Forecast accuracy in B2B sales has been stuck in the same band for years. Median accuracy across surveyed organisations sits between 70 and 79 percent, according to Forecastio’s compilation of Gartner and CSO Insights data. Fewer than half of sales leaders have high confidence in the forecast their own team produces. The same research shows that organisations embedding forecast coaching into their management process improve accuracy by up to 15 percentage points — but most companies do not invest in the coaching layer, because it is invisible compared to the call.

The forecast call itself is where the problem becomes most visible. The standard call asks one question, in seven slightly different ways: “What are you calling?” The seller answers with a category. The manager asks why. The seller offers reasons. The category is recorded. The call moves on.

This is a negotiation about confidence, not an inspection of evidence. It produces a forecast that reflects the seller’s emotional state, not the buyer’s actual behaviour.

“Most forecast calls inspect the seller’s feelings about the deal. The forecast that holds inspects the buyer’s behaviour. Those are different conversations.” — Mark Southgate

The fix is not a new methodology. It is seven specific questions, asked of each deal in commit and best case, that replace seller confidence with buyer evidence.

Question 1: What does the buyer believe is supposed to happen, and when?

Not “what is the close date?” That is the seller’s date. The useful question is: what does the buyer believe is going to happen, and when do they believe it is going to happen?

The difference is the entire game. A buyer who, when asked, would say “we’re planning to sign by month-end” is at one place. A buyer who would say “we’re still evaluating, hoping to decide in the next quarter” is at another. The seller may have recorded both as a close date this month. The forecast cannot tell the difference unless the question is asked.

If the rep cannot articulate what the buyer believes, in the buyer’s words, the deal does not belong in commit. It belongs in best case at the most generous interpretation.

Question 2: Who is the economic buyer, and when did they last meaningfully engage?

Every enterprise deal in commit should have a named economic buyer, identified by role and name, who has been in a substantive interaction in the last three weeks. Not “the rep emailed them.” A substantive interaction — a meeting, a call, a meaningful written exchange — initiated or returned by the buyer.

The deals that miss this criterion are the ones that slip. The economic buyer who has not engaged in 30+ days is not committed. The rep may have other strong relationships in the account. None of them substitute for the economic buyer’s active participation.

This is the single most predictive question on a forecast call. Make it the second question for every deal.

Question 3: What is the buyer’s process from now to close, and where are they in it?

The buyer has a process. It includes some subset of: security review, legal review, procurement engagement, technical validation, financial sign-off, board or committee approval. Each of these takes time. Each is largely outside the seller’s control.

A deal in commit should have a documented map of the buyer’s remaining process, with rough timelines for each step, ideally agreed with the buyer. Without that map, the close date is a guess.

The question for the forecast call: walk me through what has to happen between now and signature, from the buyer’s perspective, and how long each step takes. If the rep cannot answer in 60 seconds, the deal does not have a credible close date.

Question 4: Where is the dissent in the buying group?

Every committee deal has dissent. The committee that is publicly aligned almost certainly has private disagreement. The seller who says “everyone is on board” is operating on a model of the buying group that is wrong.

The useful question: who in the buying group has the strongest reservation, what is the reservation, and how are we addressing it?

A rep who can answer this has been doing real qualification. A rep who cannot answer it is running the deal blind on whatever the next committee meeting reveals.

The forecast category for a deal where the dissent is mapped and being addressed is different from the category for a deal where the dissent is unknown. Both can be in pipeline. Only the first deserves commit.

Question 5: What competitive pressure is on this deal, and has it changed?

Not “are we still in a competitive deal?” That is too coarse. The useful question is: which competitor is putting the strongest pressure on this deal, what dimension are they competing on, and has that changed in the last two weeks?

The reason this matters: late-stage competitive shifts are one of the highest predictors of slippage. A deal that was perceived as a single-vendor evaluation but where the buyer has revisited the incumbent in the last fortnight is a deal whose risk profile has changed. The forecast category should follow.

“The competitive landscape on a deal in week 12 is not the landscape from week 4. Most forecast calls are still using the week 4 picture. The buyer is not.” — Mark Southgate

Question 6: What is the commercial structure, and does the buyer agree to it?

This is the question that catches the largest single category of late-quarter surprises. The seller has a number in mind. Procurement, when they engage, has a different number in mind. Legal has redlines that no one warned the deal team about. The deal that the forecast had at $X closes at $0.7X, or slips while the gap is negotiated.

The commit deal should have:

  • A pricing structure the buyer has seen and not pushed back on.
  • An understanding of procurement’s expected discount or term changes.
  • Awareness of any redlines legal is likely to raise.

If the deal team cannot answer all three with specifics, the deal is in best case at most, not commit. The forecast risk is not in the rep’s confidence. It is in the gap between the rep’s price and procurement’s price, which is invisible until procurement engages.

Question 7: What is the most likely reason this deal slips, and what would we do this fortnight to prevent it?

This is the close-out question for every commit deal on the forecast call. It does two things.

First, it forces the rep to articulate a slip thesis. Reps who can name the most likely slip reason in specific buyer terms — “the security review is the long pole; the buyer has not started it yet; if it starts after the 15th, we slip” — are running deals that do not slip. Reps who cannot name a specific slip thesis are running deals that often do.

Second, it produces an action. “What would we do this fortnight to prevent it?” turns the forecast call from inspection into intervention. The forum stops being a place where slippage is observed. It becomes a place where slippage is pre-empted.

As Mark Southgate puts it: “The forecast call is the cheapest piece of executive time in the company. Spend it on assumptions, not commitments. The number will be more accurate as a side effect.”

What this changes about the call

Seven questions, asked consistently, of every deal in commit and best case, do something the standard forecast call does not. They surface risk in advance of the slip. They make the call shorter — because most of the questions can be answered in 30 seconds for a healthy deal — and they make the categories more meaningful, because the categories now reflect inspected evidence rather than negotiated confidence.

This is the inflection point that makes forecasting advisory work compound. The category definitions tighten. The deal review forum starts pre-loading the answers. The pipeline review runs the four pipeline properties (coverage, freshness, distribution, motion) that the forecast inherits. Forecast accuracy rises, not because forecasting got better, but because everything upstream of it did.

The diagnostic

Run the seven questions on your top ten commit deals this week, with each owning rep. Time how long the answers take.

If most deals can be answered cleanly in under three minutes, the rep is running the deal with the buyer’s behaviour in view. Those deals will mostly close.

If most deals take longer, or hit “I don’t know” on more than two questions, the deal is being run on confidence, not evidence. Those deals will slip in roughly the proportion the research predicts.

The questions are not magic. They are just specific. The forecast call that asks them produces a different forecast. The forecast that comes from a coaching-supported operating cadence is more accurate than the forecast that comes from a clever model. The seven questions are the operating cadence, in question form.

Most forecast calls negotiate confidence. The useful ones inspect evidence. The difference is a Tuesday’s worth of conversation. The compounding return is the rest of the year.