Home · Business · Sales · Pipeline & forecasting
As of 13 August 2026, AI can estimate when a deal will close.
Most people should hand this to a purpose-built tool.
Can you do it?
30 minutesto a draft.
1 hourto something you’d act on.
Cost, all in£0
Skill neededchat-fluent
Who has to check ityou
What the alternative costsNo comparable alternative price is provided in the available tool data.
If this goes wrong: you treat a weak opportunity as committed, distort the forecast and make staffing or revenue decisions on a date the buyer never agreed to.
What to actually do
Use a tool built for this
The route this page recommends
Do it yourself
Second choiceA chat interface, chat-fluent skill, and roughly 1 hour until you can act on the result.
How to actually do it
- Open your CRM and export the active deals with stage, stage-entry date, current close date, value, owner, next activity, last activity and any recorded decision-maker or blocker.
- Open the notes or recordings for the latest customer meetings in Fathom, tl;dv or your meeting-notes system, then copy the sections covering agreed actions, decision process, procurement, legal review and timing.
- Add the relevant email summary and a small set of comparable won and lost deals, removing customer names and other unnecessary private data before pasting.
- Paste the complete dataset into the prompt and ask the model to keep agreed customer dates separate from internal target dates.
- Check every estimated window against the CRM stage history, the customer's stated next step and the latest meeting notes, then correct any date or fact the model has misread.
- Send the resulting ranked list to the sales owner or forecast meeting with the missing evidence and slip reasons attached, rather than entering an AI estimate as a confirmed customer commitment.
Prompt
Estimate when each sales deal below is likely to close. Use only the information provided and do not invent buyer actions, dates, probabilities or deal facts. For each deal, provide: 1) estimated close window, 2) confidence level expressed as low, medium or high, 3) the evidence supporting the estimate, 4) the strongest reason it may slip, 5) the next customer-controlled milestone needed, and 6) the information that is missing. Separate an agreed customer date from an internal target. Treat a close date as uncertain unless the buyer has confirmed the remaining steps and timing. Use the deal's stage age, recent activity, next meeting, decision process, procurement or legal requirements, commercial value, champion strength, competing priorities and comparable historical deals where supplied. Do not use general sales benchmarks unless I provide them. At the end, rank the deals by likelihood of closing within the stated target period and explain which estimates are based on weak evidence. Data: [PASTE CRM EXPORT, CALL NOTES, EMAIL SUMMARY, DEAL STAGE HISTORY AND RELEVANT HISTORICAL DEALS HERE]
Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.
Hand it to a person
The distant thirdA person who owns the outcome does this end to end, worth it when the failure is dear.
What it gets wrong
- AI cannot know whether a buyer who has gone quiet still intends to purchase.
- AI cannot replace a salesperson's relationship knowledge about internal politics, trust or competing priorities.
- AI cannot verify that a customer-controlled milestone will happen on the stated date.
- AI cannot make an internal target date become a buyer commitment.
- AI cannot take responsibility for the forecast or the commercial decisions made from it.
Even on a YES, the friction has a name: real time truth, context depth and judgement under ambiguity.
How we scored this
Five axes, each scored nought to two by hand: ten means AI carries the task cleanly, and the thresholds that turn a total into YES, PARTLY or NO are published in the methodology. Each axis name links to its definition.
| Axis | Score (0–2) |
|---|---|
| Output | 2 |
| Inputs | 1 |
| Verification | 1 |
| Liability | 2 |
| Effort delta | 2 |
| Total | 8 / 10 |
The methodology and its thresholds are published in full.
FAQ
- Can ChatGPT predict when a deal will close?
- It can estimate a close window from CRM data, meeting notes and comparable deal history. It cannot see unrecorded buyer decisions, so treat the result as an evidence-based estimate rather than a prediction.
- How accurate are AI sales forecasts?
- Accuracy depends on the completeness and freshness of your pipeline data and on whether your historical deals resemble the current ones. Check the evidence for every estimate and keep customer-confirmed dates separate from internal targets.
- What data does AI need to estimate a deal close date?
- Give it stage history, time in stage, recent customer activity, agreed next steps, decision-makers, procurement or legal requirements, blockers and comparable won and lost deals. Missing information should be reported as uncertainty, not filled in by the model.
- Can AI update my sales forecast automatically?
- It can help analyse an exported pipeline and produce a revised view, but automatic updates need a CRM workflow and controlled data access. A salesperson still needs to confirm that the evidence reflects the current customer position before the forecast is used.
Nearby answers
Assessed by gpt-5.6-luna (gpt-5.6-luna) on 2026-08-13, second-checked by an independent model. Wrong somewhere? Email [email protected] and it gets re-checked.
The newsletter
AI news, new answers and product picks, straight to your inbox.