2026-08-05
Nine Questions to Ask Before Hiring an HVAC Marketing Agency
The questions that separate agencies who measure from agencies who decorate, what a good answer sounds like, and what the dodges usually mean.
Short answer: ask for specifics they would need to have already looked to answer. Most sales calls are spent on what an agency will do. These nine questions are about what they have already checked and what happens if it does not work — which is where the differences actually show.
We are an agency. Read these knowing that. They are still the right questions, and we have included what we would say to each.
1. “What are my Core Web Vitals right now?”
Why it works: it takes two minutes to check and requires them to have looked at your site before the call.
A good answer names your LCP, INP, and CLS, and says whether you pass at the 75th percentile of real visitors. Google’s thresholds are 2.5s, 200ms, and 0.1.
A bad answer is “we’ll do a full audit once you’re onboard.” They are quoting on appearance.
2. “Which specific towns will you build pages for, and in what order?”
Why it works: the Map Pack ranks each town separately, so a real plan is a named, ordered list.
A good answer names towns and explains the ordering — usually by population, ticket size, or how winnable they are.
A bad answer is “we’ll target your whole service area.” That is not a plan.
3. “Do I own the website, the domain, and the Google Business Profile if I leave?”
Why it works: it surfaces lock-in before you sign rather than at renewal.
A good answer is yes to all three, without qualification.
A bad answer involves a proprietary platform you cannot take with you. That is a legitimate business model — but it should be a decision you make knowingly.
4. “What happens in month one, specifically?”
Why it works: good early work is unglamorous and specific.
A good answer sounds boring: technical audit, profile cleanup, schema implementation, redirect mapping, review system setup.
A bad answer is abstract — “strategy development,” “brand alignment,” “content ideation.”
5. “How do you report on AI search visibility?”
Why it works: 45% of consumers used an AI tool to find a local business in the past year, up from 6% the year before. An agency with no answer is measuring a shrinking share of how customers find you.
A good answer describes actually checking — running the queries a homeowner would use across ChatGPT, Perplexity, and Google’s AI Overviews, and tracking whether you get named.
A bad answer is a promise to “optimize for AI” with no method attached. If they lead with llms.txt, ask what evidence they have that anything reads it. Google has said it does not use the file.
6. “Do you work with another HVAC company in my service area?”
Why it works: two clients competing for the same map results means someone is being optimized against.
A good answer is no, with a stated exclusivity policy.
A bad answer is “our teams are kept separate.” The teams are not the issue; the rankings are.
7. “What’s your reporting cadence, and can I see a real example?”
Why it works: you learn what they consider progress.
A good answer comes with a redacted real report showing leads, booked jobs, and where they came from.
A bad answer is a dashboard of impressions, rankings, and “engagement.” Impressions do not pay technicians.
8. “How long before I see results, and what would tell us it isn’t working?”
Why it works: the second half is the real question. Anyone can promise. Few will pre-commit to a failure condition.
A good answer gives an honest window — site speed immediately, map results and AI citations typically three to six months of continuous work — and names what they would expect to see at 90 days that would indicate it is on track.
A bad answer promises page one in 30 days. Nobody controls that.
9. “What does it cost, and what’s in the contract?”
Why it works: obvious, and still the one contractors skip.
A good answer is a specific number, month-to-month or with a stated term, and clarity on whether ad spend is included. Typical HVAC SEO runs $1,500–$3,500 a month for a single location, more in competitive metros.
A bad answer hides pricing until a second call, or bundles ad spend into the retainer so you cannot see what you are paying for management versus media.
The pattern
Every question above asks for something checkable. Agencies who measure answer them quickly, because they already looked. Agencies who decorate deflect to process.
Three that matter most if the call is short: Core Web Vitals, which towns in what order, and what would tell us it isn’t working.
What we would say
For fairness, ours: pricing is published at $3,500 for the build and $1,250 a month, month to month. You own the domain, the site, and the profile from day one. One HVAC contractor per service area. Month one is audit, schema, profile, redirects — nothing glamorous. We would expect speed to change immediately and map and AI visibility to take three to six months, and if nothing has moved at 90 days on a site we rebuilt, that is on us to explain.
Ask us the same nine. Ask everyone the same nine. The comparison is the point.
Sources: Google Search Central Core Web Vitals thresholds; BrightLocal 2026 Local Consumer Review Survey; Google Search Central AI-optimization guidance, June 2026; published 2026 HVAC SEO pricing analyses.