A client of ours, a commercial mechanical contractor, was ready to start an AI visibility program. Budget approved, scope agreed, everyone eager to begin. We asked them to wait a few weeks and do something else first: go get Google reviews. Not because the program would not work, but because starting it that week would have wasted a meaningful share of what they were about to spend.
Reviews Are Not the Soft Part
Most owners file reviews under reputation. Nice to have, good for the ego, something to chase when business is slow. That framing is now out of date.
When an AI assistant assembles an answer about who to hire, it draws on sources it can read and verify. A company's own website is the least independent of those sources, because the company wrote it. Reviews are the opposite: recent, specific, written by third parties, and available in volume. They are some of the richest material an engine has, and they carry weight accordingly. We wrote about the mechanics in your Google reviews are talking to AI.
Which means a business with four reviews and a business with forty are not in the same position, no matter how good their websites are. The first is asking engines to take its word for things. The second has corroboration.
The Arithmetic of Starting Too Early
Here is the practical case we made.
A visibility program is an investment in signals: consistent information everywhere, structured data, content shaped around real buyer questions, and measurement over time. All of it works. None of it substitutes for third-party corroboration, and the engines weigh corroboration heavily precisely because it is the part a business cannot write itself.
So a program launched on a thin review profile spends real money strengthening three of four legs while the fourth stays short. The work is not wasted, but a share of it is working against an unnecessary handicap. Reviews, meanwhile, cost almost nothing but time and a willingness to ask. It is the highest return item on the list and the only one a client can do largely themselves.
Spending six weeks on the cheap thing before starting the expensive thing is not a delay. It is sequencing.
How to Tell If You Are Ready
Pull up your Google Business Profile and three competitors in the same market, then look at four things.
Volume relative to competitors. Not an absolute number. If the firms you lose bids to have thirty reviews and you have five, that gap is visible to both people and machines.
Recency. A cluster of reviews from four years ago reads as a business that was good once. Steady arrival matters more than total count.
Specificity. Read yours honestly. Do they describe actual work, or do they say great company, highly recommend? Generic reviews are pleasant and nearly useless as source material, because there is nothing in them for an engine to extract.
Owner responses. Your replies are public, crawlable content on the same page. A profile where every review has a thoughtful reply tells both audiences that someone is paying attention.
If three of those four look weak, the review work comes first.
The Part Where We Changed Our Own Advice
When we set this up for the client, our first draft included something that seemed helpful: a suggested review written for each customer, so nobody had to face a blank box.
We took it out, and we would recommend you do the same. Google's policies prohibit soliciting reviews with prescribed content. More practically, when several reviews for one business share phrasing patterns, that is exactly what review spam detection looks for. Reviews get flagged and removed, and profiles can get suppressed. For a business about to build a visibility program on top of those reviews, that is the worst possible failure: the foundation disappears after the work is paid for.
The replacement works better anyway. Personalize the request, not the review.
How to Actually Ask
Pick customers you genuinely served well, and send each one a short message that does four things.
Name the specific work. "The boiler replacement at the Fifth Avenue building last spring" beats "our recent work together," because it reminds them what they would be writing about.
Make it mechanically easy. Include the direct link to leave a review, and a line about signing up for a Google account if they do not have one. Every extra step loses people.
Ask for specifics without dictating them. One sentence does it: "If you have a minute, we would appreciate a review, and it helps most if you mention what we actually did for you." That single line is the difference between great company and a review that describes the job.
Then reply to every review that arrives, including the short ones.
One Exercise Worth Doing First
Before writing to anyone, open Google Maps and read your competitors' reviews. Not to feel bad, which is the usual outcome. To see what customers in your market actually write about, which problems come up repeatedly, and which language recurs. You will usually notice that the best reviews describe a specific situation and how it was handled.
You will also notice that a competitor with more reviews than you is not necessarily better than you. They asked. That is often the entire difference, and it is a fixable one.
What Happens Next
Our client is working through their list now. When the reviews are in, we will begin the visibility program on a foundation that is meaningfully stronger than it was a month earlier, for an investment of time rather than budget.
If you are considering this kind of work and you are not sure whether your foundation is ready, that is a reasonable thing to ask about before spending anything. Tell us about your business and we will give you a straight read, including the times when the answer is that you should go ask for reviews and call us in six weeks. More on the work itself: AEO Intelligence.