As of 13 August 2026, AI can only partly find a prospect's business challenges.

Most people should hand this to a purpose-built tool.

Can you do it?

15 minutesto a draft.

30 minutesto something you’d act on.

Cost, all in£0

Skill neededchat-fluent

Who has to check ityou

What the alternative costsApollo.io provides a prospect database with AI outreach sequences and enrichment, which is a purpose-built alternative for finding prospect signals.

If this goes wrong: you build a demo or outreach message around a challenge the prospect does not have and waste a sales opportunity.

What to actually do

Use a tool built for this
The route this page recommends
One pick for this task
Apollo.io
Apollo fits this task because its prospect database, AI outreach sequences and enrichment help you collect and organise prospect signals before discovery.

Do it yourself

Second choice

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

How to actually do it

Open the prospect's website, recent company announcements, relevant job adverts and other public company sources, then copy the useful text and source links into a working document.
Gather your CRM notes, previous emails, call transcripts and the prospect's stated role, removing unrelated personal information before pasting them into the chatbot.
Add your product description and the business outcome it is meant to support, then paste the complete context into the prompt.
Ask the chatbot to separate confirmed facts from hypotheses and to attach a source or note to every factual claim.
Open each cited source and compare it with the chatbot's factual summary, correcting or deleting claims that the source does not support.
Turn the strongest hypotheses into discovery questions and ask the prospect whether they apply before using them in a demo or follow-up.
Update the CRM with the prospect's confirmed challenge, evidence and next step rather than recording the AI's unconfirmed hypotheses as facts.

Prompt

Act as a sales research assistant. Identify likely business challenges for this prospect without presenting assumptions as facts.

Prospect company: [COMPANY NAME]
Website: [COMPANY WEBSITE]
Industry and market: [INDUSTRY]
Relevant role or contact: [ROLE]
Our product or service: [WHAT WE SELL]
Public sources and notes:
[PASTE PUBLIC COMPANY INFORMATION, NEWS, JOBS, REPORTS, CRM NOTES OR CALL NOTES HERE]

Use only the information supplied or clearly cited public sources. Do not invent facts, metrics, budgets, priorities or internal problems. Do not infer sensitive personal characteristics. Separate the output into:
1. Confirmed facts, each with its source or an explicit note that it came from my notes.
2. Three to five possible business challenges, each labelled as a hypothesis and linked to the evidence.
3. The likely business impact of each challenge, without estimating figures unless they are stated in the sources.
4. What is missing or uncertain.
5. Discovery questions that would test each hypothesis without leading the prospect.
6. A short, plain summary I could use to prepare for a sales conversation.

Rank the hypotheses by strength of evidence, not by how attractive they are to our sales pitch. Flag any source that may be outdated, promotional or ambiguous. Do not claim that the prospect has a challenge until the prospect confirms it.

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

Hand it to a person
The distant third
A person who owns the outcome does this end to end, worth it when the failure is dear.

What it gets wrong

AI cannot see the prospect's internal priorities, political constraints or problems that have not been recorded publicly.
It turns weak public signals into plausible-sounding challenge hypotheses that still require a conversation to confirm.
It cannot judge whether a stated problem is urgent enough to fund, or whether the contact has influence over it.
It cannot replace discovery, relationship-building or the judgement needed to choose the right question at the right moment.

What caps this at PARTLY: judgement under ambiguity, context depth and private data access.

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	1
Total	7 / 10

The methodology and its thresholds are published in full.

FAQ

Can ChatGPT find a prospect's pain points?: Partly. It can combine public company signals, CRM notes and call transcripts into evidence-backed hypotheses, but it cannot know which problems are real, urgent or worth funding until the prospect confirms them.
How do I use AI to research a prospect?: Give it the prospect's website, relevant public sources, your sales notes and a clear description of your product. Ask it to separate facts from hypotheses, cite every source and produce discovery questions rather than asserting that the prospect has a problem.
Can AI identify a prospect's business needs from LinkedIn?: It can identify signals such as stated priorities, hiring activity and company changes if you provide accessible, lawful source material. Those signals are not proof of a business need, and you should test them directly with the prospect.
Is AI reliable for sales prospect research?: It is useful for organising evidence and suggesting lines of enquiry, but it is not reliable as a source of unverified customer truth. Check every factual claim against its source and record only challenges the prospect has confirmed.

Nearby answers

All Demos & follow-up 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.