As of 13 August 2026, AI can detect buying intent in lead emails.

What to actually do

1. Use a tool built for this

   The route this page recommends

2. 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

   1. Open your CRM or email export and copy the relevant lead emails, giving each message a unique ID and removing unrelated threads.
   2. Write down your product or service, ideal customer profile, buying signals and low-intent signals in plain language.
   3. Define what high intent, medium intent, low intent and unclear mean for your sales process, including any signal that should always trigger human review.
   4. Paste the business context and emails into a chatbot with the prompt above, then ask it to classify every message without using information outside the supplied text.
   5. Compare each label and quoted evidence against the original email, correcting any label that is not supported by the wording.
   6. Send the high-priority leads to the responsible salesperson and record the classification and follow-up outcome in your CRM so the criteria can be improved.

   Prompt

   Classify the lead emails below for buying intent. Use only evidence in the emails and the business context I provide. Do not infer intent from names, job titles, writing quality or demographic details.
   
   Business context:
   - What we sell: [PRODUCT OR SERVICE]
   - Ideal customer profile: [IDEAL CUSTOMER PROFILE]
   - Buying signals that matter to us: [SIGNALS, SUCH AS ASKING FOR PRICE, AVAILABILITY, A DEMO, CONTRACT TERMS OR IMPLEMENTATION DETAILS]
   - Low-intent signals: [SIGNALS, SUCH AS GENERAL RESEARCH, UNSOLICITED PROMOTION OR NO RELEVANT REQUEST]
   - Our qualification labels: high intent, medium intent, low intent, unclear
   
   For each email, return:
   1. The assigned label.
   2. A confidence level of high, medium or low. Treat confidence as confidence in the evidence, not certainty about the person's future purchase.
   3. Two short quotations or precise references from the email that support the label.
   4. The buying question or need expressed by the lead.
   5. The next sales action, or say "do not prioritise yet" if there is no useful action.
   6. One piece of missing information that would change the classification.
   
   Separate explicit buying signals from weak or inferred signals. Mark an email as unclear when the evidence is mixed or insufficient. Do not invent budget, authority, timescale, company needs or purchase plans. Finish with a table ranked by recommended follow-up priority and a short explanation of the ranking.
   
   Emails:
   [PASTE LEAD EMAILS HERE, EACH WITH A UNIQUE ID]

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

3. 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.