Notes / Siteworks
How can a service business tell whether AI search brings qualified inquiries?
Measure AI-search visibility, visits and qualified inquiries with current Google and Bing reports, practical lead records and clear limits on attribution.
By Edu Rigonato. Published Sep 21, 2026. 10 min read.
A new referral source appears in your website report: ChatGPT. Around the same time, someone asks for a quote. It is tempting to connect the two. Before changing your marketing budget or commissioning more content, you need to know whether the report and the inquiry describe the same journey.
That question becomes more useful when you put it in business terms. Was the person looking for a service you provide? Could you serve their location? Did the conversation move toward work your team could accept? A source label cannot answer those questions for you.
You can build a useful view without tracing every conversation someone has with an AI assistant. Start with the inquiries your business actually receives, then keep the available evidence about their origin alongside the outcome. The gaps belong in the record too.
What counts as a qualified inquiry from AI search?
A qualified inquiry is a real request that meets your business's stated criteria for a worthwhile next conversation. Define those criteria before comparing channels. Otherwise, a change in what your team calls “qualified” can look like a change in marketing performance.
For a service business, the criteria might include the service requested, the area you cover, the type of customer you serve and whether the project is suitable for your team. Budget or timing may matter, but you may not know them until someone follows up. In that case, keep the inquiry awaiting review instead of forcing an early yes or no.
Consider a hypothetical commercial cleaning company. A visitor asks about ongoing cleaning for an office within its service area. The team confirms that the building and schedule fit, then arranges a site visit. That is a stronger business signal than a visitor clicking the phone number. A job application or a request for a service the company does not offer belongs in a different category.
The source evidence needs its own description. “The inquiry arrived during a session attributed to ChatGPT” and “the customer said they used ChatGPT while researching providers” tell you different things. Both can be useful. Neither should be rewritten to make the journey sound more certain than it is.
Frequently asked questions
Should we mark every form submission as a qualified lead?
No. Record a received inquiry separately from the team’s qualification decision. Exclude spam and duplicates, and keep requests awaiting review until you know whether they fit your criteria. Google’s <a class="inl" href="https://developers.google.com/analytics/devguides/collection/ga4/reference/events">recommended lead events</a> distinguish those stages, but your implementation must supply the rules.
What if the customer says ChatGPT, but analytics says Direct?
Keep both observations. Record the customer’s statement as self-reported discovery and preserve the source analytics actually captured. The disagreement does not justify relabeling all Direct traffic as AI. It may be useful context for that inquiry without establishing its complete journey.
Can we compare AI citations with revenue in the same report?
You can show them alongside each other with clear definitions and date windows. Do not imply that each citation is connected to a customer, or divide revenue by citations as though the report tracks a continuous customer journey. A shared dashboard does not create a missing attribution link.
Do we need a paid AI-visibility tool to start?
Start by identifying the question your existing reports and inquiry records cannot answer. A paid tool may help with a specific monitoring need, but it does not automatically verify lead quality or prove incremental revenue. Define the decision it would support before adding it.
Evidence note: This article draws on current platform documentation, public browsing research and retail referral research. Report capabilities were checked on September 21, 2026. The research does not establish service-business conversion benchmarks or causal attribution. The cleaning-company scenario and inquiry register are illustrative, not client results.
Explore Siteworks