Short answer: Ask every customer, at the moment they are visibly satisfied, using one repeatable method — then reply to what comes back. Volume is the part you control, and the way to grow it without risking your profile is to be boringly consistent rather than clever.
This guide is the operational one: how to get more reviews without breaking Google’s rules.
Know what you are actually optimising for
Reviews serve two different jobs, and confusing them leads to bad strategy.
The map pack. For “plumber near me,” Google shows three businesses above your links. You are not on it unless your Google Business Profile is competitive. Review count and rating are consistently ranked among the top local pack factors by industry survey, and that weight grew from roughly 16% of local pack factors in 2023 to around 20% by 2026. [4][5]
Organic and AI answers. When an assistant answers “which HVAC company in Denver is most reliable?”, it needs evidence to summarise. Review text is usually the only place where a customer’s actual experience with you specifically is written in natural language. That is the angle covered separately in Customer Reviews and AI Trust.
For a local service business, the map pack is the bigger commercial prize. Fix volume there first.
The volume picture
The gap between appearing and not appearing in the map pack is usually volume:
| Map pack position | Typical average review count |
|---|---|
| Positions 1–3 | 200+ |
| Positions 4–10 | ~150 |
| Positions 11–20 | ~80 |
These are industry-study averages, not guarantees, but the shape is stable across datasets. The useful conclusion is blunt: star rating is not the lever. A 5.0 from 12 customers is a small sample, and both search engines and customers treat it as one. A 4.7 from 200 is a track record. Chasing the tenth five-star review costs you the twentieth review, which is worth more.
The compliant ask
Google’s guidance on getting reviews is short and specific. [1]
- Use a Google review link or a QR code. Google provides the link generator, and a QR code on an invoice, job card, or counter card removes the friction of typing an address.
- Ask in a way that makes the requirement clear — customers must be signed in to a Google Account, which is a real drop-off point worth anticipating.
- Ask everyone. No filtering by how the job went.
What is prohibited, and why it matters
The list is short enough to memorise. All of it is fake engagement under Google’s policies. [3][2]
| Tactic | Status | Why it fails |
|---|---|---|
| Discount or free service for a review | Prohibited | Inducement, explicitly called fake engagement |
| Gift, raffle entry, free consultation | Prohibited | Same category |
| Only asking customers likely to be happy | Prohibited | Review gating; skews the sample |
| Buying reviews or using review services | Prohibited | Fabricated contribution |
| Asking for removal in exchange for anything | Prohibited | Same as inducement |
| Offering to fix a problem in exchange for removal | Prohibited | Extortion scam category |
Enforcement is not theoretical. Google states it may block a profile from receiving new reviews, unpublish existing reviews or ratings, and display a consumer-facing warning that fake reviews were removed, notifying owners by email. Businesses have reported losing review counts overnight. In July 2026 Google added explicit guidance covering fake and incentivized reviews to its review-markup documentation, naming manual action as a consequence.
If someone has previously offered to sell you reviews, that is a red flag worth taking seriously rather than treating as a budget line.
Timing beats wording
You do not need a clever ask. You need the ask at the right second.
Ask when satisfaction is visible. End of job, payment settled, work demonstrated. The gap between “job completed” and “invoice due” is where goodwill evaporates.
Make it one action. A link, or a QR code the customer scans on the spot. Every extra step costs you completions.
Make it routine, not personal. Whoever finishes the job sends it — not the owner, not the office. Review volume is a function of how many times the ask happens, and personal follow-up does not scale past about twenty jobs a month.
Say what you want. Customers often assume you want five stars and write “great service.” Tell them honestly that specific detail helps other customers choose, and that describing the actual job is what they should do. This is a request for detail, not for sentiment, and it is entirely legitimate.
One reminder, then stop. A single follow-up is fine. Repeated chasing is not.
Replies are free ranking surface
Google’s own reply guidance is unusually specific: keep replies short, be conversational rather than promotional, do not offer deals to people who are already customers, and do not send the same thank-you to everyone. [1]
That last point is the one businesses get wrong. A templated reply tells a reader — and any model summarising your profile — that the reviews are not actually being read.
Useful replies do one of four things:
- Add information the reviewer did not have.
- Answer a question they raised for the next person.
- Own a mistake specifically, without excuses.
- Thank them in a way specific to the job they described.
For negative reviews, Google is unusually clear that they are not automatically a problem: honest and balanced feedback can look more trustworthy than uniformly positive reviews. The reply is where you demonstrate you are the business you claim to be. Flag a review only when it breaks content policy — never because it is merely unflattering.
What to measure
Four numbers, monthly:
- New reviews added — the input you actually control.
- Median date of recent reviews — confirms the ask is still firing.
- Map pack appearances for your top service searches.
- Reply rate — target 100%, and it is the easiest number to fix.
Star rating is a lagging indicator. Volume and reply rate are leading indicators, and they respond within weeks.
What this connects to
Reviews are one input among several, and they reinforce the others. A detailed guide to schema markup lets search engines attach your reviews to the right services. Consistent business data across directories stops reviews being credited to the wrong listing. As modern search engine optimization has shifted toward summarising third-party evidence rather than matching keywords, reviews have moved from being a marketing asset to being a primary source material — which is why we treat them as a Local Search vs AI Answers problem and not only a ratings problem. And automating the ask is usually the first place where a service business gets value out of work smarter through digital efficiency.
