How to get more Google reviews: what works, and what gets you removed.

At the moment the customer is most pleased, which is almost never the moment the invoice goes out. For most trades that is when the job is finished and they can see the result.

The gap between those two moments is where most reviews are lost. A request attached to an invoice arrives while somebody is thinking about money, and the response rate reflects it.

For a service business the best moment is usually face to face, at the end, followed by a message the same day with the link. The ask happens in person and the link makes it easy, and neither works nearly as well alone.

Good moments, by trade

• Trades.

  When the work is done and you are walking them through it, before you leave the site.

• Clinics and allied health.

  At the end of a course of treatment, not after a first appointment.

• Retail and hospitality.

  On the receipt and on the table, where it is already in front of somebody.

In your own words, in person where possible, followed by a direct link. The link matters because most people will not search for your business to leave one.

Google provides a short link for exactly this, from inside the listing. Sending somebody to search for you loses most of them, because a review is a small favour and any friction is enough to lose it.

What you say matters less than people think, as long as it is specific and human. Naming the work you did for them produces a review that describes the work, which is worth more than a five star with no words.

What works

• Ask in person, then send the link.

  The in-person ask does the persuading, the link removes the effort.

• Name the job.

  "If you have got a minute to mention the hot water job" produces a far more useful review than a generic request.

• Reply to every one.

  Google counts replies, and it is the cheapest signal available to any business.

Paying for reviews, offering anything in exchange, filtering out unhappy customers before asking, and writing them yourself. All four are common and all four risk the listing.

Review gating, sending happy customers to Google and unhappy ones to a private form, is the one respectable businesses fall into without realising it is a breach. It is explicitly against Google's policy and it is easy to detect from the outside.

Buying reviews is a faster way to lose a listing than to rank one. Patterns of purchased reviews are recognisable, and the result is usually removal of every review rather than the fake ones only.

Incentives are the grey one that is not actually grey. A discount for a review is against policy whether or not the review is honest.

Google weighs how recent reviews are, so a business with fifteen from this year usually outperforms one with sixty that all stop three years ago.

This is the reason a long-established business can be beaten by a newer competitor. The reputation is real, but the signal is stale, and the signal is what the map reads.

It also means a burst is worth less than a trickle. Twenty in a week looks like a campaign, which is exactly what it is, whereas one a week for twenty weeks reads as an ordinary busy business.

For a Newcastle business in a tight suburb like Charlestown or Adamstown this is often the whole contest. Everything else about the competing listings is nearly identical, and the one with recent reviews takes the position.

There is no universal number. The only useful target is what the businesses currently above you hold, and in Newcastle marketing that runs from 57 to 134 across the visible competitors.

Measured on this market: Gorilla 360 at 134, Hunter V Digital at 100, Redback Solutions at 73, Naked Digital at 71, Black Lion at 57. A business sitting at eleven is not slightly behind that group, it is in a different bracket of credibility.

The same exercise works in any trade and takes ten minutes. Open the three businesses above you on the map, write down their counts and their most recent review date, and you have a real target instead of a feeling.

Recency does as much work as the total. A profile with 90 reviews where the newest is from last year reads worse to a person than one with 40 where the newest is from Tuesday, and a good share of readers check.

What to write down about each competitor

• The total.

  The bar you are actually clearing, not a round number somebody suggested.

• The date of the newest one.

  Whether they are still asking, or stopped.

• What the recent ones say.

  The specific thing customers keep praising is the thing you are being compared on.

One caution on the arithmetic: a competitor with several hundred reviews has usually been trading far longer or serving far more customers, and matching them is not the goal. Clearing the businesses immediately above you on the map is, and that number is almost always smaller than it looks.

Relying on somebody to remember at the end of a busy day produces about one ask in eight, and it is always the same easy customers. Every business that gets reviews consistently has removed the remembering from the process.

The two variables that decide the reply rate are timing and taps. Sent at the moment the customer knows the job went well, and opening the review box directly rather than a profile they have to search, the same message performs several times better.

The failure everybody shares is asking at the point of payment. The customer is reaching for a card and the request reads as a second invoice, which is why asking at the counter converts so much worse than asking a few hours later.

Timing has a second half that gets ignored, which is the gap between the job and the message. Same day is right for work whose result is immediately obvious, and a fortnight is right for work the customer cannot judge until it has had time to prove itself. Sending both at the same delay wastes one of them.

The last piece is asking everybody. Selecting for the customers who seemed pleased breaches Google's policy, and it also produces a page with no ordinary reviews on it, which readers notice.

A note on the ones you would rather not have. A single critical review among thirty ordinary ones does almost no damage and is quietly useful, because a page of unbroken five stars reads as arranged. What does damage is a critical review answered defensively, or not answered at all.