Most business owners already know they should be asking for reviews. That is not the gap. The gap is that asking depends on a human remembering to do it, at the end of a job, when the van is being loaded, the next customer is calling, and the invoice still needs sending.
So the asking happens in bursts. A good week where someone remembers, then three weeks of nothing. Then a panic when a competitor appears above you in the map pack with twice the review count.
The businesses that quietly pull ahead on Google reviews are not better at asking. They have removed the asking from the list of things anyone has to remember. A trigger fires, a message goes out, and the owner finds out about it when the review appears.
This is how to build that, and how to do it without tripping over Google’s policies or ignoring the customers who are not happy.
The failure point is memory, not technique
Ask ten business owners why they do not have more reviews and most will describe a technique problem. They are not sure what to say. They feel awkward. They do not want to pester people.
Watch what actually happens and the picture is different. The awkwardness is real but it is not what stops the ask. Manually asking fails because the moment passes. The job finishes at 4pm, the engineer is thinking about traffic, and by the time anyone sits down to do admin, four jobs have blurred into one and nobody can remember which customer was which.
This matters because it explains why most review advice does not work. Scripts, wording templates and "just ask at the end of the job" all assume the bottleneck is knowing what to say. If the bottleneck is remembering to say it at all, better wording changes nothing.
It also explains the pattern you see in review profiles: a run of reviews in one month, then a gap, then another run. That is not customer behaviour. That is somebody having a push and then getting busy. Google’s local results favour steady, recent activity, so an inconsistent profile undersells a business that is actually doing good work every week.
The fix is structural. Take the decision away from a person who is already at capacity and attach it to something that happens anyway.
What a job completion trigger actually looks like
The principle is simple. Find an event that already gets recorded in your business, and hang the review request off that event.
The two most reliable are:
Job marked complete. If you use a job management system, somebody already ticks a job off when the work is done, because that is how the invoice gets raised. That tick is your trigger.
Invoice marked paid. If you do not run job management software, your accounting system still records payment. Payment is a strong signal: the customer has had the work done and has been satisfied enough to settle up.
Either event can push the customer’s name, mobile number and job type into an automation, which then waits a set period and sends a message. The message carries a direct link to your Google review form, not to your Google Business Profile in general. Every extra tap loses people.
What it connects to depends on what you already run. A plumbing firm on a field service app can trigger from job status. A clinic can trigger from the appointment being marked attended in the booking system. A fit out contractor with nothing but a spreadsheet and accounting software can trigger from the invoice being marked paid. The tooling varies; the logic does not.
Two practical details decide whether this works or quietly breaks.
First, phone numbers. Automated requests sent by SMS consistently outperform email for local service work, and SMS only works if mobile numbers are captured properly at the enquiry stage. If half your records have a landline or an empty field, fix that before building anything else.
Second, suppression. The automation needs to know not to message the same person twice, and not to message customers you have deliberately excluded (a job that went badly and has already been dealt with, for example). A simple flag on the customer record is enough.
Ask everyone, including the ones you are nervous about
There is a tempting shortcut here to avoid. It goes like this: send everyone a quick satisfaction question first, then only send the Google link to the people who answer positively, and route the unhappy ones to a private feedback form.
That is review gating, and Google prohibits it. Their policies cover discouraging or filtering negative reviews, and businesses have had review content removed over it. UK consumer law has also hardened around this, with fake and misleading reviews now treated as a banned commercial practice, so the risk is not purely reputational.
It is also unnecessary, which is the part people miss. A business doing genuinely good work does not need a filter. If nine out of ten customers are happy, asking all ten produces a strong average and one honest review that makes the rest look credible. Profiles with nothing but five stars read as suspicious to customers who have been online for more than a week.
Structure the ask so filtering never becomes tempting:
- Send the same message to every completed job, with no qualifying question in front of it.
- Keep it short and specific. Name the job, thank them, give the link, say it takes a minute.
- Never offer anything in exchange. No discounts, no prize draws, no free call out next time. Incentivised reviews breach Google’s policies and distort what people write.
- Do not tell the customer what rating to give. Ask for their honest experience.
If the thought of asking everyone makes you uncomfortable, that discomfort is useful information about the work itself rather than a reason to build a filter.
The timing window that decides everything
The single biggest lever on response rate is how long you wait.
Ask immediately, while the engineer is still packing up or the customer is still in the car park, and you catch attention but before the customer has lived with the result. Ask two weeks later and the job has faded, the detail has gone, and you are asking for a favour rather than a reaction.
The window that tends to work is the same day, a few hours after the job closes or the invoice is settled. Late afternoon and early evening sends land when people are checking their phones rather than working. The customer still remembers the technician’s name and the thing that impressed them, which is exactly the detail that makes a review persuasive to the next reader.
One follow up is worth sending, roughly three days later, worded as a light reminder and nothing more. A second follow up starts to annoy people and rarely converts. Build the sequence as two messages and stop.
There are exceptions worth coding in. Work with a settling in period, such as installations or treatments with a recovery phase, is better asked about once the customer has experienced the result. Set that delay per job type rather than applying one blanket rule.
Build the negative response into the same system
Most review processes are built to collect praise and then fall apart when criticism arrives. The owner finds the one star review a fortnight later, replies while angry, and makes it worse.
Treat negative reviews as part of the same workflow, not a separate crisis.
That means an alert and a clear escalation path. Any new review below four stars should notify a named person immediately, by text or by whatever channel they actually read. Speed matters because a complaint sitting unanswered for a week reads as indifference to everyone who sees it afterwards.
It also means a prepared response pattern so nobody writes from scratch while irritated. Acknowledge the specific issue, avoid defending, state what you are doing, and offer to continue the conversation directly. Never dispute facts in public, even when you are right. Prospective customers are not judging who was correct; they are judging how you behave when something goes wrong.
Handled this way, a negative review becomes something close to an asset. A profile with a strong average and one calm, human reply to a genuine complaint converts better than an unblemished wall of five stars, because it proves a real business is reading them.
Tracking the number, or watching it drift
A vague intention to "get more reviews" produces vague results. Review generation behaves like any other part of your lead generation engine: it needs a number attached to it.
Record three things monthly:
- Total review count and average rating, so you can see whether the trend is up and whether the average is holding.
- Requests sent versus reviews received, which gives you a conversion rate. If that rate drops, something in the message, the link or the timing has broken.
- Competitor counts for the two or three businesses that outrank you locally. This tells you whether you are gaining ground or simply standing still faster.
Monthly tracking is enough. The point is not the spreadsheet, it is that a drop becomes visible within weeks instead of being discovered when a customer mentions that your rival looks more established.
Where this sits in a wider system
Reviews are one of the few assets that compound. Every month the trigger runs, the profile gets stronger, the map pack position improves, and the enquiries that arrive are already half convinced before anyone picks up the phone.
The build itself is not complicated: an event in a system you already use, a short message, a direct link, a sensible delay, an alert when something goes wrong, and a number you look at once a month. What it removes is the need for anybody to remember.
If you would rather that trigger was wired into your enquiry, booking and follow up process as one piece of plumbing rather than a bolt on, that is the sort of thing Nimble Dingo’s AI growth systems are built to do.