Commit, Best Case, Quota: The Three Numbers That Define Your Forecast
Commit, best case, and quota are the three numbers that define a sales forecast. Here's what each means, how accurate they should be, and how to defend them.
Every sales forecast comes down to three numbers: commit (what you'll close no matter what), best case (what you'll close if things go well), and quota (what the business expects you to deliver). The gap between those three numbers — and whether you can defend it with evidence from actual deals — is the entire forecasting conversation.
Get the definitions right and the forecast becomes a tool for running the business. Get them wrong and you end up in the 77% of sellers who miss their number even after quotas were lowered 13.3%.
What Each Number Actually Means
Most forecasting problems start with sloppy definitions. Reps call something "commit" because they want to look strong in the pipeline review. Managers call something "best case" because they don't want to defend a number. By Friday, nobody knows what any of it means.
Here are the definitions that hold up.
Commit: Deals the Rep Stakes Their Credibility On
Commit is the number the rep is willing to be wrong about publicly. It should close barring something unexpected — champion leaves, procurement blows up, budget pulled. Everything else is noise, not a reason to pull a deal from commit.
Working criteria for a deal to qualify as commit:
- Mutual close plan signed with the economic buyer
- Procurement and legal are engaged, not "scheduled"
- Verbal commitment to sign in the quarter
- No unresolved red flags (budget confirmed, decision criteria met, no competing active evaluation)
Target accuracy: 90%+. Well-run teams hit their commit number 75-90% of the time. Below 70%, you have a definition problem or rep overconfidence — not a sales execution problem.
Best Case: Commit Plus Deals Where One or Two Things Need to Go Right
Best case is commit plus the deals that have a realistic path to closing this quarter if specific things happen. "One or two things need to go right" is the test. If five things need to go right, it's pipeline, not best case.
Working criteria for best case (on top of commit):
- Clear next step booked with the decision maker
- Known blockers, each with an owner and a date
- Realistic close date based on sales cycle, not wishful thinking
Target accuracy: 60-70%. Best case is where coaching happens — this is the band where rep judgment either sharpens or stays fuzzy.
Quota: The Number the Business Committed To
Quota is not a forecast. It's the plan. The forecast is your honest read on whether you'll hit it.
The three numbers have different jobs:
| Number | What it is | Who owns the accuracy | Target variance |
|---|---|---|---|
| Commit | Deals the rep bets credibility on | The rep, backed by evidence | ±5% (90%+ accuracy) |
| Best case | Commit + deals with a realistic path | The rep + manager judgment | ±15% (60-70% accuracy) |
| Quota | What the business planned for | The CRO / planning team | N/A — it's a target |
Gartner data shows only 7% of companies reach 90%+ forecast accuracy, and 80% of sales organizations don't exceed 75%. If your commit lands within 5% of actual, you're in the top ~15% of teams.
Why Most Forecasts Miss
Pipeline reviews are mostly theater. The rep talks through deals. The manager asks about champions, budget, next steps. The rep answers from memory, from Slack threads, from the last call they vaguely remember. Half of what gets said never makes it into the CRM. The other half contradicts what's already there.
By Friday, the forecast number is a confident guess built on soft evidence. Then the deal slips, and nobody saw it coming — even though the champion stopped replying three weeks ago and the last call had three objections the rep never logged.
The forecast isn't wrong because the math is wrong. It's wrong because the inputs are wrong:
- Commit criteria aren't enforced. Reps commit deals that don't meet the bar because the bar exists on a Confluence page nobody reads.
- Best case is a feeling. "One or two things need to go right" turns into "I feel good about it" without naming the one or two things.
- Deal risk surfaces too late. Champions go quiet, buying signals weaken, objections get raised on calls — and none of it makes it into the pipeline review until the deal has already slipped.
Forecast categories should have different expected accuracy — commits at 90%+, best case at 60-70%. If yours all perform the same, the categories aren't doing any work. They're just labels.
How to Defend Each Number With Evidence
A defensible forecast ties every commit and best case deal to specific evidence from actual customer conversations. Not a rep's memory of what the champion said. The actual transcript, the actual next step, the actual risk signal from the last call.
This is where the pipeline review stops being theater. Instead of "how confident are you?" the question becomes "show me the moment in the last call where the buyer committed to a signature date." If the moment exists, it's commit. If it doesn't, it's not.
Demodesk is built for this. Every sales call — video, phone, in-person via the mobile app — is recorded, transcribed, and analyzed on the same record your CRM syncs from. AI Deal Insights flags deal risks and momentum changes from what was said: champion engagement dropping, buying signals weakening, objections raised but not resolved. AI CRM Concierge updates deal fields after every call so the pipeline you review on Friday reflects what the buyer said on Thursday, not what the rep remembered to type in on Wednesday.
That changes the pipeline review. Instead of debating whether a deal belongs in commit, the manager can ask Demodesk's AI Analyst across every conversation on the deal — "where did the economic buyer confirm the close date?" — and get a cited answer with the clip. Commit stops being a feeling. It becomes evidence.
Reps get something out of it too. AI Coach scores every call against your methodology (MEDDIC, BANT, or your own) in seconds, so reps know before the next call what they missed. Deals don't slip because the champion went quiet in week six — they get flagged in week three.
No new process required. The same deals, the same CRM, the same forecast categories. The forecast gets built on the record of what happened, not on what reps remembered to write down.
The Weekly Rhythm That Makes the Numbers Hold
Three numbers, three cadences.
Monday — commit lock. Every commit deal passes the criteria or drops to best case. No exceptions. The rep names the evidence from the last call or the deal moves.
Wednesday — best case stress test. For every best case deal, the rep names the one or two things that need to go right and who owns each. Can't name them? The deal drops to pipeline. "The champion will probably get back to me" is not an owner — the deal drops to pipeline.
Friday — forecast call. Commit and best case are locked. The call covers the handful of deals where new evidence came in during the week: a call that happened Thursday, a champion who went dark on Wednesday, a procurement conversation that changed the timeline.
Teams that run this rhythm consistently land within 5-10% of actual, putting them in the top ~25% of organizations. The ones in the top 15% — within 5% — are the ones where the evidence trail for every commit deal is visible to the whole team, not locked in one rep's head.
