Your Sales Pipeline Is Not a Forecast.

I want to know what the business can collect, when it can collect it, and what it has to spend to get there.

You show me a dashboard with $4.2 million in pipeline and expect me to feel good about the business.

I have questions.

What does that number represent? Who entered it? How old are the opportunities? What has the customer actually agreed to? When could the work begin? When would the company get paid?

And what would be left after acquiring the customer and delivering the service?

If nobody can answer those questions, I don't have $4.2 million of confidence. I have a very expensive fucking conversation about a chart.

I follow the money through the operation

My starting point is the business. I want to understand how it earns money, where it loses money, what resources it has, and which decisions constrain the result.

The sales pipeline belongs in that examination. It doesn't get a pass because the software looks professional.

An inquiry, a qualified opportunity, a signed agreement, work performed, revenue recognized, and collected cash are different events. They may happen on different dates, at different values, with different risks between them.

For a high-ticket local service business, those differences can determine whether growth is affordable.

If you have to pay for advertising, intake, specialized labor, materials, or outside help before the customer pays, more signed business can increase the amount of capital tied up in the operation.

I want to see that funding requirement before management commits the money it expects the pipeline to produce.

“We should close it” doesn't tell me how you're financing delivery if you do.

A customer who can afford it still has to want it

Somebody asks for a price. Sales enters an opportunity.

Fine. What else do we know?

Can they authorize the purchase? Does the service fit the problem? Are they ready to act? Do they prefer your offer enough to choose it? Is the timing real?

Being able to afford something doesn't establish a desire to buy it. Enthusiasm doesn't establish authority. Receiving a proposal doesn't establish agreement.

You can keep early conversations in the system. They belong somewhere useful. Just don't grant them commercial meaning they haven't earned.

If “qualified” means one thing to marketing, another to intake, and whatever the salesperson needs it to mean on Friday, the field isn't a shared measurement.

It's an argument waiting for a meeting.

Tell me what changed in the buyer's world

“Proposal sent” tells me your employee sent a proposal. It doesn't tell me the buyer has moved closer to a decision.

I want the evidence supporting the stage. A confirmed requirement. An identified decision-maker. An agreed next step. A documented purchasing timetable. Whatever is relevant to how this particular business sells.

The criteria need to be explicit enough that another competent person can review the record and understand why it's there.

So do the conditions for moving backward, pausing, or closing the opportunity as lost.

If nobody wants to remove dead opportunities because management celebrates the total, you've built an incentive to preserve the illusion. Fix the incentive before you complain that the CRM is full of shit.

Your percentage needs a reason to exist

Suppose ten hypothetical opportunities are worth $100,000 each. Five sit at a stage assigned a 50 percent probability. Five sit at a stage assigned 75 percent.

The weighted total is $625,000: $250,000 plus $375,000.

That is multiplication. We haven't established that it is a useful forecast.

Now suppose relevant historical records show close rates of 20 percent and 40 percent for those two groups. Applying those rates to the same stated values produces $300,000: $100,000 plus $200,000.

The $325,000 difference came from the assumptions. Nothing about the customer changed while we did the arithmetic.

Neither amount is guaranteed. These invented numbers illustrate why I want to examine the probabilities before management plans around the result.

Were the historical records complete? Did lost deals get recorded? Are we comparing similar services, sources, sizes, and buying conditions? Do those rates mean eventual closure or closure inside this forecast period?

If the business has a handful of irregular, high-value deals, percentages can create an appearance of precision the evidence doesn't support. Separate scenarios and explicit judgment may be more useful.

And if the offer, market, pricing, or capacity has changed, old performance may describe a business you're no longer operating.

Use the history. Challenge whether it applies.

I want four dates before I trust one number

When might the customer commit? When could the business deliver? When could it bill? When might it collect?

There may be more dates that matter. Those four are enough to start exposing the fantasy.

A contract signed in March might require spending in April and produce cash in June. A customer might be ready now while your delivery schedule is full until September. A deposit might fund part of the work while leaving the company exposed on the rest. The applicable revenue-recognition treatment may describe yet another date and amount. SBA planning guidance separately calls for forecast income statements, balance sheets, cash-flow statements, and capital-expenditure budgets, while the SEC explains that a cash-flow statement tracks actual cash inflows and outflows. [1][2]

Show me the timing and terms. Then show me the downside if the deal slips, costs rise, or payment arrives late.

If the owner tells me there's a million dollars available to fund growth, I want to establish that the capital is actually available after existing obligations. A forecast doesn't make that money exist sooner.

You cannot spend a close probability on payroll.

The departments need to reconcile the same journey

Marketing records an inquiry in January. Sales records a contract in February. Operations performs the work in April. Finance collects in May.

Those reports can all be accurate while telling different stories about the same customer.

I want to connect the events rather than force them into one misleading number. Keep the source, qualification evidence, service, value definition, stage history, expected dates, actual dates, delivery costs, and collection outcome understandable to the people responsible for them.

Assign owners for the definitions and the handoffs. Decide who can change a value and why. Preserve enough history to reconstruct the forecast management actually used.

And don't give everyone unrestricted access to sensitive customer or financial records just to make reporting easier. Share what the role requires and protect the rest.

The point is a coherent operating picture. A wider distribution list won't create one.

AI doesn't get to bless the forecast

AI may help inspect records, summarize changes, or surface deals that need attention. Whether it helps depends on the information and the tool.

It cannot establish that a buyer has a budget by making the salesperson's notes sound more convincing.

It cannot repair a missing definition by predicting a precise number from inconsistent records.

I want the basis for the estimate, the limits of the information, and a person who owns the judgment. Sensitive deal information belongs only in systems approved for it.

The machine can help examine the claim. Management still has to decide how much capital to risk on it.

Keep the forecast you got wrong

Save dated versions. Compare them with outcomes. Track what closed, what slipped, what changed value, and what actually collected.

Separate a reasonable miss from a preventable fiction. A customer delaying approval is different from a salesperson leaving the same imaginary close date in the system for six months.

Use the difference to improve the process. If every miss becomes a punishment, you may teach people to hide uncertainty. If every miss disappears into a revised dashboard, you teach the company nothing.

A well-defined pipeline can support a useful forecast. I'm not objecting to pipelines or pretending uncertainty can be eliminated.

I'm asking you to make the uncertainty visible enough that we can decide what the business can afford to do.

Tell me what might sell. Tell me the evidence. Tell me when the cash could arrive, what delivery will consume, and what happens if you're wrong.

Then we can discuss the fucking forecast.

Continue Reading on anuhlest

• Your Business Made a Million Dollars. Where the Fuck Is It? — why revenue, profit, working capital, distributions, and available cash answer different questions.
• How Much Should I Spend on Marketing? That Depends on What the Fuck the Next Customer Is Worth. — why customer economics and incremental acquisition cost belong in the spending decision.
• You Got the Customer. Now Keep Them. — why the first sale does not establish retention, repeat contribution, or lifetime value.
• Your Prices Are Based on What? — why price, scope, capacity, contribution, and payment timing have to support the same transaction.