Agencies: build a client's prospect list from their rivals' reviews

Agencies have a structural prospecting problem. The businesses you sell to are numerous, individually small, geographically scattered, and almost entirely absent from the databases that B2B tooling is built around. There is no intent platform that will tell you a dental practice in Hamilton is unhappy with its website.

But that practice's customers are telling you, publicly, with a date attached, in the one place every local business is legible: its reviews.

This is the play, in full, including the parts that do not work.

Why reviews are the right surface for this

The BrightLocal Local Consumer Review Survey 2026, run against 1,002 US adults in February 2026, found that 97% of consumers read reviews before choosing a local business, and that 41% now always read them, up from 29% the year before. Standards moved too: 31% will only use a business rated 4.5 stars or higher, up from 17%.

Two things follow, and both matter commercially.

First, reviews are the highest-stakes public surface a local business has. A rating below 4.5 is now disqualifying to nearly a third of the market. That is not a vanity metric any more, and a business owner knows it.

Second, and this is the prospecting insight: a negative review is a customer describing an operational failure in specific, dated, public language. "Nobody answered the phone for three days." "Booked online and they had no record of it." "The website said they were open." Each of those is somebody naming a problem that a particular kind of agency fixes for a living.

You are not inferring a need. You are reading one.

The play, step by step

1. Pick a vertical and a geography, narrowly

The most common mistake is going broad. "Dentists in Ontario" is not a campaign, it is a database. "Dental practices in three adjacent suburbs" is a week of calls where you can credibly say you know the local market.

Narrow also makes the pitch better. A list of nine practices within ten kilometres of each other, all with the same complaint theme, is a market insight you can open a conversation with. Nine hundred practices across a province is a mail merge.

2. Resolve the actual businesses

This step gets skipped and it is where lists go wrong. You need real, resolvable businesses (a name, an address, a place record), not a scrape of search results. Chains, closed locations, duplicate listings and businesses that moved are all common enough to poison a list, and there is nothing worse than opening a call by referencing a review of a location that shut last year.

3. Gate on star rating before you spend anything

The discipline that makes this economical. If you are looking for operational pain, a five-star review cannot contain it. Filtering on rating before any expensive analysis is both a precision and a cost decision, and the same logic applies whether the expensive step is a language model or an account manager reading.

Set the ceiling by what you sell. Web and booking problems show up in three- and four-star reviews as often as one-star ones, because a mildly annoyed customer still gives four stars and then explains exactly what went wrong. If you only read one-star reviews you get the angriest customers, who are frequently describing something nobody can fix.

4. Theme the complaints, do not list them

This is the step that turns a list into a deliverable. Twelve individual reviews are noise. "Seven of the nine practices we looked at have reviews in the last quarter about phone calls going unanswered" is a finding.

Themes also tell you which service to lead with, which is different per geography and is the thing you cannot know in advance. Sometimes the theme is booking friction, sometimes it is response time, sometimes it is that nobody can find the opening hours. Three different pitches.

5. Open with the evidence, not the offer

The reason this play converts better than a list buy is that you can open a conversation with something true and specific: a dated, public statement by their own customer about a problem you fix.

Say it neutrally. "Three of your reviews since June mention people not being able to get through on the phone. Is that something you're already working on?" is a question a business owner will answer. "I noticed your reviews are bad and I can help" is one they will not.

Two ways agencies use this

For your own pipeline. The obvious one. A themed, evidenced call list per vertical per geography, refreshed on a schedule.

As a paid deliverable. The less obvious and often more valuable one. The same analysis run on your client's competitors is competitive intelligence your client cannot easily produce themselves: here is what your rivals' customers complain about, here is where you are strong by comparison, here is the messaging angle nobody in this market is using. That is a recurring engagement, not a one-off audit.

The second framing also solves the agency margin problem, because you are selling the analysis rather than absorbing it as a cost of sale.

Where this does not work

• Verticals with thin review volume. B2B professional services, trades that get work by referral, anything where customers do not review. If the median business in your target vertical has eleven lifetime reviews, there is not enough signal.
• Problems reviews do not describe. Nobody writes a review about a business's accounting software, payroll provider or cyber insurance. Reviews surface customer-facing failures, which is a specific and limited set.
• Markets you cannot service. Obvious, and still the most common way this list gets wasted: a beautifully themed list of businesses four hours away.
• When you cannot move fast. A review describes a problem that was live when it was written. Six months later the business has either fixed it or stopped caring.

What Openpulse does with this

The places listener is built for exactly this shape of work: you give it a vertical and a geography, it resolves the real businesses, reads their Google reviews within a 90-day window (longer than the 30 days used for conversations, because an operational problem moves more slowly than a thread), gates on your star ceiling before spending a model call, and classifies what survives as pain_confirmed, at_risk or healthy.

The deliverable is a roll-up: one row per business, with the themes underneath and the individual reviews as the evidence. Every rejected review keeps its reason, so you can tell "this market is quiet" apart from "my star ceiling is throwing everything away".

On the Go plan that is $49 a month for one listener running daily or weekly, after a 14-day trial. Most agencies want Pro at $499 for five listeners, because five listeners is five vertical-and-geography combinations running at once, plus the competitor tracking for the client-facing version of the play.

The method is covered in more depth in one-star reviews are a qualified lead list, and the category page is local lead generation.