The old argument for skipping a website
There was a version of this argument that used to be reasonable. A contractor with a full truck, a good reputation, and a Google Business Profile could run for years on referrals and map-pack calls. Facebook covered the "prove you're real" job. A website felt like overhead for a business that was already getting the phone to ring.
That case was never airtight, but it was defensible through most of the 2010s. Search results were mostly links. A strong GBP listing with reviews and photos could out-rank a thin, dated website. Contractors who put their money into trucks and crews instead of a website weren't necessarily leaving work on the table.
What's changed isn't that GBP got worse. It's that two new layers got stacked on top of it, and both of them read your website, not your Facebook wall.
- Google's local algorithm now weighs website signals (content depth, service-area pages, schema markup) more heavily in map-pack ranking than it did five years ago.
- AI answer engines (ChatGPT, Perplexity, Google's AI Overviews) build their "who does this well in this area" answers from crawlable web content, not social posts.
- Homeowners doing bigger-ticket jobs (roof replacement, whole-home remodel, panel upgrade) still click through to a site before they call, even when they found you in the map pack.
There's a second-order effect too: Google Business Profile itself now leans on your website to validate what's in the listing. A GBP with mismatched services, no linked site, or a link to a dead page reads as a weaker signal than one tied to a real, active site. The listing and the website used to be separate lanes. They're not anymore; the listing partly gets judged by what it points to.
Facebook and GBP alone can still generate calls for a low-ticket, high-frequency trade in a low-competition market. For most established contractors reading this, that's not the market they're in anymore.
What a website does that a Facebook page and GBP listing can't
Google Business Profile and a Facebook page each do one job well and stop there. GBP proves you exist at a location and shows reviews. Facebook proves you post occasionally and shows a comment section. Neither one can hold the depth of content that ranks for the actual services and towns you work.
A website is the only asset you own that can do all of the following at once:
- Rank for specific service-plus-city combinations ("metal roof replacement Winter Park") instead of just your brand name.
- Carry the schema markup (Service, FAQPage, LocalBusiness) that both Google and AI engines read to understand what you actually do and where.
- Show proof: real job photos, warranty terms, licensing, financing, the stuff a homeowner checks before they let a crew on their roof.
- Load and convert on its own terms. GBP and Facebook are rented land. The platform can change the rules on you overnight; a site you own can't be de-ranked by a policy update to someone else's app.
There's also a trust gap that's easy to underestimate. A homeowner who finds you in the map pack and then finds no website, or a template site that hasn't been touched since 2019, reads that as a signal about how the business is run. Fair or not, no website now reads closer to no truck than it does to old school. That perception moved fast over the last few years and it isn't moving back.
Ownership matters more than it gets credit for, too. A Facebook algorithm change, a GBP suspension over a flagged review, a platform policy update: any of those can wipe out your visibility overnight and there's no appeal process that moves fast enough to save the week. A website you own isn't immune to every risk, but it isn't rented from a platform that has zero obligation to keep your listing intact.
None of this means GBP and Facebook stop mattering. They're still part of the stack. They just can't be the whole stack anymore, and they were never built to carry the AI-search layer covered in the next section.
The part most contractors haven't caught up on: AI search
This is the piece that's actually new, and it's the reason the old "I don't need a website" argument expired. When someone asks ChatGPT or Google's AI Overview "who's a good roofer in [city]" or "how much does a bathroom remodel cost in [city]," the answer engine is synthesizing that response from crawlable web pages. It needs a source to cite. A Facebook wall and a bare GBP listing don't give it one.
Contractors with a real website, built with clear service pages, FAQ content, and proper schema, are showing up inside these AI-generated answers already. Contractors without one simply aren't part of the conversation. There's no map-pack equivalent for AI answers yet where a strong review count alone gets you cited. The engine is reading pages, not stars.
This is the single biggest reason "I don't need a website" doesn't hold up in 2026 even if it held up in 2016. It's not just a ranking factor shift inside Google. It's an entirely new discovery layer that has zero use for a business with no crawlable content.
What actually earns an AI citation:
- Dedicated pages per service and per service area, not one page trying to cover everything.
- FAQ content that answers the specific questions homeowners type into a chat window, in plain language.
- Schema markup that tells the engine explicitly what you do, where, and how you're structured as a business.
- Content that's actually true and specific to your trade and your town, not boilerplate swapped in with a find-and-replace on the city name.
This layer is also moving faster than local SEO ever did. Map-pack ranking shifts over months. AI answer behavior can shift with a single model update, and right now there's no established playbook the way there is for classic SEO. The contractors building this content depth now are establishing a citation history while it's still relatively uncrowded. That window doesn't stay open forever once every contractor in a market catches on.
None of that lives on a Facebook page. All of it lives on a website, and only a website built with this layer in mind from the start.
What kind of website actually matters here (and what doesn't)
Not every website earns its keep. A five-page template site with a stock photo of a hard hat and a contact form isn't going to rank, and it isn't going to get cited by an AI answer engine either. The gap between "has a website" and "has a website that produces work" is where most of the wasted spend in this industry happens.
What tends to actually move the needle:
| Does the job | Doesn't do much |
|---|---|
| Dedicated pages per core service, built out with real depth | One "Services" page listing everything in a bullet list |
| Service-area pages for every town you actually work, not just HQ | A single city name in the footer |
| Schema markup (Service, FAQPage, LocalBusiness) wired in on every page | No structured data at all, or a generic plugin default |
| Real photos of your actual jobs, licensing, and warranty terms | Stock photography and vague "quality craftsmanship" copy |
| Load time under 2 seconds on mobile | A bloated builder-platform site that takes 6+ seconds to load on a job-site 4G connection |
The platform matters less than the structure. A hand-coded static site, a well-configured WordPress build, or a mainstream site builder can all work, provided the service-page depth, the schema, and the load speed are actually there. What kills results isn't the platform choice, it's a thin site that never got past the template stage.
Volume matters here too, and it surprises most contractors. A site that genuinely covers a trade in a market, roofing alone, across every service line and every town in the service area, typically runs 94-plus cluster pages once it's built out properly. That's not padding. That's one page per real search a homeowner actually types. A six-page site simply isn't structured to compete against that, no matter how nice it looks.
This is also where a lot of contractors get burned by a $99/month "website included" offer bundled into a lead-gen platform. Those sites are usually thin by design (they're meant to funnel you back to the platform's own paid leads, not to independently rank you). Worth knowing before you sign anything long-term.
Does the answer change by trade or by how big the job is?
Mostly no, but the urgency shifts. The core answer (yes, you need one) holds across roofing, HVAC, plumbing, electrical, remodeling, landscaping, and the rest of the trades. What changes is how much a missing website costs you in a given month, and that's mostly a function of ticket size and how much research the homeowner does before calling.
- High-ticket, considered-purchase trades (roofing, remodeling, whole-home HVAC replacement, window/door replacement): homeowners research these hard before they call anyone. They'll check your site, your reviews, and now increasingly ask an AI assistant to shortlist options. No website here means getting filtered out before the phone even rings.
- Emergency and urgent-need trades (plumbing leaks, electrical failures, storm damage, HVAC no-cool/no-heat calls): the map pack and GBP carry more of the weight because the homeowner is moving fast and calling whoever shows up first with good reviews. A website still matters for the calls that aren't a five-alarm emergency, and for building the review and content base that keeps you visible in the map pack at all.
- Recurring and lower-ticket trades (lawn care, gutter cleaning, pest control): word of mouth and GBP carry more weight day to day, but a website still matters for capturing the commercial or higher-ticket jobs that come with a bigger sales cycle.
Company size and years in business change the calculation too, though maybe not in the direction you'd expect. An established contractor with 10 or 15 years in a market has more to lose from staying invisible, not less. A newer competitor with a well-built site and a thinner track record can out-rank a veteran shop that's coasting on reputation alone, simply because the algorithm and the AI engine have nothing from the veteran shop to read. Longevity helps once someone's already checking you out. It does nothing to get you found in the first place.
Across all of them, the AI-search layer covered above doesn't discriminate by trade. Whether someone's asking an AI assistant for a roofer, a plumber, or a landscaper, the engine is pulling from the same kind of source: real, structured, crawlable content. No trade is exempt from needing that. Some just feel the absence of it faster than others.
How to tell if your current site (or lack of one) is actually costing you work
Most contractors don't have a clean way to measure what a missing or weak website is costing them, because the lost calls never show up as lost calls. They show up as a competitor's name on a homeowner's counter instead of yours. A few honest checks can tell you where you stand without guesswork.
- Search your top 3 services plus your city, in an incognito window. If you don't show up on page one of the map pack or organic results, you're losing that search to whoever does.
- Ask ChatGPT or a Google AI Overview a question a homeowner would actually ask ("best [trade] in [city]" or "how much does [service] cost in [city]"). If your business isn't named or your site isn't cited as a source, you're absent from a channel that's only growing.
- Load your own site on a phone with the WiFi off. If it takes more than a couple seconds or looks broken on mobile, that's homeowners bouncing before they ever read your pitch.
- Count your actual service pages. One page that lists six services in a paragraph isn't six chances to rank. It's one thin chance.
- Check whether your site has any structured data at all. View source and search for "schema.org." If it's not there, neither Google nor an AI engine has a clean way to understand your business.
Run these checks against two or three of your actual local competitors too, not just yourself. It's the fastest way to see whether the gap is a you problem or an industry-wide problem in your market. In a lot of trades, half the competitors will fail the same checks, which means the opportunity isn't just to catch up, it's to be the one contractor in the area an AI engine and the map pack can actually cite with confidence.
If more than one or two of these come back bad, the honest read isn't "I should eventually get a website." It's that you're actively losing jobs to competitors who show up in the same three checks right now, today, while yours doesn't. A free audit will run these same checks for you and hand back exactly where the gaps sit, instead of leaving you to guess.