Your crew did excellent work this morning. The customer was delighted. She said so, out loud, to a technician who was already thinking about the next job.
Nobody asked her to write it down. Not because the team is lazy, but because the ask depends on a person remembering, at the end of a physical day, at the exact moment they are least likely to remember anything.
So you sit at 40 reviews, which is a respectable number and a terrible one. Respectable because they are real. Terrible because the company down the road has 240 and is winning searches you should own, on work that is no better than yours.
Reviews are not a marketing activity. They are an operational output, and outputs get automated.
What 40 reviews costs you
The benchmarks here are unusually well supported, and they compound.
Add the two figures that explain why. Between 93 and 97% of consumers read reviews before buying, around 92% for local services, and roughly 85% say they trust them about as much as a personal recommendation. Your reviews are not a badge. They are the sales conversation that happens before anybody contacts you.
The volume math is simple and slightly painful. Say you complete 45 jobs a month and roughly 4% of customers leave a review unprompted. That is under two a month, so 40 reviews took you two years. A well-built sequence typically lands somewhere between 25 and 40% on happy customers.
| Per month | Today | With a sequence |
|---|---|---|
| Jobs completed | 45 | 45 |
| Review requests actually sent | ~6 | 45 |
| Reviews received | 1.8 | 13.5 |
| Months to reach 200 reviews | 89 | 12 |
Seven years of drift, or twelve months of a sequence nobody has to remember to run.
The sequence, step by step
Eight steps. The order matters more than the wording.
What the two messages actually say
Message one, within two hours of completion. Name the customer, name the technician, name the job in three words, then one question: how did we do today? If the answer is positive, the review link follows immediately. If it is anything short of positive, it goes to a human instead, and that human calls rather than texts.
Message two, at 72 hours, only if there is no review yet. One line, no guilt, with the link repeated. Something close to: No worries if not, but if you have ninety seconds, it genuinely helps people find us. Then it stops, permanently, for that job.
Two messages, both short, both specific. Length is what kills review requests. Nobody has ever written a review because the ask was thorough.
What changes once it is in
Before: a review arrived when somebody was unusually moved to leave one, which meant the delighted and the furious, in roughly equal measure.
After: every completed job asks, once, at the right moment, with a temperature check in front of it. The rating climbs because the happy majority is finally represented, and the unhappy minority reaches you before it reaches Google.
The knock-on effect is the one that pays. Higher volume and a better average rating lift you in local search, which lifts inbound inquiries, which flows into the same pipeline you already fixed. This is why the review sequence is usually the last piece I build and the first one owners mention six months later.
You are not asking for a favour. You are asking a delighted customer to spend ninety seconds telling the truth.
The measurable: reviews per completed job. Under 5% today. If it is not above 20% within a month of go-live, the timing or the link is wrong, and both are easy to fix.
