Google Maps business data, done through the licensed route
Local business data is the most requested thing on this list and the one where I turn down the most work. Not because it is hard to collect, but because most people asking want it collected in a way that breaks Google's terms and puts their own product at risk. There is a licensed path that works, costs money per request, and produces a dataset you can actually build a business on. That is the version I build.
- Starting at
- from £1,200
- Refresh
- Weekly, or monthly for a stable territory
- Access
- licensed
What comes out of it
- Business name, address, and geolocation
- Category, opening hours, and price level
- Rating, review count, and review text where licensed
- Phone number and website where published
- Stable place ID for long-term matching
What makes Google Maps business listings awkward
Coverage is the real problem. The API returns a bounded number of results per query, so a city is not one request, it is a grid of overlapping searches tuned to local density, with deduplication by place ID afterwards. Get the grid wrong and you silently miss businesses in dense areas while burning budget on empty squares. The second issue is matching listings to your existing records, where names and addresses rarely agree exactly.
The Places API charges per request and forbids storing most fields beyond 30 days, with place IDs the notable exception. That single constraint dictates the whole architecture: you cache IDs permanently, refetch details on demand, and design the product so it never needs a full local mirror. Teams who ignore it build something that works in testing and violates terms in production.
What this is good for and what it is not
It works well for territory analysis and market sizing: how many businesses of a type operate in an area, how they cluster, and how that compares with somewhere else. It works well for enriching a list you already have, matching your customers to their listings so you get categories, hours, and ratings without manual lookup.
It works badly as a contact scraper. Phone numbers and websites are published for many listings, but building a cold outreach list from Maps runs into both the terms and, in most of Europe, data protection rules on unsolicited contact. If that is the goal, mention it in your enquiry and I will tell you honestly whether there is a compliant route or whether you would be buying a problem.
It also works badly as a review mining tool at scale, since review text has the tightest licensing restrictions of anything the API returns.
How the build works
First the grid: your target area is decomposed into search cells sized to expected business density, with denser cells subdivided until each returns fewer results than the API cap. This is where budget is won or lost, because a naive uniform grid over a large region costs several times what a density-aware one does for the same coverage.
Then collection, with place IDs cached permanently and detail fields fetched only for listings that pass your filters. Compliance with the caching rules is built into the storage layer rather than left as a policy nobody enforces, so the retention behaviour is a property of the system instead of a promise.
Then matching, where listings are reconciled against your existing records using name, address, and geolocation together, with a confidence score surfaced so uncertain matches get reviewed instead of silently merged.
Cost, including the part that is not mine
Workflow automation starts from £1,200 for a bounded extraction from one permitted business directory into a spreadsheet or database. Territory coverage, scheduled refresh, CRM matching and a review interface are quoted separately.
Separately, you pay Google per request. For a mid-sized city that is usually tens to low hundreds of dollars for an initial build, and less for refreshes since cached IDs mean you refetch details rather than rediscovering listings. I estimate this before the build so it is a decision rather than a surprise on your first invoice.
If those running costs make the project uneconomic, that is worth finding out in a twenty minute call rather than three weeks into a build. I would rather tell you the numbers do not work than take the money.
Where the line is
Licensed data access, arranged with the provider.
Scraping the Google Maps interface directly breaks Google's terms of service, and I do not build that. What I do build runs on the Places API under its licence, which permits the use and charges accordingly. This matters beyond principle: products built on scraped Maps data have been shut down after acquiring customers, and the licensing cost is far smaller than discovering the problem once you have a business depending on it. I will quote the API costs alongside the build so the running cost is clear upfront.
MathersLabs is not affiliated with or endorsed by Google Maps business listings. Work is scoped to publicly accessible data and to compliance with applicable terms and law, and feasibility is assessed per engagement before any fixed price is agreed.
Related work
SignalSauce
250 qualified leads a month, fully automated
Built an end-to-end intelligence pipeline that ingests from 5 curated news sources twice daily, scores every announcement against a custom algorithm, and enriches the ones that pass with 21 structured data points per company.
Questions people ask first
Is your source on this list?
Tell me what data you need and where it lives. I will tell you whether it is feasible, what it would take, and whether it is worth building before you commit to anything. Working hours overlap with US Eastern and Pacific.