A siding contractor can tell AI search is producing leads when callers mention asking ChatGPT or Gemini for a recommendation, when referral traffic in analytics shows visits from ai.perplexity.com or chatgpt.com, and when estimate requests rise on weeks the business appears in AI-generated answers about local siding companies. None of these signals require guesswork if the owner asks the right question at the right moment and checks the right report.
Signs that AI search is driving inquiries
The clearest evidence that AI search is generating work is a homeowner repeating language back that sounds like a summarized answer rather than a search result, such as "I asked which siding companies near me had good reviews for James Hardie installs." Contractors should treat that phrasing as a signal worth logging, not a one-off anecdote, because it points to a specific channel worth tracking over time.
AI search refers to how tools like ChatGPT, Gemini, Perplexity, and Google's AI Overviews generate direct answers to questions instead of just listing links. When someone asks one of these tools "who does siding repair near me," the tool pulls from web content, reviews, and business listings to name specific contractors. A siding company that shows up in that answer gets a warm introduction before the homeowner ever visits a website.
Asking callers how they found you
The single most reliable measurement tool a siding contractor has is a consistent intake question: "How did you hear about us, and if you searched online, what did you type or ask?" This one habit, applied to every call and every quote request, surfaces AI-driven leads that no software will catch on its own, because many AI answers don't pass a trackable link back to the website.
Front-desk staff or the owner fielding calls should listen for specific phrases: "I asked ChatGPT," "an AI search," "I asked Google for recommendations and it gave me a list." These are different from a customer saying "I searched online" or "I found you on Google," which usually means a traditional search results page. Logging the exact wording, even in a simple spreadsheet next to the customer's name and job type, turns anecdotal impressions into a running count a contractor can review monthly.
It also helps to ask a follow-up: "Did it mention us by name, or did you pick us from a list it gave?" That distinction matters. Being named directly in an AI answer is a stronger signal of visibility than being one of several options a homeowner had to compare, and it tells the contractor whether their content and reviews are strong enough to be the recommended answer, not just a mentioned one.
Watching for AI-referred traffic and citations
Website analytics can show visits arriving from AI platforms, appearing as referral sources such as chatgpt.com, perplexity.ai, gemini.google.com, or copilot.microsoft.com, separate from traditional organic search traffic. A siding contractor who checks this traffic source monthly can see whether AI-driven visits are growing, flat, or nonexistent, and can compare that trend against the same period's estimate requests.
Most standard analytics setups group all traffic by source and medium, so isolating AI referrals means filtering the traffic source report for these platform names rather than looking at overall visitor counts. A contractor doesn't need to understand the filtering mechanics to ask a marketer or web person to set up a simple monthly view: "sessions from AI platforms" alongside "sessions from Google search."
Separately, some AI tools cite the sources they pull from when generating an answer, meaning a contractor's website, Google Business Profile, or review pages might appear as a named reference. Searching the business name plus terms like "siding contractor" or "siding repair" directly inside ChatGPT, Perplexity, or Gemini and reading how the tool describes the business is a manual but useful check. If the AI tool describes services accurately and names the business among recommended options, that's a sign the content and profile information are being read and trusted by the system.
Tracking estimate requests against visibility work
The number that matters most to a siding contractor's bottom line is estimate requests, not traffic or mentions, so any measurement of AI search should end at that number. Comparing estimate request volume before and after a period of visibility work, such as updating a Google Business Profile, adding project photos, or refreshing service pages, shows whether the effort is translating into homeowners actually reaching out.
This comparison works best when a contractor picks a consistent tracking window, such as month over month, and logs three things side by side: how many estimate requests came in, how many of those mentioned an AI tool during the "how did you hear about us" question, and whether any visibility changes were made that period. Over several months, patterns emerge. A jump in AI-attributed requests following a Google Business Profile update, for instance, suggests that update is worth repeating for other service pages or locations the business covers.
It's worth resisting the urge to judge results after a single month. Homeowners researching a siding project, especially a full replacement rather than a repair, often take weeks or months between the initial search and the actual call. A contractor should expect a lag between visibility improvements and a visible increase in estimate requests, and should track across a longer window rather than expecting immediate spikes.
Deciding what to double down on
Once a siding contractor has a few months of data, the next decision is which channel deserves more attention: traditional search rankings, AI-referred traffic, or direct calls citing AI recommendations. If AI-attributed leads are a small but steadily growing share of estimate requests, that trend justifies continued investment in the content and profile information AI tools rely on, such as detailed service pages, accurate business information, and customer reviews that mention specific services.
If, after consistent tracking, no callers mention AI tools and no analytics referrals appear from AI platforms, that's useful information too. It means the contractor's current customer base is finding them through other channels, and resources are better spent reinforcing what's already working rather than chasing a channel that isn't yet producing local traffic. Reassessing every few months makes sense, since AI search behavior among homeowners is still shifting and a channel that's quiet now may not stay that way.
The decision shouldn't be based on impressions or competitor chatter about AI search. It should be based on the contractor's own logged data: caller responses, referral traffic, citation checks, and estimate request counts tracked over a meaningful window. That data tells a specific business what's actually happening with its own customers, which is more useful than any general trend.
What to ask a marketer before hiring them for AI search
Any marketer claiming to help a siding contractor with AI search should be able to answer specific questions, and their answers reveal whether they understand the space or are just repackaging traditional search engine optimization under a new name. Ask them: "How will I know if this is working, specifically for my business?" A real answer names concrete signals, like the ones above, not vague promises of "more visibility."
Ask them how they'd check whether the business is currently mentioned in AI-generated answers, and ask them to show that check happening, not just describe it. Ask what they'd change on the website or business profile specifically because of how AI tools read and summarize content, as opposed to changes made for traditional search rankings alone. Ask how often they'll report on AI referral traffic and citation checks, and what that report will look like.
If a marketer can't describe a specific way to track results tied to estimate requests, or if their pitch relies only on general claims about AI search growth, that's a sign to keep looking. A contractor evaluating this kind of help deserves someone who can point to the same caller questions, traffic filters, and citation checks described here, applied specifically to their business and their market.