Roughly six in every ten B2B deals that get forecast for a quarter slip into the next one. That number, from CSO Insights and summarised in Forecastio’s analysis of forecast accuracy, has held remarkably steady for years. Median forecast accuracy across surveyed sales organisations sits between 70 and 79 percent. Fewer than half of sales leaders say they have high confidence in their own forecast.

The conventional response is to tighten the forecast. More fields. Sharper categories. More inspection on the call. AI overlays. None of it materially moves the slippage number, because the slippage is not happening on the forecast call.

The deal slipped 60 days earlier, in a moment of unverified buyer behaviour that nobody flagged.

“The deal you are explaining on the forecast call slipped two months ago. The forecast call is just where you found out.” — Mark Southgate

The work of catching slippage is not forecasting work. It is risk-inspection work, and it has to happen in the operating cadence before the forecast call ever opens.

What slippage actually is

In a separate dataset, Landbase’s 2026 B2B sales benchmarks report that 89 percent of B2B buyers experienced at least one stalled purchase in the past year. Stalled, not lost. The deal did not go to a competitor. The deal did not get cancelled. The deal just stopped moving.

That is the texture of most slipped forecast deals. They were not killed. They lost momentum on the buying side, and the seller did not see it in time to do anything about it. By the time the manager asked “what’s the status?”, the buying group had already drifted onto another priority.

The failure mode is not dramatic. It is gradual. A meeting gets rescheduled, then rescheduled again. The economic buyer goes quiet. The mutual plan does not get updated. Procurement starts asking questions that suggest they are seeing the deal for the first time. Each of those signals is a 30-day-early indicator that the deal is going to slip. Each of them is visible to a manager who knows what to look for.

The reason they get missed is that the deal review is asking the wrong question. It is asking “are you still going to close?” rather than “what has changed in the buyer’s behaviour in the last two weeks?”

The seven leading indicators of slippage

These are observable, by someone other than the seller, in any deal review:

  1. The economic buyer has not been in a meeting for more than three weeks. Not “the rep has been in touch.” Has not been in a meeting. The longer that gap runs, the higher the slippage probability.
  2. The mutual plan has not been updated in the last fortnight. If a mutual plan exists at all (and on most deals it does not), an out-of-date one is a stronger negative signal than a missing one.
  3. Procurement or legal has been engaged late. Late engagement means a compressed end-of-cycle. Compressed end-of-cycles slip approximately 40 percent of the time, often into the next quarter.
  4. The success metric has shifted. The buyer was solving Problem A two months ago. They are now describing the project as solving Problem B. Either the buyer’s priorities have shifted (high slippage risk) or the seller misread the original problem (high re-scope risk).
  5. A new stakeholder has appeared in the last 30 days. New stakeholders almost always add cycle time. If a new stakeholder appears late, the deal is mathematically unlikely to close on the current date.
  6. The deal has no confirmed close date in the buyer’s calendar. Not the CRM close date — the buyer’s actual scheduled signature day. If the buyer’s calendar does not hold a slot for it, the date is fiction.
  7. The rep cannot answer “what would make this slip?” in under a minute. Reps who can articulate the slip risk in specific buyer terms tend to be running deals that do not slip. Reps who cannot are running deals that often do.

None of these requires a new tool. All of them require a different question in the deal review.

What inspection looks like when it works

The inspection that catches slippage is short, specific, and runs weekly on the small number of deals that matter.

The shape of the conversation:

  • For the top 15–20 deals in the quarter, the manager runs a 5-minute structured check on the seven indicators above.
  • Where two or more are flagged, the deal moves into a deeper 20-minute review that week.
  • The deeper review does not ask “are you going to close?” It asks “what has changed in the buyer’s behaviour, and what action this week can change the trajectory?”

This is the discipline that turns deal reviews into an actual operating system rather than a status meeting. The forecast call that follows it has different texture. Risks are named in advance. Slip candidates are flagged before they slip. Categories are debated on evidence, not feel.

As Mark Southgate puts it: “A useful forecast call is boring. All the work happened in the deal review. The forecast just records what the deal review already knew.”

What the forecast coaching research adds

Forecast accuracy is not a fixed property of an organisation. Gartner research, also covered in Forecastio’s analysis, shows that organisations embedding forecast coaching into the sales management process improve accuracy by up to 15 percentage points.

Read that carefully. The improvement is not from a new tool. It is from coaching managers on how to inspect risk earlier. The forecast call is downstream of the coaching. The coaching is what produces the manager who asks the right questions in the deal review.

This is the part most forecast improvement initiatives skip. They invest in the call (more cadence, more rigour, more fields) and skip the coaching (training managers to inspect buyer behaviour). The investment goes into the wrong place. The accuracy does not improve.

“Forecast accuracy is a coaching problem disguised as a tooling problem. The manager who can ask the right risk questions on a Tuesday makes the Friday forecast call accurate by accident.” — Mark Southgate

The structural fix

Slippage drops when three things are in place:

A risk-inspection rhythm on the top deals, weekly. Not the whole pipeline. The top deals. The ones that determine the quarter. Five minutes each, against a defined set of buyer-behaviour signals.

A deal review forum that asks about change, not status. “What has changed in the buyer’s behaviour this fortnight?” is a different question from “what is the status?” The first is answerable with evidence. The second invites narration.

A forecast call that inherits the work above. By the time the deal reaches the forecast call, its risk profile has already been inspected. The call categorises rather than discovers. That is the only forecast call that runs in under an hour and produces a number that holds.

The diagnostic

Pick your last quarter’s slipped deals — the ones that moved from commit or best case into the next quarter. For each one, ask: when was the first 30-day-early signal of slippage observable, and was it discussed in any deal review before the slip became public?

If more than half of the slipped deals had visible signals that were never inspected, the operating cadence is the problem, not the forecast. Tightening the forecast will not change the number. Tightening the inspection will.

Six in ten deals slipping is not a permanent property of B2B sales. It is the steady-state outcome of an inspection rhythm that misses the early signals. Change the rhythm, and the number changes.