a number with hidden work attached. Each department inherits a different version of the same deal, and the forecast remains green because it stopped measuring before the consequences began.
A useful forecast has to answer more than whether a deal might close. It has to show whether the sequence around that deal can actually happen. I use four clocks to make that sequence visible. They can run at different speeds, but they have to agree on the order of events. When one clock is missing, the forecast borrows certainty from work nobody has checked.
The first is the buyer clock. It tracks what has happened outside your company: the decision process, the stakeholders, the unresolved risk, and the next buyer-owned event. This is where evidence replaces the rep’s impression. A deal is not progressing because the conversation felt good. It is progressing because the buyer did something that changed the state of the decision.
The second is the commercial clock. It covers the path through legal, procurement, security, signature, billing terms, and the conditions that turn an agreement into something the business can recognise. A verbal yes can be genuine and still be months away from a usable contract. Treating those as the same event is how a forecast pulls revenue forward before the buying process is ready to release it.
The third is the delivery clock. It asks whether the company can fulfil what Sales is promising, when it is promising it. Give the team a sellable service catalogue, visible capacity, and a clear rule for exceptions, and Sales can keep selling without creating promises the rest of the company cannot support. Without that structure, the forecast counts demand while ignoring the bottleneck that demand is about to hit.
The fourth is the value clock. It follows the customer from handover to first value. When does the customer experience the thing that made the purchase worth making? Who owns that moment? What information has to survive the handoff? Revenue that signs but never reaches value is not predictable revenue. It is churn being reported early as growth.
A forecast becomes operational when the company traces the deal through all four clocks instead of stopping at the sales stage.
What has the buyer actually done, and what buyer-owned event happens next? Evidence lives in action, not sentiment.
What still stands between intent and a usable agreement: procurement, legal, security, signature, or payment terms?
Can the company deliver the promise on the proposed date without creating an exception that damages margin or service?
Who owns the handover, and when will the customer reach the first measurable result they bought the service to achieve?
CRM probability looks precise because it is expressed as a percentage. In practice, that percentage often compresses several unresolved questions into one convenient field. Is the buyer committed? Is the paperwork moving? Is delivery available? Is there a real path to first value? A rep can be right about the first question and the company can still be wrong about the revenue.
That is why forecast debates become personal. The manager asks why the deal is at 80 percent. The rep defends the relationship. Finance discounts the number. Operations is not in the room. Nobody is arguing from the same object, so confidence becomes a proxy for evidence and seniority becomes the tie-breaker.
Adding more stages or demanding longer CRM notes will not repair the operating sequence. Make each clock visible and require a named event where responsibility changes hands. The forecast then has something stronger than opinion: a chain the company can inspect.
I saw this clearly in a mid-sized enterprise software company after six stagnant months. The CEO and financial controller believed the sales team was failing to close enough business and were considering replacing the reps. We mapped the customer acquisition flow alongside the compensation structure on a whiteboard. The sales team was being rewarded for short-term upfront contract value. The onboarding function was already constrained. Forty percent of new clients were leaving within ninety days.
The sales team had not ignored the company’s system. It had responded to it. The incentive said maximise the contract. The forecast said count the signature. Neither instrument carried responsibility through onboarding and first value, so growth on one clock created failure on another. Replacing the reps would have preserved the conflict and changed only the names of the people caught inside it.
That room went quiet because the problem moved. What looked like weak closing performance became a design contradiction between compensation, capacity, and retention. The forecast had failed to predict the outcome. Worse, it had directed attention away from the mechanism causing it.
A forecast can be numerically tidy and operationally impossible.
AI can summarise calls, identify dates, flag missing stakeholders, and prepare a forecast review before the team enters the room. That is useful. It can reduce the administrative cost of finding evidence across a messy commercial system. It can also produce a polished explanation from fragmented inputs and make weak operating discipline look complete.
McKinsey’s 2026 B2B Pulse places AI, data quality, connected workflows, and disciplined commercial execution inside the same growth conversation. That connection matters. AI does not remove the need for a coherent revenue process. It raises the value of having one, because automation can move reliable evidence faster and can move unreliable assumptions faster too.
The sensible order is evidence first, automation second. Define what progress looks like on each clock. Decide which events matter. Make ownership visible. Then use AI to retrieve, compare, and challenge that evidence. If the underlying process cannot distinguish a buyer action from a rep’s confidence, no model can repair the definition by summarising it more fluently.
You do not need to rebuild the CRM before you can test this. Take the five deals carrying the most weight in the current forecast and put Sales, Finance, Operations, and Customer Success around the same table. For each deal, ask what the next observable event is on each clock. Do not ask whether the deal feels good. Ask what has happened, what must happen next, who owns it, and what capacity or dependency could stop it.
The exercise makes the number useful. One deal can remain highly likely to sign while revealing that delivery cannot begin for six weeks. Another can have capacity reserved but no buyer-owned event. A third can be commercially complete while the handover information is too thin for Customer Success to reach first value. Those risks require different decisions. One probability percentage cannot tell the company which work to do.
Write the gaps beside the deals. Assign the next owner. Change the forecast only where the evidence justifies it. Then watch which gaps repeat. Repetition is the system showing itself. If legal appears in every late deal, that is a commercial path problem. If custom promises repeatedly overwhelm delivery, that is an offer and governance problem. If signed clients arrive without context, that is a handover problem. The forecast review becomes valuable when it exposes infrastructure the company can build once instead of firefighting deal by deal.
The purpose of a forecast is not to make leadership feel certain. It is to help the company prepare. Sales needs to know what evidence will advance the deal. Finance needs to know when the economics become real. Operations needs to see incoming demand before it becomes an emergency. Customer Success needs enough context to turn a promise into value. A useful forecast coordinates those decisions before the consequences arrive.
This is the problem I built the 45-minute working keynote Why Your Forecast Is Fiction to open up with leadership teams. They already have formulas. The number becomes more honest when the company can see the system underneath it together.
Start with the four clocks. If the buyer clock is real but the delivery clock is impossible, the forecast has found the work. If the commercial clock is complete but the value clock has no owner, it has found the work again. A forecast must predict what can happen and tell the company what must be ready next.
Twelve years across three continents rebuilding the infrastructure B2B companies use to turn good people into predictable revenue. Now working from Sweden, with a smaller calendar and a tighter focus. Thanks for reading, new essays land here most weeks.
One essay a week on the work underneath B2B revenue. No pitch.
A 45-minute call. I tell you where the leaks are, whether or not we work together.
Let's talk