Three buckets. One forecast. Most enterprise sales orgs fill them wrong every single week. Here's what each category is actually supposed to mean, the three ways AEs break each one, and a 90-second test to fix your forecast before Thursday's call.
Every major enterprise sales platform — Salesforce, Clari, Gong — uses the same three forecast categories: Pipeline, Best Case, and Commit. Every CRO uses them to build the number they defend on the board call. And most AEs fill them wrong every single week.
It's not a CRM problem. It's not a training problem. It's a misunderstanding of what the categories are supposed to represent — and what evidence is required to use each one honestly.
The result: weekly forecast meetings that turn into interrogations. VPs who adjust numbers anyway. Quarters that miss, even when Tuesday's call looked fine.
Fix the categories and you fix the call. Here's how.
These definitions aren't opinions. They're what Clari enforces in its commit-tracking workflows, what Salesforce Collaborative Forecasts uses for roll-ups, and what most CROs mean when they review the number.
Early-stage deals that have cleared basic qualification and belong in the CRM, but where the outcome is genuinely uncertain. These deals don't go in the forecast number. They go in the coverage number — the raw inventory a team needs to produce enough Commit and Best Case deals. A Pipeline deal has a defined problem and an interested buyer. It does not have a confirmed decision process, an identified Economic Buyer, or a next step with a hard date.
Deals that are more likely than Pipeline but don't yet meet the bar for Commit. The buyer is engaged, the deal is progressing, and there's a plausible path to close this period — if everything goes well. "If everything goes well" is the operative phrase. Best Case is your upside number, not your safety net. It should contain deals you'd genuinely be surprised to lose, where the remaining risk is execution rather than qualification.
Deals you are signing your name to. In Clari's framework, a Commit requires: a confirmed Economic Buyer who is actively engaged, a defined decision process with a timeline you can verify, a close date inside the forecast period that the buyer has confirmed (not just one you've put in Salesforce), and a documented next step that the customer has agreed to. Commit is not "I'm pretty confident." It's "I can defend this deal in 90 seconds with hard evidence."
Those three definitions give you the framework. Now here's what actually happens.
The most common mistake in enterprise forecasting: an AE puts a deal in Commit because they feel good about it. The champion is warm. The last call went well. The prospect didn't say no.
None of that is evidence. Feeling 80% confident about a deal is not Commit. A champion who's been responsive but hasn't introduced you to the Economic Buyer is not Commit. A close date you set without confirmation from the customer is not Commit.
What does abuse look like in practice? A $240K deal goes into Commit in week three of the quarter because the AE got a verbal. No mutual action plan. No confirmed timeline. No documented next step from the buyer's side. The call slips twice. By week eleven the deal is "probably Q3." The VP had to chase the same deal on six consecutive Monday calls before it disappeared.
Commit should be uncomfortable to use. If committing a deal doesn't feel like staking your credibility on it, your bar is too low.
Best Case gets used as a parking lot. A deal was in Commit, it slipped, and now it lives in Best Case while the AE figures out what to do. A deal was never fully qualified but the AE doesn't want to cut it from the forecast, so Best Case it is. A deal has had no Economic Buyer contact in 30 days, but Best Case feels safer than Pipeline.
The signal for Best Case abuse: if you can't name a specific buyer action that happened in the last two weeks, the deal isn't Best Case. It's Pipeline at best. A deal with a stale champion, no decision-maker engagement, and a close date that's been pushed once already has no business in the upside number.
When Best Case fills with deals nobody actually believes in, it stops being useful. The VP stops trusting it. You start getting questions like "what's actually in Best Case that's real?" — which means the category has already failed.
Pipeline carries deals that have no real qualification signal — no confirmed pain, no Economic Buyer identified, no defined buying process, no timeline. They stay in Pipeline because cutting them would mean acknowledging coverage is thin.
The problem isn't that the deals are in Pipeline. The problem is that they inflate coverage metrics, which managers use to decide whether the team needs help. If Pipeline contains five deals that would fail a basic MEDDIC qualification check, your coverage number is a fiction. You'll hit week ten of the quarter and realize you're three Commit deals short — and the Pipeline was never real.
The 90-second rubric was built specifically for this problem: a five-axis scoring check that takes less than two minutes and tells you whether a deal belongs in Pipeline, Best Case, or nowhere near the forecast.
Score any deal in 90 seconds. The rubric is embedded in the 6-section template — fill it out, share it, walk into forecast with a number your VP can trust.
Download Free →Here's the rule: before you move any deal into Commit, you need to be able to run a 90-second deal review out loud, from memory, with four things:
If you can produce all four in 90 seconds, the deal might be Commit. If you stumble on any of them, it's not. Full stop.
This is the same framework behind a deal review your VP will actually read — the evidence-forward structure that separates a real forecast conversation from a status update.
Walk through these four deals and place each one. Apply the 90-second test.
Situation: Champion is the VP of Operations. You've had four calls. He loves the product. Close date in Salesforce: May 30. No contact with the CFO. No mutual action plan. Last buyer action: they forwarded you a competitor's proposal three weeks ago.
Category: Best Case at best — probably Pipeline. No Economic Buyer engagement. No confirmed close date from the buyer. No next customer commitment with a hard date. This deal is not Commit regardless of how warm the champion feels.
Situation: CFO was on the last two calls. Legal received the MSA yesterday. Procurement sent a kickoff email for vendor onboarding. Close date May 22 confirmed by the CFO directly on the call. Next step: security review call May 16, already on calendar.
Category: Commit. Four signals: Economic Buyer engaged, contract out, procurement active, close date buyer-confirmed. Run the 90-second review — you'd clear all four items without hesitation. This is what Commit is for.
Situation: Two discovery calls. Strong pain identified — they're losing three hours per week to manual forecast prep. Budget confirmed verbally. Decision maker identified but not yet on a call. No proposal sent. Close date: "probably end of quarter."
Category: Pipeline. Good qualification signal, but no proposal, no Economic Buyer engagement, no buyer-confirmed timeline. This deal could become Best Case by week eight if it progresses — it's not there yet.
Situation: Was in Commit two weeks ago. Close date slipped from May 15 to "sometime in June." Champion says procurement is backed up. No contact with the Economic Buyer since week five. No next step with a date.
Category: Pipeline — remove from Best Case immediately. A Commit that slipped without a buyer-confirmed new date is not Best Case. It's Pipeline until the buyer re-engages with a hard commitment. Check the benchmarks data — the average deal slips 31 days after the first push. Two slips without re-engagement is a red flag, not a delay. See the 7 hidden causes of forecast slip for the common reasons deals go quiet between calls.
When AEs use forecast categories correctly — Commit only with buyer evidence, Best Case only with active engagement, Pipeline honestly — three things happen immediately:
Forecast accuracy goes up. According to Gartner, 93% of sales leaders can't call the quarter within 5% accuracy even two weeks out. The primary driver is Commit categories stuffed with hope. Tighten the definition and the number becomes defensible.
The Monday call gets shorter. When every Commit has a clear buyer action and a confirmed date, there's nothing to interrogate. The VP reviews the evidence, moves on. The call that used to take 90 minutes runs in 35.
The VP stops drilling on the same deals every week. The reason VPs revisit the same five deals every call is that those deals never had the evidence to justify their category in the first place. Move them to the right bucket and they either progress fast or get cut — either way, they stop consuming the meeting.
The discipline of honest forecast categories compounds. Once a team knows the standard, they stop sandbagging Best Case and stop inflating Commit. The forecast becomes a real planning tool instead of a weekly performance.
Pull up your current Commit list. Pick the deal you're least confident about. Run the 90-second review. Close date the buyer confirmed? Amount both sides agree on? Documented next customer commitment? One piece of hard evidence?
If you can't clear all four — move it. Best Case if engagement is recent. Pipeline if it's gone quiet.
That's the whole exercise. Do it for every Commit deal before Thursday's call and your VP will notice.
The Deal Review Template walks through the full evidence structure — CONFIRMED, ASSUMED, AT RISK — so you can run this for every deal in your book, not just the ones your VP is already watching. See a live example at /p/acme-q3-180k.
The CommitTrack Deal Review Template has the evidence structure built in — CONFIRMED, ASSUMED, AT RISK. Walk into every forecast call with a number your VP can't rewrite.
Download Free →See how the Commit evidence structure looks inside a live pack — the Acme Q3 sample has category evidence graded across all five axes.
Start Free → See a live sample pack →The Deal Review Template has the evidence structure built in. Score any deal in 90 seconds — walk in with a number that holds up under questioning.
Download the Free Template → Or try CommitTrack free — build a full Deal Review Pack in minutes →