Google Reviews for Local Service Contractors: Plumbers, Electricians, HVAC & Roofers

There is a difference between a business that wants good reviews and a business that needs them. A restaurant wants reviews because comparison-shopping happens over lunch. A dental office wants them because people schedule non-emergency care weeks in advance and do compare three or four providers before booking.

A plumbing, HVAC, electrical, or roofing contractor operates on a completely different clock. When a water heater dies at 6:40 p.m. on a Tuesday, the homeowner has three options and roughly ninety seconds of patience to evaluate them. They will not read a services page. They will not study your certifications page. They will glance at the Local 3-Pack, look for a star rating, scan one or two review snippets for words like honest, fast, fixed it right the first time, and tap the phone number of whichever company looks safest.

That is the entire game. Everything in this article is built around winning that ninety-second evaluation, and doing it in a way that survives an FTC inquiry, a Google policy review, and a former employee who decides to talk.

1. The High-Stakes Nature of Trade Reviews: The 90-Second Decision Window

Understanding the decision window changes how you price the effort of review collection. In most local service categories, a bad review is an inconvenience — it nudges a few prospects away. In emergency-adjacent trades, one bad review visible at the moment of need can cost you hundreds of thousands of dollars in annual revenue, because the homeowner does not retry. They call the next result.

Here is roughly how fast the decision happens by trade, based on how often the category is researched in an urgent state:

Note the wide variance in call value. A plumbing call is a $500–$1,200 job with a 20–40% close rate on a single inbound call. A roof replacement is a $6,000–$15,000 job that may take four weeks of sales cycles but lands once. If you are comparing "what is one extra five-star review worth" across trades, you get wildly different answers — and that answer should drive how aggressively you pursue reviews in that trade.

What the Local 3-Pack actually rewards

The Local 3-Pack (or 3-Pack: the map pack showing three businesses per query) is not a ranking of "best." It is Google's attempt to satisfy a very specific user intent: I need this service near me, now, and I want to not get burned. Three factors dominate it:

1. Proximity — distance from the searcher. This one is largely fixed; you cannot change it except by creating real service-area presence.
2. Prominence — how many independent signals vouch for the business. Reviews are the loudest of these signals, which is why volume and quality compound.
3. Relevance — service-match signals, including review text relevance and category/service declarations.

Reviews hit all three. A service-area business in a metro like Denver that reviews mention "furnace" and "boiler" and "condo townhome" is matching on relevance. A business with 280 reviews across a decade of job types is matching on prominence. And a business whose reviews are geographically clustered in the neighborhoods it actually serves is matching on proximity in Google's local ranking model.

The compounding math most contractors miss: A business going from 38 to 280 reviews does not get "7.4x more reviews." It gets a different competitive category. Below roughly 40 reviews, you are invisible in the 3-Pack for anything but the cheapest, most generic queries — "plumber near me." Above 150 reviews, with recent velocity and a 4.7+ average, you begin appearing in the money queries — "emergency plumber in Lakewood," "AC repair same day." Those mid-tail queries convert at multiples of the head term because the searcher already has intent and geography narrowed. The rank change compounds with the review count, which is why reviewvelocity matters as much as review total.

The four things a homeowner is actually deciding

Strip away the star rating and what the ninety-second reader is really answering is four questions. Every tactic in this article maps to one of them.

2. The Job-Site Collection Workflow: Engineering the Handover Moment

Most contractors attempt review collection by email. You send a link after the invoice goes out, the homeowner clicks "pay invoice" and never opens the review request, or opens it three days later when they've forgotten the technician's name. Response rates on that model typically land between 2% and 8%.

The fix is to stop treating review collection as a marketing email and start treating it as a job-completion handover ritual — a defined moment at the end of the visit where the technician is handed a small physical or digital artifact, says five scripted words, and moves on. Homeowners complete reviews at a rate of 30–60% in this model, because the request arrives while the experience is still vivid.

The four-part handover sequence

This is the exact sequence. It takes 45 to 90 seconds and works in residential and commercial settings alike.

1. Close-out statement (the tech says it). "So the new water heater is installed, it's been pressure tested, and I've got you set up for a six-month flush." Then: "One quick thing before I head out — I've got a card here with my name and the job number. If anything's not right in the next few days, text that number and I'll come right back out."
2. Hand over the artifact (physical). A magnetic fridge card, a door-hanger sticker, or a job-folder insert with the QR code printed large and the URL in plain type beneath it.
3. The ask (one sentence, non-negotiable wording). "If you were happy with the work, I'd appreciate a quick Google review — it takes about thirty seconds and it helps other families find someone they can trust."
4. The escape hatch (this is what makes the ask comfortable). "And if anything isn't right, please call me first instead — I'd rather fix it than have you write about it."

Step four matters more than people expect. A request that feels like a trap — as if the technician is trying to prevent a bad review — triggers defensiveness. A request that frames the review as a favor for the neighborhood gets cooperation. The homeowner with a legitimate complaint almost always calls you first, which is exactly the outcome you want, because a phone call is fixable and a review is not.

Leave-behind design: what to print, exactly

The card is doing three jobs: it carries the QR code, it proves the technician was real (name + license number + company branding), and it survives long enough to be used. A business card is thrown away at the door. A card that lives on the fridge for three months is used.

QR code placement rules that actually matter

• Minimum size 0.9 inches square for the printed code; smaller and older phones plus car windshields fail to focus. Go 1.25 inches if it fits the format.
• High contrast, quiet zone. Dark code on white, with a white margin of at least 0.25 inches around the edge. A QR code dropped onto a busy logo or a photo of a job site can drop scan rates by half.
• Never print a QR code on a magnet that will sit in a refrigerator covered in magnets. If you are doing fridge magnets, leave the lower-left quadrant of the card completely clear of other artwork.
• Print the vanity URL beneath the code — something like yourcompany.com/reviews or review.yourcompany.com. Roughly one in twelve people who see a QR code will not scan it, but nearly all of those people will type the URL if they trust the brand. You are capturing that person by having the domain on the card.
• One code, one destination. Do not put a QR code on the magnet that goes to your homepage. It must deep-link straight into the review flow: the Google review link for the correct business profile, pre-filled if you use a Google review URL shortener that auto-selects your location.

SMS dispatch triggers and timing

The SMS is what converts a reluctant homeowner into a reviewer. Timing matters more than copy. The rules:

Copy for the first SMS — under 160 characters, one message, no links required (the card has the QR):

"Hi {{first_name}}, {{tech_name}} finished your {{job_type}} at {{address}} today. If everything looks good, a quick Google review means a lot: {{short_link}} — and if anything needs a touch-up, just reply here. — {{company}}"

Second ask (48 hours, no review):

"Hi {{first_name}}, still hoping you'll take 30 seconds to leave a Google review of the {{job_type}} — it helps us keep the doors open. Here's the link: {{short_link}}. Questions? Call {{phone}}."

Two things to avoid in SMS copy: multiple links (it triggers link filters on some carriers and reads as spam), and asking for a rating. "Rate us 5 stars" is the single most common trigger for a one-star "they asked for five stars" review, and those are disproportionately damaging because they are read as a trust violation by every future prospect.

3. The Photo-First Strategy: Reviews With Job-Site Photos Get Materially More Engagement

Across the trade-contractor accounts we track, reviews that include a photo from the job site accumulate roughly 3.2x more engagement in the Local 3-Pack — meaning more impressions, more click-through, and more calls — than text-only reviews at the same star rating. The mechanism is straightforward: Google surfaces photo-bearing reviews with visual thumbnails in the mobile map pack, and a thumbnail of a clean, finished install (a new water heater, an orderly electrical panel, a new shingle run) does the credibility work faster than any sentence of copy. It is the fastest possible proof of "this is a real company that does real work."

Two caveats worth stating plainly. First, this is a directional figure from our own sample, not a published Google statistic — Google does not disclose engagement multipliers. Treat it as a strong hypothesis to test in your own account, not a law. Second, the mechanism is not "photos are magic." It is "photos move the decision forward faster," and they only do that if the photo is genuinely good.

The photo types that drive the Local 3-Pack

The photo request, added to the handover script

Add one sentence to the talk track, and train every technician on exactly where to stand:

"Would you mind if I took a couple of photos of the install? I'll put them in your text thread along with your warranty — you'll have the paperwork for your records."

This works because it frames the photo as a benefit to the homeowner (documentation and warranty records) rather than as marketing material the contractor wants. Do not ask the homeowner to take the photos on their phone. You will get ceiling shots, blurry photos of a half-finished install, and a picture of the box the new unit came in. You will get none of them. The technician shoots it, on their own phone, before they pack up.

How to turn photos into reviews (without asking for reviews of photos)

Homeowners rarely attach photos to a review on their own — the upload friction in Google's review flow is real, and most people stop at the star rating. The move is to make the photo part of the thing they are already sending you: the payment, the completion photo album, the warranty document.

1. Technician's completion album. After the job, the tech sends the photos via your own SMS/email thread: "Your install is complete — here are 4 photos and your warranty certificate."
2. The homeowner reuses them. Because they are already in the thread and already on the homeowner's phone, forwarding them into the review upload takes seconds.
3. Review request includes the images. Many review-request platforms and schedulers support sending the completed-job photos inside the request itself. The homeowner sees their own job and their own job is the easiest thing in the world to write about.

Repurposing job photos beyond the review

The same photo library feeds three other assets at near-zero marginal cost: your Google Business Profile photo posts, your service-area page (city-specific landing pages with real local install photos — this is a serious ranking lever that most contractors ignore), and your proposal templates. A roofing company in Austin with 60 genuine local job photos distributed across its city pages is invisible to competitors doing the same thing, because almost nobody does it.

4. Handling Trade-Specific Negative Reviews

Negative reviews in the trades are not vague disappointments. They are specific, recurring, and highly predictable failure modes. Which means you can prepare responses in advance instead of writing them from scratch while angry at 9 p.m. on a Sunday.

First, the anatomy of a response that works. Every one of these follows the same four-move structure:

What you must not do: get defensive, dispute the facts publicly, blame the customer, mention that the job was "complex" or that the customer "didn't approve the scope" in a way that sounds like an excuse, or paste the same response on multiple reviews. Google's guidelines and consumer expectations both punish responses that argue. A response that makes other reviewers uncomfortable costs you more than the original review.

Script A: "They charged way more than I expected"

This is the single most common trade review category, and it is usually about expectation-setting, not about the number on the invoice. The homeowner did not get a quote, or got a verbal range, or got a quote and a discovery that revealed more work. The gap between expectation and invoice is what generates the complaint.

"Thanks for taking the time to write this, and I'm sorry the final number wasn't what you had in mind. Before we were on site, you'd seen the range we discussed on the phone, and the additional cost came from the corroded shutoff valve behind the old heater — which we only confirmed once the wall panel was off. That's on us for not flagging the possibility of it beforehand, and we've since changed our process to photograph and discuss any wall-access contingency before starting. Your invoice was for {{amount}}, which included {{specific scope}}. If you'd like me to walk through the line items, call me directly at {{phone}} and I'll take as long as you need. — {{owner_name}}, {{company}}"

Three things make this work. It acknowledges the gap without disputing the price. It explains the cause in a way that is verifiable (a corroded valve is a real thing, and it makes the homeowner feel their own home was the issue, not the company). And it names a process change — because "we've changed our process" is the single most persuasive sentence you can put in a trade review response, since it shows the complaint produced an outcome. If you have not actually changed the process, do not claim it. Do not manufacture process changes to win one review; it will not survive the next bad review and someone will notice the pattern.

Script B: "The technician was late / didn't show up"

Window-showing complaints are lethal because they are unfalsifiable from the outside and the reviewer will usually describe it as hours late. What a future customer needs to see is not a contradiction but a system.

"That's a fair complaint and I'm sorry it put you through it. Dispatch handles a lot of calls and there are days where a truck runs behind — on your date, the morning job overran by {{actual, honest detail}}. We've since moved to a tighter dispatch window: technicians text you a 20-minute arrival notice, and dispatch calls you the moment a truck is running more than 30 minutes behind rather than letting you wait. If you'd like the arrival window you were given on {{date}}, call me at {{phone}}. — {{owner_name}}, {{company}}"

Only include the real cause if you know it. If you do not, say "our morning schedule overran" without a specific fabricated incident. Do not write "we are sorry if you felt" — the conditional apology reads as a non-apology. And if the technician was late because the company oversold the schedule, say that. Prospects read the response not for the apology but for the honesty, and the honest version is more persuasive than the polished one.

Script C: "Hidden emergency / weekend surcharge"

This one requires the most nuance, because the reviewer is often factually correct — there was a surcharge, and it was not disclosed up front. Your response must not say the surcharge was unnecessary if it was disclosed in a text the reviewer missed. Instead, address the disclosure failure, which is the legitimate complaint.

"The after-hours fee is a real part of how we price emergency work — it covers the technician's overtime and the parts we carry on the truck — so I won't pretend it wasn't part of your invoice. What I will own is that the amount should have been confirmed before you authorized the work, and that is on us. We've changed our dispatch process so that the after-hours and any additional fees are stated in the estimate you approve by text before the technician starts. If you want to go back through the estimate you signed, call me at {{phone}}. — {{owner_name}}, {{company}}"

This response does something smart: it validates the fee without surrendering the fee, and it converts the complaint into evidence of a better system. Never offer a refund in a public response to a disputed fee — you will get a comment thread with two or three people whose "the same thing happened to me" comments will outrank your reply.

The 24-hour rule and the escalation ladder

• Respond within 24 hours of a negative review appearing. Same-day response is the standard to aim for; it reads as an active company, not a monitoring service.
• Take the conversation private fast. Every response should end with a phone number and an offer to talk. Reviews are public; resolutions should be private.
• Never ask the reviewer to take it down in exchange for a refund. That is a FTC problem (16 CFR Part 255 treats a review conditioned on a monetary incentive as a material connection that must be disclosed, and offers contingent on sentiment are the core prohibition) and a Google problem. A refund-plus-deletion request is the single most common trigger for a fake-review removal action against a business profile.
• Escalate by lifetime value, not by emotion. A one-star review on a $300 job from a customer you are never seeing again deserves a professional response and a system fix. A one-star review on a $120,000 commercial HVAC replacement from a property manager with 40 units deserves your owner, a phone call within the hour, and a written account of what happened.

Rule of thumb worth memorizing: The goal is not to make the review go away. The goal is that the next person reading the review understands the company handles problems well. A five-star response to a one-star review frequently converts skeptics; a defensive response to a one-star review confirms the reviewer's suspicion that this company is hard to deal with. Reviewers with negative experiences who get a genuinely good response frequently update their rating. Reviewers who get a legal-sounding response frequently do not.

5. Technician Incentive Programs: Reward the Behavior Without Violating Google's Rules

Technicians generate reviews because they are in the room and the homeowner likes them. But technicians also generate them because the company pays for it, and how you design that payment is where well-intentioned companies accidentally build a policy violation.

What Google prohibits

Google's Maps content and fake engagement policies prohibit a specific set of behaviors. The relevant ones for review programs:

The doctrine underneath all of it: a review must be an honest account, solicited without condition, and not purchased. What you are permitted to reward is the process — the behavior that produces reviews — not the reviews themselves.

Compliant incentive structures

Here are the models that hold up, with the specific mechanics that keep them on the right side of the line.

Model 1 — Pay for the review conversation, not the review

Reward the technician for a completed review ask, documented by a checkbox in the field software, regardless of whether a review is posted. The trigger is the ask, not the outcome.

• Structure: $10 per completed handover, verified by CRM field ("Review request delivered: Yes"). No tracking of the resulting review, no link to the tech's name, no payment condition on the review existing.
• Why it holds: The technician is paid for doing the job they are already paid for — asking a happy customer for a review. The review's existence and sentiment are entirely outside the compensation equation.
• The line: If the bonus is per-posted-review, you have crossed it. Same dollar amount, totally different legal posture. If a customer complains and you claw back the bonus, you have proven the bonus was contingent. Don't claw back.

Model 2 — Reward volume-based outcomes unrelated to review content

Tie bonuses to metrics the technician controls that correlate with reviews but are not reviews: jobs completed with full documentation, customer satisfaction survey scores, callback rate, warranty re-visit rate.

• Structure: Quarterly bonus tier on "jobs closed with 100% documentation" and "first-time-fix rate above 90%."
• Why it holds: No review content, no sentiment, no review existence is involved.
• Bonus practice: Add a non-review satisfaction metric — a one-question SMS survey to the homeowner ("Were we great? 1–5") that is never published and never linked to a review ask. This gives you technician-level quality signal without touching the public platform. Important: the survey must be sent to all customers, and you must not selectively invite only satisfied customers to leave Google reviews.

Model 3 — Give the customer something, ask for something unrelated

Give every customer a thank-you gift (a maintenance check, a $25 credit, a 6-month tune-up). Ask for a review separately, at a different moment, with no stated connection.

• Why it holds: The gift is compensation for patronage, not a quid pro quo for a review. It is also the single most effective at-home conversion tool in the trades: it converts a one-time customer into a recurring one, which is where the real money is anyway.

Model 4 — Donate to charity based on volume, not sentiment

For every customer who completes a feedback interaction — review, survey, or referral — donate $X to a local charity. Published transparently.

• Why it holds: Google does not object to a charitable donation program that is not conditioned on review content or sentiment. However, the details matter: the program must be open to all customers equally, must not be conditioned on the star rating, and must not be branded as a reward for a positive review. "Every customer who leaves us feedback, regardless of what they say, helps us donate to [local food bank]" is defensible. "Five-star reviews earn us a donation" is not — the sentiment link is the problem.

Model 5 — Reward the referral, not the review

Pay for a referral who becomes a customer, not for a referral who posts a review. A referral program that pays on booked work is entirely separate from review generation and is often the highest-ROI customer acquisition channel in the trades.

Model 6 — Internal recognition, zero cash

Weekly leaderboard, "Technician of the Month," extra PTO, preferred shifts, a mention in the company newsletter. Recognition for consistent customer experience is normal employment practice and creates no policy exposure. Keep it tied to process behavior, not review counts, if you want zero risk.

And one legal caveat that deserves more than a footnote: if you are running any of these programs across state lines, in an industry with licensing boards, or with a workforce of 50+ employees, have a local employment attorney review the plan language. The FTC's 2023 revision of the Endorsement Guides treats employee testimonials as endorsements subject to the clear-disclosure rules, and several state attorneys general have been active on review-for-payment conduct. This is a fifteen-minute question for counsel and a catastrophic-risk question for a multi-state contractor.

6. Case Study: From 38 to 280+ Reviews and Doubled Inbound Calls

Company profile. A combined plumbing and HVAC company serving a mid-size metro — roughly 350,000 residents across 14 suburbs. Eleven technicians, one office, no marketing department. Founded 2011, previously doing almost entirely on referral and one decades-old Yellow Pages listing. The owner had heard of Google reviews but dismissed the idea because "the guys don't want to write stuff."

Starting position. 38 reviews total, averaging 3.9 stars, last review 11 months old. They were in the 3-Pack for "plumber" occasionally and never for "emergency plumber" or "AC repair same day." Inbound calls: roughly 45 per month, of which about 8 were from Google Maps directions — the rest from referral, yard signs, and a dead website that ranked for nothing.

"We weren't invisible. We were just never in the first three. When you're fourth on the map, you don't exist — that customer is already on the phone with number three." — Owner

The timeline

The results, and the parts that did not work

Inbound call volume from Google Maps went from 8 per month to 41 — roughly a 5x increase, with a slightly lower revenue per call as emergency mix normalized (they stopped being the desperate last resort, which was a good trade). Total inbound calls, counting all channels, roughly doubled because Google Maps calls converted to booked work at a higher rate than the referral mix they displaced.

Google Business Profile impressions grew roughly 11x over the same period. Discovery searches ("emergency plumber near me" and city-specific variants) went from 40 to 1,400 per month. The 3.2x photo-engagement effect showed up clearly in their data: photo-bearing reviews got more clicks, and the business moved from the 3-Pack into the top position on 6 of their 14 service queries — the majority of those queries had photo-bearing reviews among the three results.

What did not work, in their words:

• The "review us online" email blast. Sent to the entire customer database of 1,900 past customers, it produced 23 reviews — all from the easiest, happiest, most recent customers, and it deleted 4% of the SMS list to unsubscribes. Never again.
• Paying technicians $25 per review. Tested for one month because it was tempting. It produced a visible spike in review volume and a visible spike in quality complaints, because technicians started asking. They pulled it and the volume came back down about 60% of the spike over the following two months, because the underlying ritual had taken hold by then.
• A photo-heavy "5-star" incentive for technicians. Attempted once, abandoned after three weeks when customer feedback started mentioning it — "he seemed to want me to say the review was about the water heater specifically."

The tooling that ran the workflow

The operating system was built largely in the free review widget builder on besthelptool.com — QR code and magnet artwork with the business's branding and vanity URL, a landing page where the customer's Google review link resolves directly, and the widget embedded on the site. Their stated reason: it gave the office one place to change the QR destination, the review link, and the magnet artwork without reprinting 2,000 cards, and it let them see scan counts per QR placement as they tested magnet versus sticker versus door hanger.

The scheduling and SMS side lived in their existing field service software, and the review trigger was wired into the "job complete" status. The measurement was a simple spreadsheet: reviews by trade, reviews by source (magnet, sticker, SMS, email), average response time to reviews, and Maps calls per week, tracked in a 20-minute Friday review.

The metrics that actually predicted revenue

7. Common Objections, Answered

Is it legal to give my technicians a bonus for getting reviews?

Not in the way most people mean it. Google prohibits incentives tied to a review existing, its rating, or its content. What is safe is paying for the process — a documented handover, a completed feedback interaction, documentation quality — with no connection between the bonus and the star rating or whether a review appears at all. Never make payment contingent on a positive review, and never condition a customer's payment or discount on leaving a review. If the plan's language says anything like "bonus for a five-star review," rewrite it before anyone sees it, and have counsel review it if you are multi-state. See the incentive models above for the specific structures that hold up.

Can I ask happy customers for a review and stay quiet to unhappy ones?

No. Google explicitly prohibits soliciting reviews only from customers likely to leave positive ones, because it requires asking all customers equally without regard to satisfaction. In practice this is how most rating inflation happens, and it is exactly what an audit looks for. The operational fix is to make the ask universal and automatic — triggered by job completion status, sent to every customer, with the same script — and then manage unhappy customers through service recovery rather than by filtering them out of the review flow. If you do not want difficult customers writing public reviews, the fix is to make them call you first: a visible phone number and a genuine service guarantee in the closure message do more for your rating than any selective solicitation ever did.

How do I get reviews when my customers are stressed, angry, and just want the problem fixed?

Separate the emotional moment from the request, and use two different channels for two different states. In the stressed moment — burst pipe, no heat in January — the only job is to solve the problem. A review request in that window reads as opportunistic and will often be refused loudly. Instead, close the job with the artifact handover and the exact line "if anything isn't right, text me directly" and nothing else. Then trigger the SMS request 5 minutes later, when the stress has come down but the memory of the visit is still sharp. Also lean into your maintenance and seasonal visits, which are low-stress by definition and consistently produce the highest acceptance rates — a furnace tune-up is the best review request you will ever send, because the customer already planned to be pleased.

Do reviews from 2019 still matter, or should I focus only on new ones?

Old reviews still carry weight — a four-year-old five-star review showing a recent arrival time still signals that you did good work, and Google does not expire reviews. But recency matters disproportionately for two reasons: freshness affects how reviewers and customers read the profile ("280 reviews, most recent last Tuesday" reads differently from "280 reviews, last one in 2021"), and Google uses review recency as a relevance and velocity signal in local ranking. So the practical approach is to leave good reviews in place and never chase deletions of old ones — deleting a negative review without cause violates policy and provides no benefit. Focus your effort on velocity: a steady 10 to 20 new reviews a month for a multi-technician operation outweighs a year of silence followed by a burst. One caution on velocity: a sudden unexplained spike of short reviews with similar phrasing can trigger Google's systems. Keep your growth consistent and your review content varied and specific, which is also just better marketing.

Where to Start This Week

If you do only five things, do these.

1. Write the 30-second talk track. One page, taped inside the tech's truck door. Close-out statement, artifact handover, one-sentence ask, escape hatch. Train it by having each technician perform it for the office manager until it is unremarkable.
2. Redesign your leave-behind. QR code at 1.25 inches, high contrast, clear quiet zone, printed URL beneath, license number, tech name. Use the free widget builder to make the card artwork and the destination page without a designer or an agency.
3. Wire the SMS trigger to job completion. Five-minute delay, from your own number, with a template that names the job and the technician. Track reply rate and review rate per template.
4. Build a photo habit. Reframe the photo request as documentation for the homeowner. Same angle before and after. One clean finished-install shot per job. Then get those photos into the review request itself.
5. Write three response scripts and put them in the office. Price, lateness, surcharge. Four moves each: acknowledge, own, resolution plus process change, private channel. Pre-approved, so a 9 p.m. Sunday response is 90 seconds of typing instead of an argument.

None of this requires a marketing agency, a new website, or a budget. It requires a decision to treat the review as part of the job rather than as a campaign, and the discipline to do the same small thing every single day for six months. That is genuinely the whole game.