# Close Date Is a Wish. Your Forecast Is Built on It Anyway. > Every CRM has a close date field. Almost no team treats it as the lagging, rep-controlled vanity metric it actually is. By Greg McCallum, founder of PeerLab. Published 2026-10-15 in Pipeline & Forecasting. Source: https://www.peerlab.ai/blog/close-date-is-a-wish-not-a-forecast-signal/ Every week, in every B2B sales organisation with a CRM, someone opens a forecast call by asking "what's your close date?" and then types a number into a spreadsheet as though that number means something. It does not mean what they think it means. Close date is the field that feels like a forecast input and functions like a mood reading. The rep sets it, the rep moves it, and the rep's main incentive when moving it is to avoid a difficult conversation rather than to report reality. Built on that foundation, your forecast is not a prediction. It is a performance. ## Why Close Date Feels Valid When It Isn't The appeal is obvious. Close date is concrete, it fits neatly into a pivot table, and it gives a forecast call a conversational hook. "You said end of Q3. We're two weeks out. Where are we?" That feels like rigour. It isn't. Three structural problems make close date a poor signal. **It is set by the rep without external verification.** No buyer has signed anything, agreed to anything, or even been asked about a timeline in a meaningful way. The rep typed a date. That's the audit trail. **It is updated retrospectively to absorb slippage, not predict it.** Deals don't slip on a forecast call. They slip quietly on a Tuesday when a rep moves the date from 31 March to 30 April and says nothing. By the time it shows up in a pipeline review, the narrative has been reconstructed. "They pushed back procurement sign-off" sounds like information. It is usually explanation-after-the-fact. **It creates a self-reinforcing story.** Once a date is in the CRM, the rep organises their activity around defending it rather than interrogating whether it was ever realistic. The date becomes the story. The story becomes the forecast. The forecast gets presented to a CRO who presents it upward. Nobody in that chain independently verified a thing. ## What the Field Actually Measures Close date is a proxy for rep confidence, occasionally seasoned with wishful thinking about quota attainment. That is not useless, but you should know that is what you are measuring. Reps with strong pipelines set optimistic dates. Reps under pressure set aggressive dates. Reps who have had a good discovery call but have not yet identified a decision-maker set dates based on vibes. The field is a sentiment indicator at best, and the sentiment it captures is the rep's, not the buyer's. If you want to know what close date correlates with, the honest answer is: quota attainment deadlines, end of quarter, and the rep's last conversation with their manager. These are not buying signals. ## Three Observable Signals That Actually Correlate With Timing There is a version of this conversation where I say "use buyer intent data" or "trust the AI forecast" and disappear. I am not going to do that. Here are three CRM-capturable signals that, used together, outperform close date as a timing predictor. Each has a failure mode by role, because each one gets gamed. **1. Confirmed next step with a named buyer contact and a specific date** Not "following up next week." A calendar invite accepted by someone on the buyer side, tied to a specific agenda item. This is the strongest single timing signal in a deal. If the buyer has committed time, there is momentum. If next steps exist only in the rep's notes as intentions, the deal is drifting. *Failure mode for reps:* Booking a placeholder call with an internal champion who has no authority, then logging it as a confirmed next step. Looks identical in the CRM. The fix is requiring the contact's job title and whether they are a decision-maker or economic buyer to be logged alongside the activity. **2. A documented mutual action plan or close plan with buyer-agreed milestones** This is different from a rep-authored close plan. A mutual plan that the buyer has seen, responded to, or modified is evidence of shared commitment to a timeline. A plan the rep wrote and never shared is creative writing. *Failure mode for managers:* Accepting "we've got a close plan" at face value. Ask to see it. Ask what the buyer said when they reviewed it. If the rep looks at you like you have asked them to produce a unicorn, the plan does not exist in any useful sense. **3. Procurement or legal engagement confirmed from the buyer side** Once legal or procurement is involved, you have a structural forcing function on timeline. These departments introduce real-world friction and real-world deadlines. Commercial terms, data processing agreements, security reviews: they take as long as they take. But their involvement is also evidence that the deal is real and that the buyer organisation is spending resources on it. *Failure mode for RevOps:* Not building a field for this. If the only evidence procurement is involved is a note in a free-text activity log, you cannot aggregate it, you cannot trend it, and you cannot use it in a forecast model. Build the field. Make it a picklist. Require it to be updated. If you want a structured way to track these against open deals, the [Deal Health / Risk Scorecard](https://www.peerlab.ai/library/tools/deal-health-risk-scorecard/) gives you a ready-built framework rather than starting from scratch. ## What to Do With Your Forecast Call Stop opening with "what's your close date?" and start opening with: - "What is the confirmed next step, who owns it on the buyer side, and when is it?" - "Has the buyer reviewed a mutual close plan? What did they say?" - "Is legal or procurement involved yet? If not, when do you expect that to change?" These three questions give you information the rep cannot easily construct from thin air. They also surface the deals that are theatre faster than any pipeline review matrix. A good [Win Rate & Forecast Accuracy Tracker](https://www.peerlab.ai/library/tools/win-rate-forecast-accuracy-tracker/) will also show you, over time, whether close dates in your team's pipeline are predictive or decorative. Run the correlation. Most teams that do this find close date accuracy sits somewhere around the same predictive value as the rep's horoscope for that quarter. ## Why This Persists The uncomfortable answer is that close date theatre serves everyone in the short term. Reps get to feel organised. Managers get something to report. CROs get a number to present. RevOps gets a clean field. Nobody is incentivised to surface the fact that the foundation is rot. The only person who suffers is the one trying to run a business on the forecast, and that person is usually also the one who signed off on the forecast cadence in the first place. That is the cycle. Breaking it requires admitting that a field your entire pipeline review was built around is measuring confidence, not evidence, and then building something harder to game in its place. The three signals above are not perfect. They can be gamed too, and I named the ways. But they require the rep to describe buyer behaviour rather than internal optimism, and that is a fundamentally different thing to ask. Your close date field is not going anywhere. But the next time someone presents a forecast built primarily on it, you now know exactly what they are presenting. **The tool for this:** [Close Date vs. Deal Evidence Scorecard: Three-Signal Forecast Audit](https://www.peerlab.ai/library/close-date-deal-evidence-scorecard-three-signal-forecast-audit/), free and no signup.