Everything, answered properly

It turns a business category plus a geography into a spreadsheet of contactable businesses. You pick something like "dentist" and somewhere like "Austin" — or a whole state, or a whole country — filter down to the ones worth calling, and export 64 columns of contact data: name, main and secondary types, full address, coordinates, timezone, phone with a mobile/landline classification, email, website, reviews, working hours and every social profile we could resolve.

Under two minutes for a city-sized search, because nothing is scraped while you wait. The data already sits in our index; a search is a query against it, not a crawl. A country-wide extraction with several hundred thousand rows is queued and delivered as a file — you get an email and a webhook when it lands rather than a spinner.

No. An email address and a password is the whole of it. We do not ask for a card, we do not place a hold on one, and there is nothing to remove afterwards. The trial runs for 14 days with 100 lead credits and every feature unlocked — country-wide search included, the same as on a paid plan. On day 14 it simply stops: nothing converts, nothing is charged, and everything you exported stays yours.

Enough to check the data against a market you already know, which is the honest use of a trial. Pick a category and a town where you can recognise the businesses, export 100 rows, and ring three of the phone numbers. If the data is right you will know in ten minutes; if it is wrong you have lost nothing.

The product is a web app — nothing to install, and it runs the same on macOS, Windows and Linux. There is also a REST API for everything programmatic.

Yes. Your plan sets how many people are licensed to use it — one on Starter, five on Growth, unlimited on Scale — and credits are pooled across the account rather than split per person. Per-user logins are not built yet, so today your team shares the one sign-in. That is allowed within your seat count; it is a licensing term, not something the software polices. Enterprise gets SSO.

Google Maps, plus the businesses' own published websites and the social profiles those websites link to. That is the whole list — one map, read properly. Every row carries the Google ID it is keyed on and a Link straight to the public listing it was read from, so any row in an export can be checked against its source in one click. Addresses found on a business's own website additionally carry the URLs they were found on, inside All emails.

No, and this is the single biggest architectural difference between us and a browser automation tool. We crawl ahead of demand into our own index and re-crawl on a rolling schedule. Your search queries that index. That is why a whole-country extraction is possible at all, why results are instant, and why two customers searching the same city do not cause two crawls.

Re-crawl frequency is weighted by category volatility. Restaurants, salons and retail — where a third of listings change in a year — come round far more often than notaries or funeral directors. Rather than ask you to trust an average, every row carries a Link to the listing it was read from, so you can check any row against the live source yourself. First seen on records when a business entered the index, which is a different fact and is filterable before you export.

However many that market has — the index fills on demand. Search a category and a place, and if we already hold it you get the rows straight away; if we do not, we go and get it from the map while you wait, usually in about three minutes, and email you when the list is ready. Either way the count you see is the true count for that search, shown before you spend a credit, so you never pay to discover a market is thin. If you want the standing total anyway: 216,497 establishments, 137,388 of them in the United States.

Records are matched on name, coordinates, phone and domain, and merged into a single row. Where two readings disagree, the most recently read value wins and the alternatives are retained internally. This matters more than it sounds: a busy restaurant is routinely pinned twice under two spellings, and you should be billed for it once.

From the business's own published contact points: its website, its map listing, and the social profiles it links to. We do not guess patterns. If a business publishes no address, the Email column comes back empty rather than invented — a blank cell is worth more than a plausible-looking address that bounces.

There is no single true/false verified column, because one boolean flattens several different facts into the one people over-trust. The signals are published per address inside All emails instead: whether the domain has a valid MX record, which provider runs the mail, whether it is a consumer webmail or a known disposable domain, and the URLs the address was published on. None of that means somebody reads the mailbox. Treat it as a strong signal, not a guarantee.

Every one, on every plan including the trial. No country list to check against and no region behind a tier — the cheapest plan reaches exactly what the most expensive one does. Two markets are already filled in and instant: the United States, which is by far the largest we hold, and Norway, where we started. Anywhere else we fetch from the map when you first search it, which takes about three minutes and then behaves like the rest. Either way the count in the app before you export is the real number for that exact search, not an estimate, and looking costs nothing.

137,388 US establishments, across 2,332 cities in 45 states, under 1,589 different business categories. That is the number in the warehouse right now, not a projection of the market. A real search for a single category in Austin returned 5,714 rows.

It is also not a ceiling. If your category and city are not in there yet, searching for them fetches them from the map in about three minutes, and you can watch the rows arrive before you spend anything.

Yes, and for a web designer or a marketing agency it is usually the only filter that matters. 23.6% of the US businesses we hold have no website at all — roughly one in four. They have a phone number, an address, reviews and hours, and nowhere to send anyone. Turn the filter on and the count you see is exactly how many prospects that is in your area, before a single credit moves.

Of the US businesses we hold: 93.4% have a phone number, 76.4% have a website, and 28.8% have an email address already stored. The rest are enriched when you ask for them — we go and read the business's own site.

Do not plan a campaign off those averages, though. Coverage swings hard by category, and the product makes it exact for free: switch on "Has email" and the number on screen is how many rows in your search carry an address. Every exported row then has one. Check your own category before you pay for anything.

Name, main type and all types; Google ID and a link to the listing; phone in national and international form plus its type; email; website and root URL; eleven address columns from full address down to country code; longitude, latitude and timezone; reviews rating and count; closed status and working hours; six social links; website title, meta and language; five contact pages plus the full list; the email set split by role — individual with first and last name, contact, sales, marketing, finance and administration — plus every address as JSON; six columns holding every social profile found; website technologies; ad pixels; and first seen on. Download the sample file and look at it yourself. The website-derived columns are the part of the schema still being rolled out — the header is always there, the value is empty until we have read it.

We deliberately do not publish a single accuracy percentage, because it would be meaningless across 4,000-odd business types and every market we index, and every vendor's number is self-graded. What we publish instead is per-row provenance: the Google ID the row is keyed on and a link straight to the public listing it was read from, so you can check any row against the live source in one click. Sample your own market during the trial and grade us on it.

They stay in the index with permanently_closed set, and are excluded from exports by default. You can opt them back in — competitive analysis and market-churn research both need them — but you will never be silently charged a credit for a business that shut last year.

We collect only information a business has published about itself for the purpose of being contacted: its listing on a public map, its own website, and the social profiles it links to from that website. Nothing is taken from behind a login, nothing is bought from a data broker, and no personal inbox is ever inferred. Collecting publicly available business contact information is lawful in the EU, the UK and the US; what is regulated — heavily — is what you then do with it, which is the next few answers.

Two hard limits we hold ourselves to: we do not knowingly index sole traders' residential addresses as business addresses, and we honour a removal request from a business within 30 days regardless of whether the requester has a legal right to make one.

Yes. CAN-SPAM is an opt-out regime, not opt-in: you may email a business without prior consent, provided the header and subject line are not deceptive, the message is identifiable as an advertisement, it carries a valid physical postal address, and the unsubscribe mechanism works and is honoured within ten business days. Penalties are per email, so the arithmetic gets bad fast.

Phone is a different statute and much sharper. The TCPA restricts autodialled and pre-recorded calls and texts, and the federal and state Do Not Call registries apply to numbers that belong to individuals — including plenty of sole proprietors whose cell is their business line. Scrub against the DNC registry before dialling; the phone_type column tells you which numbers are mobile.

The CCPA as amended covers business contacts as well as consumers, so a Californian sole proprietor in your list has the right to know what you hold, to delete it and to opt out of its sale or sharing. Publicly available information is excluded from the definition of personal information, which covers much of what a map listing contains, but that exclusion is narrower than people assume once data is combined. If you market into California, publish a privacy notice, run an intake for deletion requests, and answer them.

Sometimes, and you should assume yes. info@thebakery.com is not personal data — it identifies an organisation. marie.dupont@thebakery.com is, because it identifies a person, and so is a sole trader's mobile number, because a one-person business is a person. GDPR applies to the second kind, and we treat every record as though it might be. That is why provenance is attached to every field rather than to the file as a whole.

For B2B prospecting the usual basis is legitimate interests, GDPR Article 6(1)(f). It is not automatic: you have to run and record a three-part balancing test — that the interest is legitimate, that the processing is necessary for it, and that it does not override the rights of the individual. Direct marketing is named in Recital 47 as a legitimate interest, which helps, but a recorded assessment is what an authority will ask for. Do not rely on consent you do not have; a published address is not consent.

Article 14 then applies because you did not get the data from the person: you must tell them where it came from, generally within a month or at first contact. In practice that is a line in your first email plus a linked privacy notice naming FindAndClose and the public map listing as the source. The Link column, and the per-address sources list inside All emails, exist precisely so you can make that statement truthfully.

It depends on the member state, and this is where most people get it wrong. The ePrivacy Directive requires prior consent for unsolicited email to natural persons and leaves B2B to national law, so the rules genuinely differ across the single market. Several states — Ireland, the Netherlands, the Nordics — permit unsolicited email to a corporate address with a working opt-out. Others are far stricter: Germany's UWG effectively requires prior consent for email even B2B, and Italy has taken a restrictive line too. Treat "EU" as twenty-seven regimes, not one, and check the country before the campaign.

Wherever you send, the mechanics are the same: identify yourself and your company, say why you are writing, offer a one-click opt-out, and honour it permanently across every list you own.

We are an independent controller for our own index — we decide what to crawl and how long to keep it. Once you export a file, you become the controller of your copy, and you decide what it is used for. That distinction matters when a request arrives: deleting a record from our index does not delete it from your CRM, and we cannot delete it for you. We sign a DPA where the relationship warrants one; it is available on request on any plan.

Send them to our removal form, or forward the request to us. We suppress the record within 30 days and add the identifiers to a suppression list, so a later re-crawl does not quietly reinstate it — a deletion that undoes itself on the next pass is not a deletion. If you already exported that business you must also delete it from your own systems; we have no access to them. Suppressed records do not consume anyone's credits.

No, and we do not intend to. The index contains businesses. We do not build consumer profiles, we do not append demographic or intent data from third-party brokers, and we do not guess personal mailbox patterns from a company domain. If you want firstname.lastname@ guesses, another vendor sells them; we would rather hand you a blank field.

Not under the standard terms. Exports are licensed for your own sales, marketing and research use — not for redistribution, resale, or as training data for a model you then ship. If that is what you need it is a different licence and a conversation with us, not something any of the four plans covers — and we would rather have it up front than discover it later.

No. It is a plain-English summary written by the people who built the index, current as of July 2026, and it cannot account for your jurisdiction, your sector or your specific campaign. Where the stakes are real — a large outbound programme, a regulated industry, a market you do not know — have your own counsel review it. We will happily answer their questions about how the data is sourced and what we retain.

One business, exported once, in a rolling 30-day window. A fully populated row and a sparse one cost the same single credit. Re-exporting the same business inside 30 days is free, and previewing results in the app costs nothing at all — only the export deducts credits.

Starter is $49 a month for 10,000 lead credits, Growth $99 for 50,000, Scale $249 for 250,000, and Enterprise is quoted from 1,000,000. Annual billing takes 20% off every rate — $39, $79 and $199 a month respectively, which is $120, $240 and $600 saved over a year. What rises with the price is the credit allowance, not the reach: every plan searches every market we index, a whole country in one pass, any radius and any polygon. The calculator on the pricing page will size a specific market against all four.

On Starter they do not — the allowance resets on your billing date. On Growth unused credits carry forward one month, on Scale three months, and on an Enterprise contract twelve. A carried balance is only drawn on once the current month's fresh allowance is spent, so rollover extends a quiet month rather than doubling a busy one.

Exports stop and nothing is billed. There is no overage charge. You are warned in the app at 80% and 100%, search and filtering keep working, and the export endpoint returns 402 credits_exhausted until you upgrade or the month resets.

Within 14 days of a first payment, if you have used under 20% of that period's credits. Annual plans cancelled after 14 days run to the end of the term rather than being refunded. If an export came back materially broken we credit it regardless of the window — that is a defect, not a change of mind.

Card on every plan; bank transfer against an invoice or a purchase order on Enterprise. All prices are in US dollars and exclude tax — EU and UK businesses with a valid VAT number are reverse-charged, everyone else is charged at their local rate.

Growth and above. Programmatic access is one of the few things the tiers actually separate: the REST API switches on at Growth at 10 requests a second, rises to 50 on Scale and 200 on Enterprise. Starter is the app. What the API never changes is reach — it runs on the same account, the same credits and the same markets the app already gives you.

A conventional REST API over HTTPS with bearer-token auth, cursor pagination and idempotent export creation, so a retried request cannot double-charge your credits. A search is a GET /api/v1/map/search with the same filters the app exposes; an export is a POST that returns a job you poll or a webhook you receive. Rate limits are published in response headers rather than discovered by being cut off.

Two ways, both from Growth upwards: outbound webhooks fired when a scheduled export finishes, or the REST API if you would rather pull. There is no bespoke one-click CRM connector — a webhook into your own automation is more durable than an integration that breaks when a vendor changes an endpoint.

Yes, from Growth upwards, alongside the API and webhooks. Each run can be set to return only businesses that are new or changed since the last one, which is usually what you want and is far cheaper in credits than re-exporting the whole market every Monday.