The Search Actually Starts on Google Maps, Not Google Search
Type "roofer near me" or "HVAC repair [city]" into Google and the first thing that loads above any blue link is the map pack: three pinned businesses with a star rating, a review count, and a distance. That's the real front door for local-service searches now, not the organic list below it, and for most trades it's the only part of the results page a homeowner looks at before tapping a name.
Homeowners scan that pack in seconds. They're not reading anything yet. They're pattern-matching: does this business have a real address in my area, does the star rating look solid, does the review count look like a business that's actually busy. A contractor sitting in position four or lower on that pack, or missing from it entirely, loses most of that first wave of lookers before a single word gets read. There's no consolation prize for being a great contractor with a weak or unclaimed profile; the homeowner never gets far enough to find out.
What moves a business into that pack isn't luck. It's a complete, verified Google Business Profile with the right primary and secondary categories, consistent name/address/phone data across every directory that lists the business, accurate service-area boundaries (not a blanket radius that includes towns the crew never actually reaches), and steady review velocity. That's a mechanical, fixable problem, not a popularity contest, and it's also one of the most commonly neglected pieces of a contractor's online presence, because it doesn't feel like "marketing" the way a new website or an ad campaign does.
- Homeowners check the map pack first because it answers "is this close and is this legit" in one glance.
- Star rating and review count are read together. A 4.9 with 6 reviews reads as newer or thinner than a 4.6 with 200.
- Profile completeness (photos, hours, services listed, Q&A answered) signals an active business over a stale listing.
- Category selection and service-area setup directly affect which searches trigger the profile at all.
This is the exact mechanic we work in our Local SEO & Google Maps and Google Business Profile Management services: getting the profile itself right, with the right categories, consistent data, and an active review flow, so the map pack has a reason to show it in the first place.
Reviews Get Read for Patterns, Not Just Stars
Once a homeowner has a short list from the map pack, they open reviews. Most don't read all of them. They skim the most recent five to ten and look for two things: does this business still answer the phone and show up on time, and how does the owner handle a bad one. Very few homeowners scroll past page one of reviews unless something specific catches their eye.
A single one-star review sitting unanswered for a year reads worse than three one-star reviews that each got a calm, specific reply from the owner. Homeowners have learned that no business is spotless. What they're actually judging is whether this contractor treats a complaint like a fire drill or like a normal Tuesday. An owner's reply that says the business made it right and explains what was done builds more trust than a page of five-star reviews with no owner presence at all.
Recency matters as much as volume. A contractor with 150 reviews but nothing new in eight months looks like it might not be the same crew anymore, might have changed hands, or might be sliding. A steady trickle of recent reviews, even just one or two a month, reads as a business that's still doing the work and still asking for feedback. Volume alone, without recency, can work against an older, established contractor who stopped asking for reviews once the phone was already ringing.
The job type mentioned in a review matters too. A homeowner needing a full re-pipe reads more into a review that specifically describes a re-pipe job than into a generic compliment with no detail. Specific reviews read as real. Vague ones, even genuinely five-star ones, read as thin.
| What homeowners notice | What it signals to them |
|---|---|
| Recent reviews (last 60-90 days) | Still active, still doing the work |
| Owner responses to negative reviews | Accountability, not just PR |
| Specific detail in reviews (job type, timeline) | Reviews look real, not bought |
| Review count relative to competitors | Relative trust signal, not absolute |
This is the review side of the funnel we run under Reputation & Reviews Management: building a steady, honest review flow and making sure every negative one gets a real answer, fast, not a form letter three weeks later.
The Website Is the Tie-Breaker, Even for a Five-Minute Visit
Once the map pack and reviews have done their filtering, most homeowners click through to the website for one reason: to confirm the business is real and does the specific job they need. They're not reading a full site. They're scanning for a phone number, a service list that matches their problem, and some sign the business has done this kind of job before.
A slow-loading site, a site that looks like it hasn't been touched since 2015, or a site with no clear way to call or text kills confidence fast. Homeowners don't file a complaint. They just go back to the map pack and click the next name down. That drop-off is invisible to the contractor. There's no lead in the inbox to explain what didn't happen.
What actually earns the click-to-call from a website visit is simple and mechanical: the phone number visible without scrolling, a load time under two seconds, service pages that match what the homeowner searched (not one vague "services" page for twelve trades), and some visual proof of past work, whether that's project photos or a clear before/after.
Site speed carries more weight than most contractors assume. A homeowner standing in a driveway or scrolling on a lunch break has no patience for a page that takes six or eight seconds to load. That delay alone, before any content is even read, pushes a real share of visitors back to the map pack to try the next name. Under two seconds is the practical bar to clear, not a nice-to-have.
- Homeowners give a site roughly 5-10 seconds to prove it's legitimate before bouncing.
- Mobile matters more than desktop. Most of this browsing happens on a phone standing in a driveway or kitchen.
- A missing or buried phone number is the single biggest silent job-killer on a contractor site.
This is the exact gap our Contractor Websites work closes: sites built to survive that five-second scan and convert it into a call, not just look nice in a portfolio.
A Growing Slice Now Asks AI Before They Ever Search
A newer pattern is showing up alongside the map pack habit: homeowners typing a full question into ChatGPT, Google's AI Overview, or a voice assistant instead of a bare keyword. Something like "who's a reliable roofer in [city] for a full tear-off" or "best rated plumber near me for a water heater replacement."
These tools answer in a short list, sometimes with a sentence or two of reasoning, pulled from whatever public information is structured clearly enough to summarize: review content, service pages, schema markup, and consistent business data across the web. A contractor that's invisible to AI answer engines simply doesn't get mentioned, no matter how good the actual work is.
This is still an early-stage shift, but it moves in one direction only. The contractors positioning for it now, meaning clean structured data, clear service pages, and a review profile with real substance, are the ones who'll show up as AI search keeps taking a bigger share of that first-touch moment. Waiting until it's the dominant channel means starting from zero after competitors have already built the signal.
The mechanics behind this are less mysterious than they sound. AI answer engines aren't inventing opinions about local contractors. They're summarizing what's already public and structured: FAQ content that directly answers a question, service pages that clearly state what a business does and where, and review text that gives the engine something specific to quote or paraphrase. A contractor with a bare-bones website and a handful of one-line reviews gives an AI engine almost nothing to work with, so it defaults to naming whoever does have that structure in place, whether or not that business does better work.
AI-search visibility is a newer layer on top of the map pack and reviews, not a replacement for them. A contractor with a thin, unstructured web presence has nothing for an AI engine to summarize in the first place. Getting the fundamentals right (profile, reviews, site) is also what makes a business legible to AI answer engines. It's the same underlying work, read by a different kind of reader.
The Full Path, Start to Finish
Put the four checkpoints together and the 2026 homeowner decision path looks less like a straight line and more like a funnel with hard cuts at each stage.
- Trigger: something breaks, ages out, or a homeowner decides to finally fix it.
- Map pack scan: three to five names get a first glance based on rating, review count, and distance.
- Review skim: the surviving two or three names get judged on recency, response to complaints, and specificity.
- Site check: a fast visit to confirm legitimacy, service match, and find the phone number.
- AI cross-check (a growing share of homeowners): a direct question to an AI tool, sometimes before the map pack, sometimes after, to sanity-check the shortlist.
- Contact: call, text, or form, usually to the two or three names that cleared every filter above.
Timing compresses this whole path more than most contractors expect. A homeowner dealing with a failed water heater or a leaking roof during a storm might run all six steps in under fifteen minutes on a phone. A homeowner planning a kitchen remodel might stretch the same steps over several days, revisiting the shortlist more than once. The order of the steps rarely changes. Only the pace does.
Notice what's missing from that list: a homeowner reading a full "About Us" page, browsing a blog, or comparing full-page quotes from five contractors. That version of shopping mostly doesn't happen anymore for local-service work. The filtering happens fast, mechanically, and mostly before contact.
The practical takeaway for a contractor is that marketing dollars spent only on the bottom of this funnel (a nicer truck wrap, a Yelp ad) miss where the actual cuts happen. The map pack, the review flow, and the website are the three checkpoints doing the real work, with AI visibility now riding on top of all three.
What Contractors Get Wrong About This Path
The most common mistake is treating these four checkpoints as separate projects instead of one connected system. A contractor spends on a new website but never touches the Google Business Profile feeding traffic to it. Or a contractor chases five-star reviews but the site those reviewers might click through to still takes eight seconds to load on a phone.
A second common mistake is assuming the highest star average wins. It doesn't. Homeowners read patterns, not averages. A business with a lower average but recent, detailed, well-answered reviews frequently beats a higher average that's gone quiet.
A third mistake is ignoring the AI layer because it feels early or unproven. The businesses treating it as optional in 2026 are the ones who'll be playing catch-up when it's not optional anymore. The work to prepare for it (clean schema, clear service pages, structured FAQ content) is the same work that improves the map pack and website checkpoints anyway. There's no separate budget line required to start.
| Common contractor mistake | What actually happens |
|---|---|
| New website, old/incomplete GBP | Traffic never arrives because the map pack doesn't surface the business |
| Chasing star average over review substance | Homeowners skip a stale 4.9 for an active 4.6 |
| No plan for AI search visibility | Invisible to a channel that's only growing share |
| One generic services page for every trade | Homeowner can't confirm the site matches their exact job, bounces |
The fix isn't a single big purchase. It's sequencing the existing pieces correctly: a verified, complete Google Business Profile feeding an active review request habit, a website fast enough and clear enough to convert the traffic that profile earns, and structured content clean enough for an AI answer engine to summarize accurately. Each piece reinforces the others. None of them work well in isolation.
None of these are expensive fixes in isolation. They're commonly just never done in the right order, or never done at all, because each one looks like a small task until you add up what it costs in missed calls.