How we read the data

Google’s public business listings for pest control across the United States and Canada — 17,520 businesses in 2,052 towns and cities, captured 2026-08-24.

This is not an exhaustive register. It is the set of pest control businesses Google surfaces for each place, capped at around twenty per locality. A company can be perfectly good, licensed and busy and not be here. We describe these pages as what Google surfaces, never as every business that exists.

A place page lists the businesses Google surfaces when you search that place. That is not the same as the businesses whose office is inside its boundary, and for a trade where companies drive to you it should not be.

Around 43% of listings that carry a street address show a different town from the one they are filed under — a firm in Edmonton covering Sherwood Park, a branch in Saint John covering Moncton. So a listing says a business serves a place, never that it sits in it, and every listing shows its own address so you can judge the distance yourself.

Google collects a full five-bucket breakdown of every business’s reviews and displays a single average. Those are very different objects. Two businesses can both average 4.9 while one has almost no critical reviews and the other has hundreds.

Across the whole corpus, 3.7% of 5,410,348 reviews are one or two star. That figure is taken from Google’s own histogram over every review, never from the subset of review texts we hold — our held sample is selected and skews heavily positive, so a rate computed from it would be flattering and wrong.

Google derives topics from the text of reviews and counts how many reviews mention each one. When a listing shows “termites 41”, that means forty-one reviewers wrote about termites.

It is a measure of volume and subject, not of quality. A business named often for a pest does a lot of that work; it does not follow that it does it better. We rank pest pages on that count because it is the honest available signal for specialism, and we say so on every one.

Lists are ordered by a rating adjusted for how much evidence sits behind it. A raw average treats a 5.0 from four reviews as better than a 4.8 from nine hundred, which is obviously wrong.

We shrink each rating toward the corpus mean (4.772) in proportion to its review count, with a prior weight of 50 reviews. A business with 50 reviews sits halfway between its own average and the corpus average; one with two thousand sits essentially on its own. No business pays for position and there is no advertising on this site.

Where a business has a website, we record the claims it makes there — licensed, insured, family owned, 24/7 emergency cover, and so on. Those appear on listings marked as self-reported.

They are marketing copy, not verification. We have not checked any of them against a licensing board and we never present them as a tick. Applicators are licensed in every state and province, so the absence of a “licensed” claim on a listing means we found no such sentence on that company’s website — it is not evidence that anyone is operating without a licence, and we do not style it as one.

We looked. Pest control licensing in North America runs through structural pest control boards and departments of agriculture, fragmented state by state, and the public-record signal available to us covers under 1% of the corpus. When we inspected those matches they turned out to be name collisions — a recreation department, a transport department, a mosquito abatement district — rather than licence records.

So this site makes no claim whatsoever about any business’s licence status from public records. Checking a licence is worth doing and it is worth doing at the source; your state or provincial board is the place.

We hold review texts for most listings and use them for exactly one thing: counting how often the owner replied. We do not republish the reviews themselves. They are other people’s words on someone else’s platform, about a fifth of them are ratings with no text at all, and reprinting a selected fifth of them at scale would misrepresent the whole.

Where a reply rate is shown it is always over the texts we hold, and the listing always names both numbers — what we hold and how many reviews exist. “Based on 412 reviews” while holding twenty would be a lie.

Opening hours are on 92.2% of listings and they are a durable fact. The source also carries a live-looking status string — “Open · Closes 5 PM” — captured at a single moment. Every page on this site is a static file built ahead of time, so that string would be wrong the moment it was published. We do not carry it and we never render it.

It means we did not collect it — not that the value is zero and not that the answer is no. Those are different things and conflating them is the easiest way for a directory to lie. Wherever a listing is missing something, it says so in words rather than showing an empty space.

3,677 records in the source are not businesses a consumer can hire — county mosquito abatement districts, municipal animal services, wildlife rehabilitation charities, and a scattering of retailers and visitor attractions that share a category with pest control. They are excluded rather than listed.

If your local vector control district treats for mosquitoes, that is often free and worth knowing. It is simply not something this site should present as a company to call for a quote.

The corpus is still being collected and every count on this site changes as it grows. Nothing here is hard-coded: each figure is rendered from the snapshot named beside it. If a number looks stale, check the capture date in the provenance line under it.

Write to help@exterminatorguild.com and it will be looked at. See for business owners.