Why review count decides who gets the call, not just who ranks
When a homeowner's AC dies at 4pm on a July Friday, they don't read ten websites. They pull up the phone, search "AC repair near me," and look at the Map Pack: three listings, star ratings, review counts. That's the whole decision. A 4.9 rating with 340 reviews next to a 4.6 with 40 reviews wins almost every time, even if the second company is better at the actual work. Reviews are the fastest trust signal a scared, hot, impatient customer can read in four seconds.
This matters more in HVAC than in almost any other trade because the lead is often an emergency. Roofing leads get shopped over days. HVAC leads in a heat wave get shopped over minutes. The company with the review count wins the click, wins the call, and the other two companies never get considered.
Google's local ranking algorithm also treats review signals (count, average rating, and how recently reviews arrived) as part of the relevance and prominence factors that decide Map Pack placement, alongside proximity and how complete the Google Business Profile is. A stalled review count doesn't just look worse to customers, it actively costs rank.
The franchise down the road usually wins this fight by volume of trucks and years in the market, not by a superior review process. That's the opening: an independent shop that systemizes the ask can out-review a franchise with three times the truck count, because most franchise locations still leave review requests to whichever tech remembers.
- Map Pack clicks concentrate on the top 3 listings. Being 4th means being invisible on mobile.
- Review recency signals an active, still-operating business. A profile with its last review from 14 months ago reads as stale to both Google and the customer.
- Star rating differences under 0.3 rarely change customer behavior. Review count differences of 100+ almost always do.
The moment that matters: asking at job close, not job forget
Every HVAC company means to ask for reviews. Fewer than half have a system that survives a busy Tuesday. The fix isn't a better speech, it's moving the ask out of the technician's memory and into the workflow.
The highest-converting moment is the payment moment: the invoice is signed, the system is running, the customer is relieved. That's when a text with a direct review link goes out, ideally triggered from the dispatch or invoicing software the tech already closes the ticket in, not a separate app they have to remember to open.
A direct link matters more than people think. "Please leave us a review" sent with a link to search "[Company] Google reviews" loses most of the audience to friction: they have to search, find the right listing, tap through, then find where to write a review. A direct short link (Google gives every Business Profile a shareable review link) drops a customer straight into the review box. That single change routinely doubles completion rate on the same ask.
| Ask method | Typical friction | Best use |
|---|---|---|
| Tech asks verbally on-site | Customer forgets before they open the app | Warm-up only, follow with text |
| Text with direct review link, sent within 1 hour | Lowest friction, one tap to review box | Primary channel for every ticket |
| Email sent next-day batch | Buried in inbox, low open rate | Backup only, not primary |
| QR code on invoice or truck | Customer must be prompted to scan | Passive assist, not a substitute for the ask |
The system, not the script, is what makes this repeatable across every technician on the truck fleet, including the ones who hate asking for anything.
Building the ask into your seasonal service calendar
HVAC demand isn't steady. It spikes hard during heat waves and cold snaps, then goes quiet in the shoulder months. That rhythm should shape when and how you ask for reviews, because a review push during a heat-wave week competes with dispatch chaos, and a review push in a slow month has room to actually get attention from both techs and office staff.
During peak season (the two or three weeks a heat wave or cold snap hits), keep the ask mechanical and automatic: the text goes out the moment the invoice closes, no manual step for a tech juggling six calls a day. This is the highest-volume window for reviews because it's the highest-volume window for jobs, period. Don't let the chaos be an excuse to skip the ask. It's the opposite: peak season is when you bank the review count that carries you through the quiet months.
During shoulder season, when call volume is lower, that's the window to go back and follow up with customers who didn't respond to the first text, and to specifically target maintenance-plan members and past system-replacement customers for reviews, since those are typically your highest-satisfaction relationships and the best testimonials for a $9,000+ purchase decision.
Replacement customers deserve a different cadence than a $180 tune-up call. A system replacement is a major purchase. Ask for the review about a week after install, once the system has run through at least one cooling or heating cycle and the customer has felt the difference, not the same day the crew leaves. That review reads as more credible to future replacement shoppers because it reflects the outcome, not just the sales experience.
- Peak season: automatic text-on-invoice-close, zero manual steps.
- Shoulder season: follow-up pass on non-responders, maintenance-plan member push.
- Replacement jobs: ask about a week post-install, after the system has proven itself.
- Maintenance-plan renewals: a natural, low-pressure annual ask point.
What to do when a review comes back negative
Asking more often means more reviews, and more reviews means occasionally a bad one. That's not a reason to ask less. It's a reason to have a response plan before it happens.
The instinct to argue in the reply box is the wrong instinct. A public back-and-forth about who said what on a service call reads badly to every future customer who scrolls past it, regardless of who was right. The goal of a public reply isn't to win the argument, it's to show the next reader that your company responds like a professional when something goes sideways.
The better process has two tracks. First, reply publicly within a day or two: acknowledge the specific issue without being defensive, and offer a phone number or direct contact to resolve it offline. Keep it short. Second, actually call the customer. Half of negative HVAC reviews trace back to a miscommunication (a diagnostic fee the customer didn't expect, a callback that took longer than promised) that a five-minute phone call can often resolve well enough that the customer either updates the review or at least stops mentioning it to neighbors.
Route the risk of a bad review out of the equation earlier when you can. If a job went sideways (a repeat callback, a part on backorder, a scheduling miss) don't send the automatic review-request text for that ticket. Route it to a manager for a personal check-in call instead. This single filter (never auto-ask on a known rough job) prevents most of the avoidable one-star reviews before they're ever written.
What you should never do: offer a discount, refund, or gift card in exchange for changing or removing a review. That violates Google's policies on review manipulation and can get reviews or the whole profile flagged for removal.
Google's rules on incentivized and gated reviews
Google's policy is direct: businesses may not offer money, discounts, free services, or any other incentive in exchange for a review, and businesses may not filter customers by satisfaction before directing them to leave a review (sending only happy customers to Google while quietly redirecting unhappy ones elsewhere). Both practices are review gating, and both can get reviews removed or trigger a profile suspension.
What's allowed, and what actually works better anyway: ask every customer, without exception, and make it easy. You can absolutely time the ask to the moment satisfaction is highest (right after the system is running and the invoice is paid), that's just good timing, not gating. The line is whether you're asking everyone or only the customers you've pre-screened as happy.
A widely used and compliant middle step is a private feedback form that asks "How did we do?" before offering the public review link, as long as every customer who wants to leave a public Google review is still able to, and you aren't suppressing negative feedback from reaching Google. Many shops use this as a triage tool: a 1-3 star private response routes to a manager call, a 4-5 star response gets the direct Google link. That's legitimate customer service routing, not manipulation, provided the door to Google stays open for anyone who wants to walk through it.
Fake reviews (from employees, family members, or purchased review services) are the fastest way to get a whole profile wiped and possibly suspended. Google's spam detection has gotten materially better at flagging clusters of reviews from accounts with no other activity, reviews posted in unnatural bursts, or reviewers with no history in your service area. It isn't worth the risk to a profile that took years to build.
Which techs to trust with the ask, and how to track who's asking
A review system dies the fastest way any shop process dies: it depends on one enthusiastic tech, and everyone else quietly skips it. The fix isn't a pep talk at the Monday meeting, it's removing the ask from the tech's job description entirely wherever possible.
If your dispatch or field-service software (many HVAC shops already run something like ServiceTitan, Housecall Pro, or a similar platform) can trigger a review-request text automatically the moment a ticket is marked complete and invoiced, that's the highest-reliability setup, because it doesn't matter if the tech is having a rough day or forgot. The request goes out regardless.
Where that automation isn't in place yet, the next-best fix is a simple tracked list: every closed ticket gets a manual send, and someone in the office (not the tech) owns confirming it happened, the same way you'd track that an invoice went out. Making it someone's explicit job, rather than everyone's vague responsibility, is usually the difference between a review system that runs for a month and one that runs for years.
It's worth resisting the urge to gamify this with public tech-by-tech review leaderboards tied to bonuses. That edges toward incentivizing the review itself rather than the service behind it, and it can pressure techs into asking in ways that read as pushy to the customer. Track completion of the ask (did the text go out), not the review outcome, if you want to measure tech performance without creating pressure that shows up in the review's tone.
- Automated trigger from dispatch/invoicing software is the most durable setup, no memory required.
- Manual fallback needs an explicit owner in the office, tracked like any other closing task.
- Measure ask completion rate, not review count per tech, to avoid incentivizing the wrong behavior.
What actually moves the needle: a realistic timeline
Review growth in HVAC follows the same arc as most local-service review building: slow at first, then compounding once the ask is systemized and techs stop needing reminders. It isn't instant, and any promise that it is should be a red flag.
In the first month or two of a real system (automatic text at invoice close, every ticket, every tech), most shops see the review rate per completed job climb from whatever it was (often under 5% of jobs) toward 15-25% of completed jobs, simply because the ask stopped depending on memory. That's the biggest single jump, and it happens fast because it's a process fix, not a demand-generation problem.
Map Pack movement lags behind review count by design. Google needs to see sustained review velocity, not a one-time burst, before it re-weights local ranking. Expect the ranking benefit to show up over the following few months as reviews keep arriving at a steady clip through a full season, not a single push.
The compounding part is the useful part: a shop that keeps the automatic ask running through both a summer cooling season and a winter heating season builds a review count and recency pattern competitors can't easily match without doing the same disciplined thing for the same length of time. This is one of the few local-marketing levers where being early and consistent matters more than being clever.
- Weeks 1-4: system goes live, ask rate per completed job climbs fast.
- Months 2-4: review count compounds, response-to-negative-review process gets tested and refined.
- Months 4+: Map Pack position typically starts reflecting the sustained review velocity, alongside the rest of your Google Business Profile signals.