A good sales forecast accuracy rate is one that lands inside a band you set before the quarter starts, quarter after quarter, with the deals in commit closing for the reasons the reps gave.
The short version
Most B2B teams talk about accuracy as one percentage: the number called compared with the number closed. That headline matters, because finance plans hiring and spend on it. It is also easy to hit by luck. If three committed deals slip and two unforecast deals land, the headline can look perfect while the forecast told leadership very little.
So judge it twice. First the headline, against a band you choose in advance (for example, within 10% of the week-4 call). Then the deals: of everything in commit at the snapshot, what share closed in the period? A team whose headline is good and whose commits mostly closed has a forecast it can trust. A team that hits the number on unforecast deals has a forecast that needs work, even in a good quarter.
Set the band to fit the business. A team closing four large deals a quarter will swing more than one closing 400 small ones, and a single slipped deal can move the headline by 20 points. For the large-deal team, the deal-level number is the better guide.
The benchmark, and where it comes from
The range quoted most often is that quarterly forecasts across most B2B industries land within 8% to 15% of actuals, with top performers above 95% accuracy. Salesloft's guide cites it to Eagle Rock CFO research, though the linked research page returned "Page Not Found" when we checked on 31 October 2026. Treat it as a rough guide.
Check which formula sits behind any benchmark you compare against. Salesloft's own example shows a team that forecasts $10M and closes $9.5M reporting 95% accuracy under one method and a 5.3% error rate under another. HubSpot reports accuracy as one minus the absolute error divided by actual revenue, floored at zero. Pick one and keep it.
An illustrative quarter
This example is illustrative. A team calls $2.0M at week 4 and closes $1.92M: 96% on the headline, inside its 10% band. The deal review shows something else. Of 18 deals in commit, 10 closed. The other 8 slipped, and the gap was covered by three large deals that were in best case at week 4.
The manager traces each slip to the buyer evidence that was missing at the snapshot. Five had a close date the buyer never confirmed. Three had a signer who had not heard the case. Next quarter, commit requires a buyer-owned date and a signer on a call, and each rep with a deal missing either one gets a coaching call that week to plan the conversation that fills it, with the manager, or by bringing the deal to Caty, one of our AI sales experts, on a call. Bryce, our expert for managers, gives the manager the same buyer-evidence read across every deal in commit before the forecast call.
The two numbers to track
Put both numbers on the same line of the quarterly review, every quarter, from the same snapshot.
Headline accuracy. The number you called against the number you closed. The illustrative team above closed $1.92M against a $2.0M call. Dividing actual by forecast gives 96%. The error-rate method divides the $80,000 gap by the $1.92M actual, about 4.2% off, which some teams report as 95.8% accuracy. Both describe the same quarter, which is why the formula has to be fixed before anyone compares. The full arithmetic, including how to average across several periods, is in How to calculate forecast accuracy.
Commit conversion. Of the deals in commit at the snapshot, the share that closed inside the period. In the same quarter, 10 of 18 closed, about 56%. This number does not care how big the deals were or what else landed. It asks one question: when a rep said a deal would close, did it?
Read together, the two numbers tell four stories.
| Headline | Commit conversion | What it usually means |
|---|---|---|
| Inside the band | High | The forecast is working. Keep the definitions stable. |
| Inside the band | Low | Luck. Unforecast deals covered for slipped commits. |
| Outside the band | High | Commit is honest, but too much of the number sits outside it. Look at best case and pipeline. |
| Outside the band | Low | Commit is not tied to buyer evidence. Start with the category rules. |
The second row is the one that catches teams out, because the quarter looks fine in the board pack.
Why forecasts miss
The causes repeat across most teams, and most of them sit at the deal level.
- A close date the buyer never agreed. The rep picked the end of the month. Nobody on the buyer side has a reason to sign by then. Our guide on coaching reps on close dates covers the conversation that fixes it.
- A signer who has not heard the case. The champion is keen, but the person who approves the spend has not been on a call. The deal can stall for a quarter at the approval step.
- Rep optimism. Reps want deals to close and read friendly buyers as committed buyers. This is normal and it is why commit needs written evidence behind it.
- Stale records. A deal that has not changed in three weeks still sits in commit because nobody moved it. The snapshot then counts a deal that was already drifting.
Why sales forecasts are wrong goes deeper on the rep side of this list.
What changes a reasonable target
There is no single right band. Four things move it.
Deal count. A team closing hundreds of deals a quarter averages out its surprises. A team closing a handful does not: one slip moves the headline by double digits, so it cannot tighten the band below what its deal mix allows and should rely more on commit conversion.
Sales cycle length. If most deals take nine months, the quarter's forecast is mostly made from deals already in late stages, and the band can be tight. If deals take three weeks, much of the quarter's revenue has not entered the pipeline at week 4.
Snapshot timing. The earlier the snapshot, the wider the band you should accept. Measuring at week 10 flatters accuracy because most of the uncertainty is gone; week 4 tests the forecast when leadership is using it to plan. Pick the week that matches when finance needs the number and keep it.
Change in the business. A new product, a new segment or a price change makes history a weaker guide. Widen the band for a quarter or two and say so in advance.
How to set your band in five steps
- Choose the snapshot. Agree with finance which week's commit counts as the call, usually week 3 or week 4.
- Choose one formula. Write it down with a worked example so nobody computes it differently next quarter.
- Look back four quarters. Measure the headline and commit conversion from past snapshots, using the CRM history.
- Set the band from that history. Start a little tighter than your typical miss, and set a floor for commit conversion alongside it.
- Review the slips, every quarter. For each committed deal that did not close, record the missing buyer evidence. The most common gap becomes next quarter's coaching focus.
Using the numbers in the weekly forecast call
Accuracy is measured once a quarter, but it is made every week. Three habits in the weekly forecast call move both numbers.
Ask for the buyer evidence behind each committed deal, out loud, before discussing the amount. A deal with no buyer-agreed date or no signer on a call drops to best case until the gap is filled. Keep a short list of every deal that moved category during the week and why, so the end-of-quarter review starts from notes. And when a deal slips, record it the day it happens, while the rep still remembers what the buyer said.
Done weekly, the quarterly review becomes a reading of patterns the team has already seen. Done quarterly, it becomes an argument about memory.
Common mistakes
- Moving the snapshot. Taking the number later in the quarter when the early one looks bad makes every trend meaningless.
- Rewarding the headline alone. Praising a manager for a 96% quarter built on unforecast deals teaches the team that commit does not matter.
- Mixing formulas across teams. Two regions reporting 95% under different methods are not comparable.
- Sandbagging to protect the number. Holding good deals out of commit makes the headline easy to beat and leaves finance planning on a number that is too low. Under-calling is a miss too.
- Treating accuracy as a rep metric. Reps own the buyer evidence on their deals. The category decision belongs to the manager, so the coaching starts there.
Raising the rate
- How to improve sales forecast accuracy: the seven-step procedure, from fixing the snapshot to coaching every slip.
- Deal risk guide: where deal risk comes from and how to act on it.
- How to run a forecast call: the weekly questions that keep commit honest.
FAQ
Is 90% forecast accuracy good?
For most B2B teams, landing within 10% of the called number is a solid quarter, and Salesloft's cited benchmark puts typical quarterly forecasts 8% to 15% off actuals. Check it at the deal level too: if the 90% came from unforecast deals covering for slipped commits, the accuracy was luck and next quarter can miss badly.
Should forecast accuracy be measured monthly or quarterly?
Measure it on the period you call to leadership, usually the quarter, from a fixed snapshot such as the week-4 commit. Teams with short cycles can add a monthly view. Changing the snapshot or the formula between periods makes the trend meaningless, so fix both before the first measurement.
Who owns sales forecast accuracy?
The front-line manager owns it, because the manager decides which deals go in commit. RevOps owns the definitions, the snapshot and the measurement. Reps own the buyer evidence on their own deals, which is what the category decision should rest on.
How do you calculate forecast accuracy?
Divide the revenue you closed by the revenue you called at a fixed snapshot, or take one minus the absolute gap divided by actual revenue. Use one method every period. Alongside it, divide the number of committed deals that closed by the number in commit at the snapshot.
Is under-forecasting better than over-forecasting?
Neither under-forecasting nor over-forecasting is good. Over-calling leaves finance short when deals slip. Under-calling hides pipeline, makes the team look better than its forecast deserves and leads to hiring and spend that are too cautious. Both show up as a miss against the band.