Home · Business · Customer Service · Tier-1 support & chatbots
As of 13 August 2026, AI can qualify incoming customer enquiries before you reply.
Most people should hand this to a purpose-built tool.
Can you do it?
5 minutesto a draft.
30 minutesto something you’d act on.
Cost, all in£0/month
Skill neededpower-user
Who has to check ita colleague
What the alternative costsA purpose-built platform such as Botpress is an open platform for building LLM chatbots and agents.
If this goes wrong: the system sends a promising enquiry down the wrong route or rejects it, and you lose the chance to respond properly.
What to actually do
Use a tool built for this
The route this page recommends
Hand it to a person
Second choiceA person who owns the outcome does this end to end, worth it when the failure is dear.
Do it yourself
The distant thirdA chat interface, power-user skill, and roughly 30 minutes until you can act on the result.
How to actually do it
- Open your current enquiry inbox or chat export and choose a representative set of recent enquiries, removing unnecessary personal information before copying them.
- Write the qualification rules in one document, including who you serve, what you offer, exclusions, priority cases, routing options, required details and response expectations.
- Paste your current price, availability and eligibility information into the same document, or link the team to the source that staff must treat as current.
- Paste the prompt and one enquiry into ChatGPT, Claude or Gemini, then compare the proposed status with the written qualification rules.
- Test the prompt with clear enquiries, incomplete enquiries and ambiguous enquiries, and change the instruction if it does not route uncertain cases to a person.
- Send the output to a colleague for approval before replying, then copy the approved reply and record the classification and any corrections in your inbox or CRM.
- After reviewing a batch of real results, update the rules with recurring missing details and keep a human check for every case outside the documented criteria.
Prompt
You are helping qualify incoming customer enquiries for [BUSINESS NAME]. Do not make promises, invent facts or decide that an enquiry is unsuitable unless the rules below clearly support that decision. Qualification rules: - Ideal customer: [DESCRIBE YOUR IDEAL CUSTOMER] - Services or products offered: [LIST THEM] - Customers or enquiries we cannot accept: [LIST THEM] - Required details before a human replies: [LIST THEM] - High-priority situations: [LIST THEM] - Normal routing options: [LIST THEM] - Opening hours and response expectations: [ADD THEM] - Current prices, availability or eligibility rules: [PASTE THE CURRENT SOURCE OF TRUTH] For the enquiry below, return: 1. A concise summary using only information stated by the customer. 2. The customer’s apparent need or intent. 3. The facts that match or fail the qualification rules, quoting the relevant wording where useful. 4. Missing information, separating essential information from optional information. 5. A status of QUALIFIED, NEEDS INFORMATION, ROUTE TO HUMAN, or NOT A FIT. Use ROUTE TO HUMAN whenever the enquiry is ambiguous, sensitive, urgent, outside the rules, or requires a judgement not covered above. 6. The recommended next action and the reason for it. 7. A short, plain-English reply that asks only the necessary questions or confirms the next step. Do not claim that a service, price, appointment or outcome is available unless the supplied source of truth says so. 8. A confidence note explaining what could make the classification wrong. Incoming enquiry: [PASTE THE CUSTOMER ENQUIRY HERE]
Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.
What it gets wrong
- AI cannot know whether an unclear enquiry is commercially valuable unless you give it a rule or a person makes the judgement.
- It cannot guarantee that prices, availability, eligibility and service limits are current without a maintained source of truth.
- It can extract what the customer wrote, but it cannot reliably infer unstated urgency, buying authority or intent from a short message.
- You still carry responsibility for missed leads, unsuitable replies and inconsistent treatment of customers.
Even on a YES, the friction has a name: judgement under ambiguity, context depth and stakes of error.
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 | 2 |
| Verification | 1 |
| Liability | 1 |
| Effort delta | 2 |
| Total | 8 / 10 |
The methodology and its thresholds are published in full.
FAQ
- Can AI qualify leads for my business?
- Yes. It can extract details from an enquiry, compare them with your criteria, identify missing information and draft the next reply. Keep a person in the loop for ambiguous cases and decisions your rules do not cover.
- Can AI tell which enquiries are worth replying to?
- It can rank or route enquiries against criteria you provide, but it cannot reliably judge value that is not defined in those criteria. Use statuses such as qualified, needs information and route to human rather than letting it silently reject uncertain enquiries.
- How do I use AI to qualify customer enquiries?
- Give it your written qualification rules, current service information and the enquiry, then require it to show the evidence for its classification and draft a reply. Test it on clear and ambiguous examples before allowing the output into your normal workflow.
- Should AI reply to every customer enquiry?
- No. It can handle straightforward first questions, but urgent, sensitive, unusual or commercially important enquiries should go to a person. A human should approve replies whenever the consequences of a wrong route matter.
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.