The Search You Never See

Somewhere in your service area right now, a homeowner has a wet basement and ten minutes before the kids need to be picked up. A few years ago, she’d have opened a search engine, scrolled past some ads, and started clicking through ten tabs of reviews and websites, trying to figure out who to trust. Today, a growing number of people skip most of that. They open an AI assistant, type “who should I call for water in my basement in [her town],” and get an answer with names attached. Then they call the first name on the list.

You can’t see that search happen. It doesn’t show up in your website analytics. It doesn’t show up in your Google Ads dashboard. It happens entirely inside someone else’s product, and the first sign you get is the phone ringing with a customer who says “I asked ChatGPT and it told me to call you,” or, worse, never mentions it at all because it never occurred to them that it mattered.

That’s not a reason to panic, and it’s not a reason to chase a new tactic this week. It’s a reason to understand what’s actually going on, because you can influence what that assistant says about you. And here’s the part worth sitting with: most of what it takes is work you should already be doing.

What the Assistant Actually Does When Someone Asks

Here’s the part that’s knowable, as of this writing, without speculating about anything proprietary. When you ask ChatGPT, Perplexity, or Google’s AI Overviews a local-service question, you don’t get a list of ten blue links to sort through yourself. You get a named answer with a reason attached. A recommendation, sometimes two or three, with a sentence on why this business and not another.

Where does that answer come from? On the assistants that show their work, and several of them do, you can literally watch the sources it pulls from as it builds the answer. Google reviews and star ratings. Business directory listings. Local news coverage of a storm or a company milestone. Forum threads where someone asked the exact same question a year ago and got real answers from real people. And the business’s own website, when that website actually answers the question being asked instead of describing services in vague, general terms.

It’s worth separating two things people tend to lump together under “AI search.” Google’s AI Overviews sit at the top of a regular Google search results page, and they still lean heavily on the same signals that drive the map pack and the organic results you already track: your Google Business Profile, your reviews, your rankings. Standalone assistants like ChatGPT and Perplexity are a different surface entirely. Someone might open one of those apps directly and never touch a search engine that day at all. Both are doing something similar underneath the surface: reading the public record about your business and turning it into a plain-language answer for the person asking. Neither one is inventing information about you out of nowhere. They’re synthesizing what’s already out there and public.

That’s the whole mechanism, as far as anyone outside the companies building these products can actually observe it from where we sit. No secret ranking factor to chase, no trick to learn. Just what’s publicly true and findable about your business, compressed into an answer a homeowner reads in three seconds and acts on.

What We See Running the Checks Ourselves

We’re not guessing at any of this from the outside looking in. We run a monthly AI-visibility check for a home-service client right now: the same panel of questions a real customer would plausibly ask, run every month, scored on whether the business shows up, gets recommended by name with a reason, or doesn’t show up at all.

A couple of patterns have held steady enough across that work to be worth telling you about, even without attaching a single number to them.

The first: the same business can be the assistant’s confident, named first recommendation for one kind of question and be completely invisible for another. Ask a near-me, “who should I call right now” question, and a business with a strong, recent review history and a complete profile tends to show up reliably, often at the top. Ask a how-to question instead, something the business’s own website should be the obvious authority on, and that same business is sometimes nowhere in the answer, while a competitor’s blog post or an old forum thread gets quoted instead. It’s not that the business is worse at the work. It’s that nobody ever gave the assistant anything specific to point to for that particular question, so it reached for whatever public source did answer it.

The second pattern: the sources these assistants lean on hardest are the same trust assets that already win you the map pack in a normal Google search. Review depth and recency. A complete, accurate Google Business Profile. Consistent information about the business across the directories and sites that mention it. If you’ve been building those things because they help you rank and get chosen in ordinary search, you already have most of what an assistant needs in order to recommend you. If you haven’t, the gap tends to show up in both places at once, not just one.

Your AI Resume Is Already Written

Here’s the reframe that actually matters, and it’s simpler than most people expect: you don’t optimize for AI. You give the public record about your business more true, specific things to say, and the assistant reads whatever’s already sitting there.

Four things feed that record, roughly in order of how much they seem to matter based on what we watch month over month.

Review depth and recency. An assistant weighing whether to recommend you is effectively reading the same reviews a customer would read, just faster and at higher volume. A page of five-star reviews from two years ago says less than a steady flow of recent ones that mention specifics: the technician’s name, the exact job, how fast someone showed up on a bad day. This is a system, not a one-time push, and we’ve written about the mechanics of building that flow instead of hoping for it in our piece on building a review system that compounds trust.

A complete, accurate Google Business Profile. Service areas, hours, categories, photos, the Q&A section, all filled in and kept current as your business actually changes. This is the single most structured piece of public data that exists about your business, and it’s the first place both search engines and assistants look before anything else. If yours has gaps, here’s what a complete profile actually requires, section by section.

Consistency across every directory and site that already mentions you. Your name, address, and phone number should match everywhere, because assistants cross-reference multiple sources before they’ll commit to a specific recommendation. A business that looks like three slightly different companies across the web, one address here, a different phone number there, reads as less trustworthy, to a machine doing pattern matching and to a person doing the same thing by instinct.

Content that answers real questions in plain language. Not slogans about being “the trusted choice since 1998.” Actual answers to what homeowners ask before they ever pick up the phone: how much does this typically cost, how long does a job like this take, what’s the real difference between fixing it now and waiting a season. Assistants quote answers. They don’t quote taglines, and they can’t quote a page that never actually said anything.

None of this is a new project you need to start from scratch. It’s the same foundation that has always mattered for getting found and getting chosen, whether the person on the other end is scrolling ten tabs or asking one assistant a single question. The assistants just read that foundation faster and package the answer differently.

How Big Is This, Honestly

Here’s where we tell you the plain truth instead of the sales pitch, because the honesty is the whole point of this piece. This is a small slice of how homeowners find you today. Search still dominates, by a wide margin, and word of mouth and repeat business still carry more of your book than either. Nobody has reliable numbers on how many people are actually asking an AI assistant for a local recommendation instead of searching the normal way, and we’re not going to pretend to hand you a percentage. Anyone who quotes you a precise adoption figure for this, in a sales deck or a cold email, is making it up. Treat it as a red flag, not a data point.

What we do know is that it’s growing, and that the assets feeding it, reviews and genuinely useful content especially, take months to build no matter when you start building them. That’s the actual argument for paying attention now, and it’s a quiet one, not an urgent one. Being early here costs you nothing extra if you’re already doing the reviews and the content work you should be doing for ordinary search and word of mouth anyway. What would be premature is redirecting real budget toward “AI visibility” as its own line item, or worse, buying whatever AI-optimization package starts showing up in your inbox from people who read this trend a week before you did. If someone’s selling you a shortcut to being recommended by an assistant, that’s a sign to hang up the phone, not a sign to open your wallet.

We genuinely don’t know yet whether this becomes a channel you plan a budget around or stays a rounding error next to search for years to come. What we do know, from watching it every month, is that the fundamentals feeding it are the exact same ones that have always mattered for a home service business trying to get chosen.

Answerable Is the New Findable

You don’t need a new strategy for this, and you don’t need to hire anyone specifically for it. You need the same strategy you already know, held to a slightly higher standard: reviews that keep coming in and say something specific, a Google Business Profile that’s actually complete instead of half-filled-out, business information that matches everywhere it appears online, and content that answers the questions your customers are genuinely asking instead of talking about how great you are. Build those four things, and you’re already positioned for whatever this becomes over the next few years, without ever having spent a dollar chasing “AI SEO” as its own category.

That’s the same system we build into the ads-led plans we run for home service clients: the fundamentals done consistently enough that they compound, whether the surface reading them is a search engine, an assistant, or a neighbor asking around. If you want a second set of eyes on where your reviews, profile, and content actually stand today, see the plans and find the one built around your business.