Prospecting by rating and reviews: which businesses need what you sell

Ratings and review counts are public, cover almost every local business, and say a lot about what a business needs. A low or falling rating points to a business with a problem to fix: service training, cleaning, maintenance, staffing or marketing. Very few reviews points to one that is hard to find or new. A high rating with many reviews points to a busy, proud business that may pay for quality and capacity. Decide which of those your offer helps, filter for it, and approach with an offer that fits the problem, never with the rating itself as your opening line.

This piece is for agencies, suppliers, consultants, trainers and software firms who sell to local businesses. It explains the four rating patterns and who sells to each, how to read a rating against its own market, an example, the mistakes that turn this into an insult, and how to have an AI employee build the list each week.

Four patterns and who they suit

Rules of thumb for matching a rating pattern to an offer, not measured results.

Read each rating against its own market

A rating has no meaning on its own; it has to be read against places of the same kind in the same city. Two examples from xMap's places database, November 2026:

• Near the centre of London, 295 of the 392 beauty salons nearest the centre have a rating, averaging 4.7 out of 5, and 246 are rated 4.5 or more. A salon at 4.3 there is in the bottom part of its market.
• Of the 437 supermarkets nearest the centre of London, 423 have a rating, averaging 3.6, and only 38 are rated 4.5 or more. A supermarket at 4.3 there is among the best.

The same 4.3 is a problem to fix in one trade and a strength in another. Always compare a place with the average of its own kind within its own area. In Toronto, by contrast, only 167 of the 312 hotels within 10 km of the centre have a rating at all, so for hotels there a missing rating is common and says less.

An example

A made-up example. A hospitality training company in Jakarta sells half-day service courses to restaurants. It wants restaurants that are busy enough to afford training and are being marked down for service.

1. Find every restaurant in central Jakarta with Nearby places, with its rating and number of reviews.
2. Keep places with at least 100 reviews, so the rating is based on enough customers to be trusted.
3. Look again around each one with Nearby places, passing on each restaurant with the average rating of restaurants within 1 km, and keep those rated more than 0.3 below it.
4. Ask xMap AI to score each from its rating, reviews and kind, and write one line on why it was chosen.
5. Send the best 15 a week to the sales lead.

The company ends up with a short list of busy restaurants that are underperforming against their own street. Its first message talks about the course and a free half-hour visit, not about the rating.

Mistakes that turn this into an insult

• Leading with the rating. "We noticed your 3.6 stars" makes an owner defensive. Lead with what you offer; let the rating stay your reason, not your opener.
• Trusting ratings from a handful of reviews. A 3.0 from four reviews is an opinion, not a pattern. Set a minimum number of reviews.
• Comparing across trades. Supermarkets and beauty salons sit on different scales, as the London figures show.
• Ignoring chains. A chain branch's low rating may be fixed by head office, not by the branch manager you write to.
• Reading reviews you do not have. xMap Flows scores customer opinion from the rating and the number of reviews; it does not read the text of reviews, so do not claim to know what customers complained about.

Watching for a rating that drops

A falling rating is often the best moment of all, because the owner has just noticed it too. When something changes can watch up to five places for a rating dropping below a line you set, and What changed picks out places on your list whose rating moved since the last run.

Doing it with an AI employee

1. Schedule: every Monday.
2. Area of interest: the cities or districts you serve.
3. Nearby places: your customer types, with a minimum rating or not, as the pattern you want requires.
4. Filter: reviews at least 50 or 100, and rating below or above your line.
5. Only new: so a business is not sent twice.
6. xMap LLM: a score and a reason, told which pattern you sell to and the local average to compare with.
7. Email the list to the sales lead, or Outreach with every first message approved before it goes.

What lands in the inbox each Monday is a short ranked list, each with its rating, number of reviews, the kind of place and one line of reasoning. For a lighter version, the template for businesses with low ratings and no website already builds this for agencies.