Home · Business · Sales · Pipeline & forecasting

YES

As of 13 August 2026, AI can identify at-risk deals in your pipeline.

This still needs a person who signs their name to it.

Can you do it?

15 minutesto a draft.

30 minutesto something you’d act on.

Cost, all in£0

Skill neededpower-user

Who has to check ita colleague

What the alternative costsThe available tool data gives no price for a human or specialist alternative.

If this goes wrong: you focus a rep on the wrong opportunity or remove a viable deal from the forecast because the recorded evidence was incomplete.

What to actually do

  1. Hand it to a person

    The route this page recommends

    A person who owns the outcome does this end to end, worth it when the failure is dear.

  2. Use a tool built for this

    Second choice

  3. Do it yourself

    The distant third

    A chat interface, power-user skill, and roughly 30 minutes until you can act on the result.

    How to actually do it

    1. Open your CRM and export the active pipeline with deal name, owner, value, stage, expected close date, last activity, next step, contact details and forecast category.
    2. Gather the latest call notes, meeting transcripts and material email or activity history for those deals, removing records that you are not authorised to share.
    3. Paste the export, supporting records, stage definitions and team context into the prompt, keeping a clear label for each deal.
    4. Ask the model to rank the deals by risk and require evidence, missing information and one next action for every flag.
    5. Open each deal in the CRM and compare the model's evidence with the current stage, dates, activity log and next step, correcting any stale or mismatched data.
    6. Ask the model to revise the list using your corrections and to separate confirmed evidence from inference.
    7. Send the reviewed list to the sales manager or deal owner, who decides which buyers to contact and whether any forecast category should change.

    Prompt

    Analyse the pipeline data below and identify deals that may be at risk. Use only the information provided and do not invent buyer intentions, dates, values, competitors or activity. For each deal, return: deal name, current stage, risk level of low, medium or high, the specific evidence for the risk, missing information, what would change the assessment, and one practical next action. Separate facts from inferences, quote the relevant CRM or meeting evidence where available, and mark a field as unknown when it is missing. Do not treat a late activity, unchanged stage or missing note as proof that a deal is lost. First state the risk criteria you used, then give a table ranked by urgency, followed by deals with insufficient evidence to assess. Flag any inconsistent dates, stages or next steps. Do not recommend changing the forecast category unless the evidence supports it and the relevant forecast definitions are supplied.

Pipeline data:
[PASTE CRM EXPORT OR DEAL LIST]

Meeting notes or transcripts:
[PASTE RELEVANT NOTES]

Email or activity history:
[PASTE RELEVANT ACTIVITY]

Stage and forecast definitions:
[PASTE YOUR DEFINITIONS]

Additional context from the sales team:
[PASTE CONTEXT]

    Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.

What it gets wrong

  • AI cannot know a buyer's private change of priorities unless that signal appears in the supplied records.
  • It treats incomplete CRM data as evidence unless you make the missing fields explicit.
  • It cannot reliably distinguish a normal quiet period from a deal that has lost internal support.
  • It cannot take accountability for changing your forecast or deciding where a salesperson spends time.
  • It cannot verify whether a proposed next action fits the relationship, procurement process or internal politics without your context.

Even on a YES, the friction has a name: judgement under ambiguity, private data access and verification cost.

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.

AxisScore (0–2)
Output2
Inputs2
Verification1
Liability1
Effort delta2
Total8 / 10

The methodology and its thresholds are published in full.

FAQ

Can ChatGPT identify which deals are at risk?
Yes, if you provide the relevant CRM records, activity history and meeting evidence. It can rank warning signs and explain the evidence, but you must check the flags against the live deal and your team's context.
What data does AI need to spot at-risk deals?
Give it the pipeline export, stage and forecast definitions, expected close dates, recent activity, next steps and relevant call notes or transcripts. Missing or stale CRM fields reduce the quality of the assessment.
Can AI tell me whether a deal will close?
It can identify evidence that supports or weakens the current assessment, such as an overdue next step or inconsistent stage data. It cannot know a buyer's future decision, so treat the result as a review list rather than a prediction.
Is it safe to upload my CRM pipeline to AI?
Only use a service approved by your organisation and check its data handling terms before uploading customer or commercially sensitive information. Remove unnecessary personal data, follow your company's access rules, and do not paste records into an unapproved account.

Nearby answers

All Pipeline & forecasting rulings

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.