Quick answer: Use AI for lead qualification by supplying explicit fit criteria and asking it to organize the enquiry, identify missing information and suggest a next step. Keep uncertain cases visible and require appropriate review before customer-facing messages or consequential decisions. Measure qualified outcomes and routing accuracy alongside response speed.
In this guide: Define fit in business terms · Design the intake record · Use AI to prepare a structured assessment · Create a clear handoff path · Worked example: an incomplete website enquiry · Measure more than response time · Maintain the rules as the business changes · Frequently asked questions
An enquiry workflow should help a potential customer reach the right answer. It should not create a system that rejects people because their first message is short, unusual or written in a different style.
AI can assist with interpretation and drafting, while rules and people retain control over what the business promises and how the enquiry progresses.
Define fit in business terms
Write the conditions that make your service appropriate: the problem you solve, service area where relevant, required scope, timing and other legitimate practical constraints. Use information the prospect has actually provided.
Separate “not a fit” from “not enough information.” An enquiry with no budget stated is incomplete, not automatically unqualified. A person asking for a service outside your scope may need a helpful referral rather than a generic rejection.
Do not infer sensitive personal characteristics or unrelated purchasing ability from a name, writing style or other proxy. Qualification should focus on the actual service requirements.
Design the intake record
| Field | Purpose | Handling when missing |
|---|---|---|
| Requested outcome | Understand what the prospect wants to achieve. | Ask a focused follow-up question. |
| Service or project scope | Check whether the business can help. | Route for review if unclear. |
| Relevant location | Confirm coverage where it matters. | Request the service location. |
| Timing | Understand urgency and capacity fit. | Avoid promising a start date. |
| Contact preference | Respect how the person wants to continue. | Use the established enquiry channel appropriately. |
| Source and notes | Preserve the original context. | Keep the original message available. |
Collect only information needed for the next decision. A long qualification form can create friction before a person understands the offer.
Use AI to prepare a structured assessment
Ask for a summary, evidence for each fit criterion, missing information, suggested route and a draft response. Require the model to use “unknown” when the enquiry does not provide an answer.
An example instruction is: “Assess the enquiry against the supplied service criteria. Cite the exact part of the message that supports each conclusion. Do not infer unstated budget, urgency or identity characteristics. If a required fact is missing, recommend clarification. Draft a response using only approved service information.”
Treat the output as a recommendation until the workflow has been evaluated for the relevant action. Preserve the original enquiry so a person can correct an interpretation.
Create a clear handoff path
Use routes such as suitable for a discovery conversation, needs clarification, outside current scope and requires specialist review. Assign an owner and response expectation to each route.
A workflow can notify an internal owner or prepare a draft without automatically sending it. Before an external response, check the recipient, the requested service, the accuracy of any commitment and the appropriate contact permissions.
Provide a way for staff to override the recommendation and record why. Those corrections can reveal unclear criteria or recurring interpretation errors.
Worked example: an incomplete website enquiry
A hypothetical prospect writes: “We need a new website for our installation company before our new service launches.” The message suggests a relevant project but does not establish scope, timing, content readiness or budget.
The appropriate AI-assisted output is a short summary and a request for the missing details. It should not fabricate a project value or promise delivery before the launch date. A person can then decide whether to offer a conversation and which questions are most useful.
Once the details are supplied, the workflow can prepare a project brief for review. The customer should not have to repeat information because the system failed to preserve the context.
Measure more than response time
Track time to a useful first response, correct routing, clarification rate, qualified enquiries, proposals and customer outcomes. Sample cases the system considered unsuitable to check for missed opportunities.
Record rework and complaints. A system that replies immediately with irrelevant questions may perform worse than a slightly slower, informed response. Review whether the automation helps the customer move forward.
Keep operational logs and access appropriately controlled. Use approved tools and verify data-handling requirements for your situation before connecting customer systems. This is a workflow design guide, not a determination of your legal obligations.
Maintain the rules as the business changes
Update service criteria, availability, pricing references and escalation contacts whenever they change. Test the workflow with common and difficult examples after significant revisions.
Keep a manual fallback. If the tool is unavailable, the business should still receive and respond to enquiries. Automation should support a dependable process rather than become the only route to a customer.
Frequently asked questions
Should AI reject leads automatically?
Be cautious with rejection based on incomplete or interpreted information. Route uncertain cases to clarification or review, and evaluate any automatic rule against real missed-opportunity risk.
Can AI write the follow-up email?
It can prepare a draft using approved facts. Sending should follow the business's authorization, contact preferences and review requirements.
What if our criteria are unclear?
Define them before automating. An unclear manual process usually becomes a faster unclear process when software is added.
Continue reading
- Which Marketing Tasks Should You Automate With AI—and Which Need Human Review?
- How to Define Your Ideal Customer Profile Before Spending on Marketing
- AI-Powered Marketing for Small Businesses: Where Should You Start?
Plan your next step. Discuss your website and marketing priorities with Stefano Marketing Agency. We can clarify your goals and agree on the appropriate scope.