Why your sales forecast is always wrong, and how to fix it

It's week eleven. Your forecast said $4 million. You're heading for $2.8 million. Again.

You're not alone. Gartner found that fewer than half of sales leaders have high confidence in their own forecast. Think about that. Most of the people calling the number don't believe it.

I've sat beside VPs of Sales and CROs in boardrooms across North America and Asia Pacific as they explained a missed quarter. The market. The rate cycle. The deal that slipped. The rep who left in July. I've heard every reason. Almost none of them is the real one.

This guide is the real one. It's everything I've learned since 2000 about why forecasts miss, and the simple, unglamorous habits that make them accurate.

The forecast isn't the problem. The inputs are.

A forecast is just arithmetic on top of opinions. Each rep looks at each deal and makes a call. Commit. Best case. Pipeline. The spreadsheet adds it up.

The trouble is the question behind every call: "How do you feel about this deal?"

Feelings are optimistic. A buyer who loved the demo feels like a buyer who will sign. A deal that has been "almost there" for six weeks still feels close. And nobody wants to be the rep who kills a deal in front of the team.

So hope goes in. Your forecast comes out. It looks precise. It isn't.

Revenue is an outcome. By the time you see it, it's history. You can report it, but you can't coach it. What you can manage are the activities underneath every deal. Those activities are either happening or they're not. No opinion required.

What is pipeline fiction?

Every deal in your pipeline is one of two things.

• Actionable. There is a date and a time for the next meeting, confirmed in both the rep's calendar and the buyer's.
• Falldown. There isn't. The last conversation ended without a confirmed next step. The deal may still carry a stage, a probability and a close date. But it's stalled, whether the CRM shows it or not.

A Falldown deal is not forecastable. It can't be part of a number you'd defend in front of your board. Yet in the pipelines I review, 50 to 100 percent of deals are Falldown. That gap between what the CRM says and what is actually booked is pipeline fiction.

The four checks that separate real deals from fiction

Run your top ten deals through these four questions. Be strict. "Sort of" counts as no.

1. Is the next meeting booked? A real date and time, in the rep's calendar and the buyer's. No meeting means the deal is drifting, no matter how warm it feels.
2. When did the rep last speak with the buyer? Spoke, not emailed or messaged. If the answer is "a couple of weeks ago," the deal is cooling. Dark deals don't close on the date in the CRM.
3. Has the close date held? One moved date is common. Two is a pattern. Three is fiction. A close date that keeps moving is a deal telling you the truth. Listen to it.
4. Does the manager know the real status, without opening the CRM? If your manager can't tell you what happens next and when, neither can your forecast.

Most sales leaders who run this exercise find one or two "commit" deals that fail the first check alone. Many find more. Every one of those deals is sitting inside your number right now.

The five reasons sales forecasts miss

When I trace a missed quarter back to its source, I find the same five causes. None of them is the market.

1. The pipeline was never pressure tested

The rep entered the deal. The manager reviewed it. It moved forward without a challenge. Your forecast became a summary of unverified stories.

The question that reveals this in thirty seconds: "For your top five committed deals, what did the buyer say, specifically, that tells you this will close this quarter?"

The fix: weekly pipeline reviews built on evidence questions, not status questions. Not "where is it?" but "what proof do we have?"

2. Stage names mean different things to different people

Ask ten reps what "Proposal Sent" means and you'll get ten answers. For one, it means a proposal was delivered and discussed. For another, it means a PDF was emailed into silence. Both sit in the same stage with the same probability.

The fix: stage definitions written in buyer actions, not seller actions. A deal moves forward when the buyer does something, not when the rep does.

3. Deals have no confirmed next step

This is the Falldown, and it is the most common cause of forecast failure I've seen in more than two decades of pipeline work. A call ends without a next date and time. The deal sits in the pipeline, quietly dying. Nobody knows yet.

The fix: make the confirmed next step non negotiable. No buyer agreed date and time, no place in the commit.

4. The forecast call rewards confidence, not accuracy

Sit in enough forecast calls and you'll notice it. The managers who speak with the most confidence get the least pushback. Over time, everyone learns to perform certainty instead of reporting reality.

The fix: clear definitions for commit, best case and pipeline. And a culture that celebrates the manager who called 70 percent and delivered, over the one who promised 100 percent and missed.

5. Nobody below the CRO owns the miss

When the quarter misses, the CRO answers for it. The managers whose pipelines held deals with no next step rarely do. So the same miss happens again next quarter, explained a new way.

The fix: track forecast accuracy by manager. Not as punishment. As a coaching signal that shows exactly which habits to work on.

Why the fix runs through your managers

Most organisations treat a forecast problem as a rep problem. The pipeline is padded, so they train the reps. A new methodology, a two day workshop, a binder. The forecast improves for a month. Then the old patterns return.

Here's why. The person who decides what counts as real in your pipeline, every single week, is the manager. If the manager wasn't equipped to hold the new standard, the standard leaves when the trainer does.

Your managers can do two things no system can. They can ask the question that surfaces the truth about a deal. And they can make it safe for a rep to tell that truth.

When a manager reads CRM entries aloud, they're managing data entry. When a manager asks, "Is the next meeting booked, or is this Falldown?", they're managing the opportunity. Only one of those conversations produces a forecast you can trust.

What happened at IDC

When I began working with IDC across Asia Pacific, pipeline reviews were diplomatic. Managers reviewed what reps had entered. The numbers looked plausible. The forecasts were off.

We installed one standard across the manager layer: every live deal needed a confirmed next step, a recent conversation with the buyer and a close date the rep could justify. Managers held that standard in every review, in every market.

CRM adoption reached 100 percent, not because anyone mandated it, but because the standard made honest data the easiest path. Bookings grew 15 percent. And IDC kept the work running for years, because their managers owned it.

"Before Robert, our pipeline reviews were diplomatic. After, they were accurate."
Cardi Prinzi, VP Sales, Pacnet (now Telstra)

Why better software won't save you

AI forecasting tools get sharper every quarter. But they all read the same thing: the data your reps enter. Feed an algorithm hopeful data, and you get a confident wrong answer, faster.

The fix happens upstream, in the Monday pipeline review, where managers decide what counts as real. Get that right, and every tool you own gets smarter overnight.

What accurate forecasting looks like

An accurate forecast isn't magic or luck. It's five habits, running every week:

1. Evidence based pipeline reviews. Every committed deal is backed by something the buyer did or said.
2. Buyer action stage definitions. Everyone agrees what each stage means.
3. A confirmed next step for every committed deal. No date and time, no commit.
4. Clear forecast categories. Commit, best case and pipeline mean the same thing to everyone.
5. Manager level forecast accuracy. Tracked, visible and used for coaching.

Teams that run these habits still face moving markets. The difference is that they know, precisely, what's real in their pipeline. So when the market shifts, they respond from truth instead of hope.

What to do before your next forecast call

You don't need a new system to start. You need one hour on Monday.

1. Pull your team's top ten committed deals.
2. Run each one through the four checks above.
3. Move every deal without a booked next meeting out of commit.
4. Ask each manager to name the real status of their top three deals, without opening the CRM.
5. Compare the new number with the one you were about to present.

The gap between those two numbers is your pipeline fiction. Now you can see it. And what you can see, you can fix.

Frequently asked questions