Sales forecasting for B2B tech companies: a practical guide

I learned this the hard way

Early in my career, I fell flat on my own sales forecast, more than once. I called numbers I believed in, and the quarter fell apart anyway. It took me years, and a lot of anxious quarter-ends, to understand that the problem wasn’t effort or optimism. It was that my forecast had nothing solid underneath it.

Since then, I’ve built a forecasting system and used it with tech companies as a consultant. Almost without exception, every client I’ve worked with had this as their core problem: quarters that start full of promise and collapse in the last few weeks.

What a collapsing forecast looks like

One client, a cybersecurity solutions company in Chennai with around 30–35 salespeople, is a good example.

Every quarter started the same way: high hopes, plenty of optimism and a funnel full of what looked like great opportunities. By the middle of the quarter, about 60% of the “qualified” opportunities had disappeared. By week 9 or 10, about 70% were gone. Every quarter fell apart, and nobody could say exactly why.

The reason, when we looked closely, was simple. The funnel was full of hope, not qualified opportunities. When we filtered the team’s “qualified” funnel through proper qualification rules, only 30–35 of every 100 opportunities survived. The rest had never been real.

Pipeline vs forecast: why the difference matters

Your pipeline is every opportunity your team is working on. Your forecast is your judgement of which of them will close this period, and for how much.

Most forecasting problems begin when the two get confused. A pipeline can look large and healthy while the forecast inside it is fragile. In B2B tech this is especially true, because:

• Decisions involve several people. A deal can look hot with the user and stuck with finance or the person who signs.
• Many companies sell through partners. Most of my clients are system integrators and value-added resellers. Their forecast depends on customers, OEMs and their own sales teams all moving together.
• Quarter-end pressure works both ways. Customers know you need to close, and deals that “will definitely sign” often move when the customer realises they can wait.

The three KPIs: Q, N and C

In B2B Tech SalesOS, every seller tracks just three numbers. I keep it deliberately simple. You don’t need 13 or 14 funnel stages to forecast well; between “lead” and “closed”, what matters most is how much of your funnel is genuinely qualified.

KPI	What it measures	Coverage target (X = your sales target)
Q: Qualified funnel	Opportunities that pass all five qualification factors (MANT-C, below)	At least 4–5× X
N: Negotiation funnel	Opportunities at the negotiation stage	At least 2× X
C: Commit	Deals you’re committing to close this period	At least 1.25× X

If your qualified funnel is below 4× your target at the start of the quarter, you already know the quarter is at risk, weeks before it shows up in the commit number. That early warning is the whole point.

Qualification is what makes Q mean something

The Q number is only useful if “qualified” means the same thing for everyone. We qualify every opportunity on five factors, MANT-C: Money, Authority, Need, Timing and Competition (including the option to do nothing). An opportunity that hasn’t cleared all five doesn’t count in Q, however exciting it feels.

This is also why forecasting and discovery are linked. If discovery stays on the surface, qualification has nothing to stand on. Our guide to discovery questions covers that side.

Forecast in three scenarios, not one number

Every forecast we build has three scenarios:

• Most probable: what you expect to close, based on the evidence.
• Worst case: what closes if the risky deals slip.
• Best case: what closes if the deals that are close but not certain come through.

Three scenarios force an honest conversation about risk. If the worst case and the most probable are far apart, the forecast depends on too few deals, and leadership should know that before planning cash or hiring around it.

Make the number your own

The biggest change isn’t a spreadsheet. It’s a mindset.

In most teams I meet, calling out a forecast number is casual. Someone says a figure in a meeting, and nobody treats it as a promise. The shift we work on is from that casual habit to owning the number: treating a commit as something the company’s business, profitability and future depend on. When people feel that ownership, the way they qualify and the deals they focus on change.

That ownership needs a rhythm to support it. With the Chennai team we introduced:

• Reviews twice a week, short and in the same format every time.
• A mid-quarter review, to reset early rather than late.
• A quarter-end review, to learn from what happened.

And we built MANT-C qualification and SIAM discovery into the sales process itself, not as a training exercise. That, more than anything, brought seriousness to the forecast.

What changed for the Chennai team

The first quarter after implementation, they still missed their number. But their forecast was 85–90% accurate. For the first time, the number they called was the number they got.

That gave the team enormous confidence: a reliable forecast was possible. Over the next two to three quarters, as they focused on well-qualified opportunities instead of spreading themselves across weak ones:

• Individual reps’ achievement moved from around 30–40% to 85–90%.
• The team’s overall achievement rose with it.
• The company became less dependent on a few “hero” salespeople.

The behaviour change mattered as much as the numbers. Under a clear, strict review standard, the people who took the pressure seriously shaped up quickly. The ones who had been hiding behind reasons and excuses became visible. Mediocre performance can’t survive a review process that asks for evidence every week.

The most common mistake founders make in forecast meetings

The pattern I see most often: founders put tremendous pressure on commit without asking what the commit is based on.

When the pressure is high and nobody asks why a deal is in commit, people shoot from the hip. They call a number to get through the meeting, and the forecast becomes a negotiation rather than an assessment.

A related trap: founders who want to control every opportunity and insist on being in every final client meeting. It undermines their sales team, and it often undermines the founder too, because the customer sees who really decides. There’s a right time for a founder to engage with a customer, which I cover in the guide to building a sales system in a startup.

Better questions for a founder to ask in a forecast meeting:

1. Which of these commit deals have cleared all five MANT-C factors?
2. Is our qualified funnel at least 4× this quarter’s target? If not, what’s the plan?
3. What’s the gap between our most probable and worst-case scenarios, and which deals create it?
4. Which deals moved since the last review, and what customer evidence caused it?

How accurate should a sales forecast be?

You can’t be 100% accurate. Nobody is. But a forecast built on data and KPIs shouldn’t be more than about 10–12% away from the number you call.

Whether you achieve your target is a separate question. A team can miss its target with an accurate forecast, as the Chennai team did in their first quarter, and that’s still progress. What isn’t acceptable is a number that’s far from reality because it was based on feeling.

In one sentence: forecasting is intuition based on data. Experience still matters, but the commit you call has to rest on concrete facts and KPIs, not gut feel.

If you’d like a quick read on where your team stands, the Revenue Predictability Check takes two minutes and shows your weakest areas.

Common forecasting mistakes

• Counting unqualified opportunities in the funnel, so Q looks healthy when it isn’t.
• Pressuring for commit without asking what it’s based on.
• Reviewing the forecast only at quarter end, when there’s no time left to act.
• Treating the forecast as the target. The forecast is your honest view of what will close; the target is what you need.
• Depending on a few heroes instead of a system everyone runs.

Where forecasting fits in B2B Tech SalesOS

Forecasting is step 5 of the seven steps in B2B Tech SalesOS. It sits near the end on purpose: a forecast is only as good as the qualification behind it, and qualification is only as good as the discovery conversation where it happens. Teams that fix forecasting on its own see the number improve for a quarter and slip back. Teams that fix qualification and discovery as well see it hold.

If you’d like help putting this in place with your team, see how we approach sales forecasting consulting.

Frequently asked questions

What are the QNC KPIs?

They’re the three forecast KPIs in B2B Tech SalesOS: the Qualified funnel (at least 4–5× your target), the Negotiation funnel (at least 2×) and Commit (at least 1.25×).

How often should a sales team review its forecast?

We recommend short reviews twice a week, plus a mid-quarter and a quarter-end review. The frequency matters less than the consistency: same format, every time, based on evidence.

Why did the qualified funnel shrink so much after qualification?

Because most “qualified” opportunities hadn’t been checked against money, authority, need, timing and competition. In one client team, only 30–35 out of 100 survived. Losing the other 65–70 on paper is far better than losing them in week 10.

Can you have an accurate forecast and still miss your target?

Yes, and it’s an important first step. An accurate forecast tells you the truth early enough to act on it.