Automated Google Review Requests
After every completed job a short request for a review goes out – to every eligible customer, not only the ones you expect to be happy. Anyone with an open complaint, an objection on file or a request already sent this year never makes the list.
Ask nobody and you only hear the annoyed
Almost nobody writes a review unprompted. It takes a reason strong enough to sit down in the evening and open a Google account, and irritation is a far stronger reason than satisfaction. The customer whose boiler has run quietly for three weeks has stopped thinking about you; the one whose engineer turned up two hours late has not. So a profile nobody has ever asked for does not show the average of your work. It shows the average of the jobs that annoyed someone enough to act.
The hard part is not the sending, it is knowing when a customer is actually finished. A job the system marks as done may still carry an open complaint, an unpaid final invoice or a second visit booked for Tuesday. Ask at that moment and the silence would have been cheaper. On top of that, the same person sits in three systems at once – the diary, the invoice, the ticket – so without one leading system and a memory of who has already been asked, a friendly request becomes a series within a few weeks.
Then there is the line you may not move. Asking about satisfaction first and sending the review link only to the happy ones is the obvious idea, and it is exactly what Google forbids as review gating; so is anything handed over in return for a review. That leaves three levers: ask every eligible customer rather than only the vocal ones, get the timing right, and route complaints to a person who answers them before they become public. Those three are what this page is about.
When asking makes sense
We start with the completion that occurs most often and has the clearest moment. Further occasions later share the same suppression list and the same channel.
After an on-site job
The most common starting point: the engineer closes the job, the request follows two hours later – but only if no return visit or open item hangs on it.
After delivery or collection
In retail the delivery status counts, not the order date. Let the returns window pass as well, or you are asking into a parcel on its way back.
Several locations and branches
Every location has its own business profile and its own link. The wrong one credits head office and leaves the branch that did the work empty-handed.
After a resolved service case
Only for tickets closed without escalation or reopen. A case that dragged on is a poor moment to ask, even if it ended well.
After advice or treatment
For law firms, practices and consultancies the wording decides: no case reference, no name in the text, nothing that hints at the matter.
After recurring appointments
Maintenance contracts and subscriptions produce a completion every quarter. Without a cooling-off period the same customer is asked four times a year.
What changes on the profile – and who gets asked tonight
On the left the profile as it forms when nobody asks, on the right the same profile after a quarter on paper. Below it one day of the suppression list: seven completed jobs and the decision the workflow reaches on each.
Profile without requests 3.4 34 reviews
- 514
- 45
- 33
- 24
- 18
The same profile, one quarter on paper 4.0 72 reviews
- 541+27
- 412+7
- 35+2
- 25+1
- 19+1
A worked example, not a promise: 120 completed jobs a month, roughly 100 requests once the suppression list has run, over one quarter – and the assumption that one in eight recipients writes something. Change that assumption and the right-hand side changes with it. Nobody knows your own rate before the first month.
Seven jobs closed yesterday, seven decisions
| Completed job | What the suppression list finds | Decision |
|---|---|---|
| Boiler service · Ms Albers | invoiced, nothing outstanding | request 18:10 |
| Bathroom refit · Mr Reiners | second visit booked for Tuesday | checked again afterwards |
| Callout · the Voss family | complaint open since Monday | excluded, goes to service |
| Roller shutter motor · Ms Timm | invoiced, first job for her | request 18:10 |
| Electrical check · Mr Albers | same number as row one | second contact in the household |
| Window fitting · Kranz Furniture | marketing objection on the record | excluded for good |
| Service contract · Café Sonn | asked in March this year | held until March |
On an ordinary day the first row repeats down the page; here the cases that make up the suppression list are lined up side by side. Sending window Monday to Friday, 9am to 6pm – anything falling due at 7:40pm goes out the next morning.
A task, not a template: one or two stars create a record with a deadline. The public reply is the one part of this chain nobody should automate – it is read by customers and by Google alike.
The number of poor reviews does not fall – it simply grows more slowly than the rest. Reducing it means removing the cause, not the request.
How the new reviews turn out is the real unknown, and that rests on your work rather than on the request. The reporting is built so that after four weeks it shows your own figures instead of our assumption.
From completed job to review
-
Detect completion
The trigger is not a date but a status change in the leading system: job invoiced, ticket resolved, delivery confirmed. We agree which system that is before we build, and there can only be one.
-
Run the suppression list
Before anything is sent, the list takes out everyone who must not be asked: marketing objection, open complaint, disputed invoice, a request in the last twelve months, a second contact in the same household.
-
Pick channel and review link
Email, SMS or WhatsApp, depending on which contact detail exists and what consent covers. The link belongs to the business profile of the location that did the work.
-
Send at the agreed moment
The message goes out after the configured wait, but only inside your sending window. A follow-up is optional, runs exactly once and is cancelled once the link has been opened.
-
Feed the reaction back
Click, opt-out and reply land on the customer record, so the suppression list is right on the next run. A new review below three stars creates a task for the public reply.
What changes day to day
Today
- Somebody asks whenever it happens to cross their mind
- The same customer gets the request twice in one quarter
- The link points at head office, not the branch that worked
- Reviews come almost entirely from people who were annoyed
- A one-star review sits unanswered for days on end
With automated requests
- The request goes out once the job is genuinely finished
- Each person is asked at most once in twelve months
- Each branch collects the reviews it actually earned
- Even customers who would never speak up unprompted get asked
- A poor review is on somebody's desk the following morning
What review requests cannot do
Sending is the easy part. We clarify these four before quoting, because they decide whether the project is worth doing at all:
- With low volumes it does not pay. If you close twenty or thirty jobs a month, you can write the request yourself in five minutes, and personal wording beats any template – set against the entry price, it takes years to come out even. The automation earns its keep once the record of who was asked when no longer fits in anyone's head, or once several locations run their own profiles.
- We do not build pre-selection, not even as an intermediate page. The most common request in the intro call is a small survey up front: ask for stars internally, send four and five on to Google, route the rest into a form. That is still review gating whatever the page is called, and we will not build it – not even when a competitor looks better for doing it. If that funnel is what you want, you are in the wrong place.
- More requests will not lift an average the work does not support. A firm with a genuine quality problem simply collects poor reviews faster – the automation makes your work more visible in both directions. A review cannot be deleted in any case, only reported and answered.
- Who you are allowed to contact is not a technical question. A review request by email to existing customers counts as marketing in law, SMS and WhatsApp need consent of their own, and healthcare and legal services carry extra rules. We implement what you agree with your legal advisers and log every send; the basis has to come from you.
Fits your systems
Scope and price
The entry price covers one trigger, one channel, one business profile and your suppression list. What moves the price, we say before the quote.
- Trigger from one leading system: job, ticket or delivery
- Suppression list: objection, complaint, cooling-off, duplicate contact
- Request in your own wording on one channel, with an opt-out link
- One Google business profile with the matching review link
- One follow-up that runs exactly once – or none at all
- Click, opt-out and contact status written back to your CRM
- Documentation, handover session and 30 days of support
What increases the price
- Several locations with their own profiles and reporting
- A second channel such as SMS or WhatsApp Business with separate consent
- Requests in several languages and country-specific rules
- Trigger from a till or industry system without an open interface
- Further platforms such as Trustpilot in the same report
Several locations, a second channel and requests in more than one language typically land in the range of our Workflow Advanced package from €2,490. We quote the binding fixed price after the intro call.
All prices excl. VAT · operation and further development optionally via a support package
What you get
-
Production request flow
Set up from status change to sent message, signed off in a dry run on real jobs
-
Documented suppression list
Which job never triggers a request – as a table you can have changed later
-
Monthly report
Requests, clicks, opt-outs and the state of the profile, with the send times compared
-
Handover for service and sales
How a customer is suppressed for good and who writes the public reply to a poor review
Frequently asked questions about review requests
These solutions fit alongside
Customer Feedback & NPS Automation
The inward side of the same question: feedback that never becomes public.
Complaint Management Automation
Where a complaint goes before it turns up as a review on your profile.
WhatsApp Business Automation
The second channel when email addresses are missing – with its own consent and window.
Email Automation
Following up with a hard stop: one message, then quiet.
How many of your happy customers does anyone ask?
In the free intro call we work out which system reports completion reliably, who you are allowed to contact and which jobs should never trigger a request. Afterwards you know whether it pays off at your volume.
Book a free intro callWhere automated review requests creates value in everyday work
The workflow detects completed services, checks contact permission and suppression rules, sends a neutral review request and routes responses appropriately.
Three concrete operating scenarios to compare with your own process.Ask after an on-site visit
After completion and a short delay, the customer receives a concise request for feedback.
Handle multiple locations
The correct review link is selected from location or branch.
Route replies to service
Questions or complaints create a task with customer context.
A strong fit when …
Recurring office tasks follow clear rules, consume small blocks of time every day and should run reliably without replacing existing systems.
- You handle recurring completed customer cases using repeatable rules.
- The intake, target system and accountable business role can be named clearly.
- Exceptions are allowed to remain visible and move to people deliberately.
Deliberate automation boundary
There is no review gating: all eligible customers are asked neutrally. Complaints and sensitive cases go to people, not into an automated review sequence.
Explore the technical approach and platformsEstimate time savings with your own volume
The calculator uses 5 minutes today and 1 minutes after automation as fixed example assumptions. It does not replace process analysis.
Illustrative estimate based on the visible assumptions — not a guarantee.