Quick answer: the marketing strategy a home service business needs is not a channel. It is three things held together: a brand specific enough to be chosen, coverage of the questions your customers actually ask — Learn, Compare and Act, not a page per service and city — and a website that can convert the moment an answer engine points someone at you. Owners keep shopping for the fourth thing. There isn't one.
A note on what this article is. On August 3, 2026 we published a 40-minute conversation with Lori of Eagle Pipe Heating & Air, Emily Fleniken (co-founder of Lemon Seed Marketing), and Jazmin Ramirez (Executive Director of Growth at CI Web Group). This article is not a transcript or a summary of that conversation — we have not transcribed it, so you will find no quotes from it here. It is our own article on the question the episode's title asks, published to go with the episode. The episode is embedded below; watch it for the story in their words.
Watch the episode
Lori started her HVAC and plumbing company in a barn in 2015 with zero trades experience — and she's still finding ways to be first. In this episode she sits down with Emily and Jazmin to talk growth, resilience, and why she said yes to launching one of the first AI-powered Hydra websites in the industry.
Open this episode on YouTube — about 40 minutes. One piece of honest housekeeping while you're here: Eagle Pipe's Hydra build is in preparation, not live. eaglepipemechanical.com still runs on Webflow today, and we publish no performance claims for a site that hasn't launched. You can see how we keep that line on Eagle Pipe's launch story page.
The strategy is a system, not a channel
Ask ten contractors what their marketing strategy is and you will get ten channel names. LSAs. Google Ads. SEO. Direct mail. A new website. Each answer is a purchase, not a strategy — which is why the same owner can buy all five and still describe demand as unpredictable.
A system is different. In a system, each part makes the next part cheaper. Brand makes ads cheaper because people already recognize the name. Content makes the brand credible because the company demonstrably knows things. A site built to convert makes both worth paying for, because it turns attention into a booked call. Take any one out and the others get more expensive.
That's the whole reason we co-present with a branding agency instead of pretending we do everything. Lemon Seed Marketing is a CI Web Group strategic partner because the brand layer is real work and it belongs to specialists. Our layer is the platform and the coverage. Neither one produces a strategy alone.
Start with the question chain, not the service list
Here is the part most home service marketing gets structurally wrong, and it has nothing to do with effort.
A homeowner used to run one search, look at ten links, and choose. Now they ask a question, get an answer, and ask again. A single repair decision usually takes six to ten questions before anyone picks up a phone: why is my AC freezing up, then can I fix it myself, then is it worth repairing a 12-year-old unit, then what does that cost, then who near me does this, then are they any good.
Now look at what the standard contractor site answers. Twelve services times eight cities is ninety-six pages, and it feels thorough — but "AC repair in Poulsbo" answers exactly one question in that chain: who does this here. It's the last question. The five before it, the ones asked far more often and far earlier, get nothing.
That is not a ranking problem, and optimizing those ninety-six pages harder will not fix it. It is a coverage problem. The strategy that follows from it is: write down the real sentences your customers say, in order, for your most profitable job — then find out how many of them your website answers on a page of its own. For most contractors the honest count is one.
Learn, Compare, Act — the three kinds of answers
We organize educational content on an intent ladder, because customer questions come in three distinct kinds and each needs its own page.
- Learn — how it works, what's wrong, the basics someone searches before they're ready to call anyone. Why a coil freezes. What a heat pump actually does. Why a water heater is making that noise. This is where the volume is, and it's the cheapest layer to write well.
- Compare — decision-stage questions. Repair or replace. Heat pump versus furnace. Tank versus tankless. Brand against brand. These are the questions where an answer engine is actively hunting for someone who has laid out the tradeoffs plainly and honestly.
- Act — cost guides, financing, second opinions, maintenance plans. What a replacement runs in this market and what the monthly payment looks like. This is where a cited answer converts, because the person has already decided to do something.
Multiply that out — say forty real questions per service line across the three levels, times your service lines, plus the local variants that matter — and you are not writing ninety-six pages. You are building several hundred answers, each one a separate chance to be the source that gets named. That is why we ship Learning Center content in packs rather than one page at a time. Coverage is the product.
Why this changed: answer engines cite, they don't rank
On August 3–4, 2026 we ran the same HVAC problem past ChatGPT, Gemini and Perplexity two ways, and the results are worth knowing before you set a budget.
Asked nationally — "why is my air conditioner freezing up and how do I fix it?" — the assistants pulled from 56 citations (Gemini) and 17 (Perplexity), and every single source was a national publisher, manufacturer, or large directory. No local contractor appeared anywhere. Asked locally — same problem, city attached — the citation count collapsed to 18, 20 and 6, and the answers named local companies by name.
Two conclusions to act on. Generic national questions are not winnable for a local contractor, so don't spend there. The same question localized is winnable, because the pool shrinks by two thirds and the competitive set becomes companies your size. The full test, including the parts that didn't flatter us, is written up in how consumers search with AI and how to get cited.
The most useful thing we found in that test was why one local company got recommended: Perplexity justified it by noting the company "says frozen coils are a common cause of AC problems." Not a directory listing. Not a review count. They were cited because they had published content on the thing that was asked, so they were quotable on it. They earned a local recommendation with an educational page.
A citation is only worth what your site does with it
Coverage gets you named. It doesn't get you paid. A citation delivers someone mid-decision, holding one specific question, onto one page — and if that page is a wall of prose ending in "contact us," you did the work and your competitor got the job.
The answer page has to carry the next step inside it, matched to the intent that brought the visitor: a Learn visitor is ready to understand and then compare, an Act visitor is ready to book and shouldn't have to hunt for how. Estimate tools, financing calculators, second-opinion offers, booking paths — placed on the page that earned the attention, not buried on a contact page two clicks away.
One of our clients watched both halves of that system land on the same day. Josh Osuna, Service Manager at Eco Electric, Plumbing, Heating & Air in Seattle, texted his leadership group:
"Both of our HVAC sales today were from our website. The first from microsites and the second one was recommended via AI search which cited our website and ranked us at the top."
Josh Osuna, Service Manager · Eco Electric, Plumbing, Heating & Air · Jul 22, 2026
That is one company, one day, and we're not going to dress it up as a benchmark. What makes it worth reading is the mechanism it confirms: an answer engine cited the site, the visit arrived, and the page finished the job. The full record is on Eco's launch story.
Brand is what makes the answer choosable
Here's where coverage stops being enough. Once an assistant has answered the homeowner's technical questions, the last question in the chain is who should I call — and that one is not answered by page content. It's answered at the entity level: your Business Profile, your structured data, your reviews, whether your facts are consistent everywhere they appear, and whether your name means anything specific.
Most home service brands are interchangeable on purpose. Family owned. Honest technicians. Serving the area since. All true, none distinguishing — and when three companies say the same three things, the only remaining variable is price. Sameness makes every lead more expensive, in paid, in organic, and now in AI answers.
The through-line is worth memorizing: page-level content earns question citations; entity-level signals earn the "who should I call" recommendation. You need both. A great answer library behind an inconsistent, indistinct brand gets cited on the how-to and skipped on the hire.
Being first is a habit, not a launch
The line in the episode description that stays with us is that Lori started in a barn in 2015 with no trades experience and is still finding ways to be first. Not was first once. Still finding ways.
That is a strategic posture, and it's available to a company of any size — arguably more available to a small one. It's what makes an owner with no industry background willing to try structured pricing, or a new content architecture, or a website platform built for how customers search now, before the rest of her market gets around to it. Being early is one of the few advantages that doesn't require capital.
What we can't promise you
- Citation is not instant. New content has to be published, crawled and established. In our own August tests, a site that had gone live the day before was not cited at all. Anyone promising fast AI citations is guessing.
- Our test was two questions on two days in specific markets. Your citation sets will differ by city and trade. Run the tests for your own market rather than trusting ours.
- Nobody can attribute revenue to AI citations today with the precision we attribute paid search. What is measurable is whether you are cited — track that now, because it's the input.
- Eagle Pipe's Hydra site is not live yet, so there are no Eagle Pipe results in this article. When it launches and gets measured, the numbers will be published on its launch story page and not before.
What to do this week
- Write down the question chain for your most profitable job — the actual sentences a customer says, first symptom to hiring. Usually six to ten.
- Ask an assistant each one, once nationally and once with your city attached. Note whether you're cited and which sources are. That list is your real competitive set.
- Count your coverage. How many of those questions does your site answer on a page of its own?
- Fill the Learn gap first. Largest, cheapest, and demonstrably where a local contractor wins a citation.
- Fix the entity layer in parallel — Business Profile, structured data, consistent facts, reviews.
- Say one specific true thing about your company that your two closest competitors cannot copy, and put it everywhere. If you can't find one, that's the project.
None of this is a channel you can buy on Monday and evaluate on Friday. That's the honest reason so many owners keep buying channels instead: a system takes a season to compound and can't be judged in a week. The companies that get there anyway tend to look like Lori's — willing to be early, willing to be specific, and unwilling to wait until the market makes it obvious.
Related resources
- How consumers search with AI — and how to get cited: the full August 2026 citation test, numbers and all.
- AI search guides and how generative search reads a site.
- Content intelligence: how the Learn / Compare / Act library gets built.
- Hydra OS: the platform side — structure, speed, and conversion components on the answer page.
- Featured customers: launch stories with the before state on the record.
- The Catalyst for the Trades podcast: more conversations with owners like Lori.