Email and customers · Lesson 5.5
Measuring Automated Review Requests: 6 Numbers Worth Tracking
Once automated review requests are running, the temptation is to watch the star rating and nothing else. That number moves slowly and depends on many things you do not control. A few other numbers tell you much sooner whether the system is working.
This post names six, explains how to calculate each one with a plain sheet, and gives you a simple weekly and monthly reading routine. None of the figures in the examples are real. They show the method, and the right target is whatever your own business produces. The setup the metrics belong to is taught in Lesson 5.5: Get five star reviews without chasing.
Track requests sent, reviews received, the ratio between them, reply time, rating spread and what reviews mention, and read them monthly, not daily.
Why activity numbers mislead
"We sent 400 requests this month" sounds good and means little. It measures effort, not outcome. The numbers that matter connect the ask to something that happened afterwards.
Equally, do not chase a perfect rating. Google's help says honest and balanced reviews can help potential customers decide, and that a mix of positive and negative feedback often feels more trustworthy. A profile with nothing below five stars can raise doubts, and chasing it tempts people toward rules they should not break.
The six numbers for automated review requests
1. Requests sent
How many first asks went out in the period. It is a check that the automation runs, not a measure of success. If it drops to zero, something broke.
How to get it: count rows with a date in "Asked on", or read the sent count in your email tool.
2. Reviews received
How many new reviews appeared on your Business Profile in the same period. Google lets you read them in your profile under Read reviews.
A caution: reviews may arrive days or weeks after the ask, so a month's reviews will include some from last month's requests. Also, not every review can be traced to a request, since customers can review without being asked.
3. Reviews per request
Reviews received divided by requests sent, over a month or longer. This is your main health number, sometimes called a response rate.
Example, illustrative only: if you sent 100 requests and received 8 reviews, the ratio is 8 percent. The number is not good or bad by itself. What matters is the trend in your own business after a change.
Use a long window. With small numbers, one extra review can swing the percentage a lot. Compare three months, not three weeks.
4. Time to reply
How long between a review being posted and your reply. Track the average and the longest. Google's help says a prompt response shows you value feedback, and its tips list responding in a timely manner as good practice for negative reviews.
How to get it: note the review date and the reply date in a sheet. Google also says replies are checked before posting, usually within about 10 minutes but sometimes up to 30 days, so count from when you submitted the reply.
A realistic target: one or two working days is a sensible personal goal. It is not an official figure.
5. Rating spread
Not only the average, but how many reviews at each star level. A shop with a 4.6 average from a mix of 5s and a few 3s looks different from one with a 4.6 average from a few 1s and many 5s.
How to read it: look at the lowest ratings. What do they say? If several mention the same thing, it is a problem to fix, not a number to hide.
6. What reviews mention
Not a number but a count: tally the topics that appear, such as speed, price, friendliness, quality, and delivery. A list of the top three themes each month is more useful than the stars.
AI can help: paste the month's reviews and ask for recurring themes and one representative quote for each. Check the result against the reviews yourself, because summaries can overstate patterns.
A simple tracking sheet
Make one row per month with these columns:
| Column | Notes |
|---|---|
| Month | Calendar month |
| Requests sent | From your list or tool |
| Reviews received | From Business Profile |
| Reviews per request | Received divided by sent |
| Avg reply time | In days |
| Count of 1 to 5 stars | Five small columns |
| Top theme | One line |
Fill it on a fixed day each month. It takes about fifteen minutes.
How to read it weekly
Weekly is for the machine, not the results. Spend five minutes checking:
- Did requests go out this week? If not, why?
- Are there unanswered reviews?
- Did anyone email a complaint about the message?
That is all. Do not read anything into one week's review count.
How to read it monthly
Monthly is for results. Ask:
- Is reviews per request moving in a direction over three months?
- Are replies going out in good time?
- Did a change you made, such as a new trigger day or message, coincide with a shift?
- What is the one theme to act on?
Change only one thing at a time so you can attribute a shift. Even then, be modest about what you conclude. Season, a promotion, a bad delivery batch, or a lucky run can move the numbers just as much as your message.
Numbers not to chase
- A target star rating. Do not set it. You cannot control honest feedback, and trying to creates pressure to filter, which Google's policy says merchants should not do (selectively soliciting positive reviews).
- A review count by a deadline. Google's policy says merchants should not request that staff solicit a certain number of reviews.
- Competitor review counts. Different businesses, different customers.
- Open rates alone. Email open numbers can be unreliable on their own. Treat them as a rough signal at most, and look at reviews per request instead.
Spotting a fault from the numbers
| What you see | Possible cause |
|---|---|
| Requests drop to zero | Trigger or automation stopped |
| Requests steady, reviews near zero | Link broken, emails in spam, or timing off |
| Reviews up but several mention the message | Wording feels pushy |
| Reviews go missing | Google can remove reviews that breach its policies, so check for incentives or steering in your process |
| Rating falls | Look at the themes, not the message |
If a review you expected is not showing, Google has a help page about missing or delayed reviews, so check it before assuming a problem.
Keeping the data tidy
The cleaner your customer list, the more reliable these numbers are. A list where each customer has a completed date and an "asked on" date makes requests sent trivial to count. Lesson 5.1 covers keeping everything in one simple list. For more on avoiding errors that distort the figures, see the post on five mistakes that cost time.
Questions people ask
What is a good response rate?
There is no official benchmark I can verify. Measure your own baseline and compare it with itself over time.
Should I track the star rating?
Yes, but as the spread and the themes, not as a target.
How often should I look at the numbers?
Weekly for whether the system is running, monthly for results.
Can AI analyse my reviews?
It can group themes. Read the reviews yourself to check its summary.
Start with one sheet
Create the tracking sheet today and fill in this month's figures, even if they are rough. In three months you will have something real to compare. If you want the setup that creates the numbers, with the ask, link and reply workflow, Lesson 5.5 covers it: Get five star reviews without chasing. The complete setup guide is a good first read.