Running reviews across many locations

One office can run reviews on goodwill and a reminder in somebody's calendar.

At several offices that collapses, and it collapses in a particular way. The strongest office gets stronger. The weak ones go invisible. And the report that lands on your desk shows a healthy average describing nowhere.

Why the brand average lies

Reviews rank an office against other businesses near that office. One competing in a dense city and one in a small town are in different competitions. Averaging them produces a figure that describes neither.

It gets worse. The average is dominated by whichever office has the most reviews — usually the oldest and biggest. A new office with nine reviews disappears into the brand number while being extremely visible to everybody searching nearby.

Report by office. Always.

Centralize this, not that

Centralize: the compliance policy, the request templates, the timing, the tooling, who to escalate a serious complaint to, and the training on what staff can't say in public. These are exactly what goes wrong when each office improvises, and the FTC exposure lands on the whole company.

Don't centralize the replies. A central team writing every response produces identical replies across every office, which reads as automated and kills the point of replying. The value of a reply is that it shows a human being at that office read the review.

What works: a central policy with examples, and local people writing in their own words inside it.

The part that actually breaks

Here's the failure nobody plans for. Your request process can't tell which office the client dealt with.

A central system sends review requests from head office with a generic link. Reviews land on whichever profile the client happens to find — often head office, sometimes an office you closed in 2021. Four months later the numbers look wrong and nobody can reconstruct why.

Fix it at the source. The request has to carry that specific office's review link, which means the record has to know which office did the work.

That's a data problem before it's a marketing problem, and it's worth solving before you buy any software to sit on top of it.

When software starts paying

For one office, the audit is an afternoon a year and you don't need a subscription.

The math changes with scale. Somewhere in the several-offices range, the cost of somebody manually checking listings, catching duplicates, routing requests and watching responses passes what the tooling costs.

The honest trigger isn't a count of offices. It's whether anybody is actually doing the work. Unmanaged listings are the real expense, and they stay invisible until a prospect calls a number that's been disconnected for a year.

Whatever you buy, apply the rules in buying services: you own the accounts, you own the data, you can export it, and you can leave on short notice.

What to actually read

Per office, each against its own history:

• Requests that turn into reviews. An office well below the others has a process problem, not a quality problem.
• Reviews per month. The trend matters. The lifetime total doesn't.
• Recent rating, separate from lifetime rating. A lifetime average can hide a decline for years.
• How often and how fast they reply.
• The actual text. Read it.

That last one is the real return on all of this. The same complaint turning up at one office is the cheapest operational intelligence you'll ever get. An office whose reviews have slid over two quarters has a management problem — and the reviews told you before any internal system did.