Search Trustpilot by company name or category and get every match back as a row - rating, review count, claimed and closed status, category, email, phone and address. Up to 100 companies per search term by default.
We ship two Trustpilot services and they return identical fields. The difference is entirely about what you already know.
The Trustpilot Scraper is a lookup. You give it a domain, a profile link or a category URL - things that identify a business you've already found - and it resolves them.
This is a search. You give it a company name or a category as a search term and it finds the matches for you. That's the tool when you don't have a domain list yet and the question is "who's out there in this space?" rather than "what does this company's profile say?"
Because the output is the same seventeen fields, anything you build on one works unchanged on the other. Many teams search first to discover a market, then use the lookup service later to re-check specific businesses on a schedule.
A search term is the input - a company name or a category.
Terms behave like you'd type them - a category such as real estate, or a company name.
This is the most important thing to understand before you queue a large run, because it's where the cost comes from.
Unlike a lookup, one search term is not one row. The default is up to a hundred companies per term - so ten terms is potentially a thousand rows, not ten. Lower it when you only want the top of each market.
A skip setting lets a later run start further down the same result set, so you can work through a large category in stages instead of pulling it all at once - or pick up where a previous run stopped.
The term limit is generous, which is exactly why the per-term setting matters. Combine the two thoughtfully rather than accepting both defaults on a wide keyword list.
Because row count is terms multiplied by results-per-term, it's easy to queue far more than you intended. The estimator prices the job before it runs - read it on this service in particular.
Search accepts optional enrichments, so a single job can go from a keyword to a contactable, verified list.
The documented options cover contact and lead discovery, email validation, and company website finding - the same ground as Leads & Contacts Enrichment and the Email Address Verifier. Layered here, one run turns "marketing consultants" into named contacts whose addresses have already been checked.
Whether that's the right shape depends on timing. Enriching in place means one job and one export instead of three and a pair of joins. Running the dedicated services separately gives you control over when each step happens - which matters most for verification, since a mailbox check is only as good as the day it was taken. If you're sending soon after the search, enrich in place; if the list will sit for weeks, verify separately just before you send.
A real search export returns exactly the same shape as the lookup service - so a pipeline built for one reads the other unchanged. Only the way in differs.
rating.Contact columns fill better here than on a category lookup. Across 200 rows of real search output, email arrived on 71% and phone on 55% - against 31% and 30% in the category export on the lookup page. category and location are also not guaranteed, at 91% and 96%.
All three start without a list, which is what separates this from the lookup service.
Entering a market usually starts with not knowing who's in it. A category search returns the companies with a Trustpilot presence, ranked by how customers rate them - which is a more useful first map than a search-engine results page, because reputation and review volume come attached.
Search a category, filter to low ratings or unclaimed profiles, and you have a list of businesses with a visible problem - plus the email and phone to reach them in the same row. Add the contact enrichment and the list arrives ready to work.
Searching your own company name surfaces profiles that reference it - including near-matches and lookalikes you may not know exist. Worth running periodically if your brand is one people copy.
No subscription, no minimum, no per-seat licence. Your first 500 rows are on us - after that, pay-as-you-go.
Every new account, one-time. No credit card required. Particularly worth using here to see how many companies a single term actually returns before you queue a keyword list.
Roughly $2 per 1,000 businesses. Remember the arithmetic on this service is terms × results-per-term, not one row per term - so the per-term setting is the main cost lever you have.
Volume pricing, SLAs, dedicated workers and bespoke onboarding for teams mapping many categories on a schedule. Tell us your numbers and we'll quote.
LIVESCRAPER10 at checkout.Search gives you the list. These two are what teams run on it next.
The lookup counterpart, returning these same seventeen fields from a domain or profile link. Once search has found the businesses, this is how you re-check specific ones on a schedule.
Feed the websites you've just discovered into twenty firmographic columns - industry code, headcount, revenue, founding year. Reputation plus firmographics qualifies a list properly.
The same comfortable footing as the lookup service, for the same reasons.
A review platform's search results and the business profiles behind them are published so customers can find companies and judge them. Collecting publicly available business information for research is long-established practice, and as long as the data is publicly available and the process doesn't disrupt the site there are no federal laws prohibiting it.
There's no personal data in the seventeen fields - the contact details belong to the organisation - and because this service returns a review count rather than the reviews, no reviewer is named or quoted anywhere in the output. That keeps it clear of the questions that attach to user-authored content and to anything identifying individuals.
The platform's terms restrict automated access, so that remains a terms question as it does everywhere. We touch nothing behind a login, run no third-party trackers on the data layer, and your exports auto-delete after 30 days.
The questions we hear most. Anything else? Talk to us - humans, not bots, write the answers.
profile_image is protocol-relative - it begins with // instead of https://. Fine inside a web page, broken in a spreadsheet or a server-side fetch. Prepend the scheme. Also test for both null and empty string when checking for missing values; the documented response uses both in a single row.Enter a category or company name and get every match with rating, status and contact details. Your first 500 rows are free - enough to see how big a category is before you commit.
Every market-research exercise begins with the same gap: you cannot enrich a list you do not have. The Trustpilot Search Scraper closes it by taking a search term rather than an identifier. Enter a company name or a category - up to 1,000 terms in a single run - and each term returns the businesses matching it, every one carrying seventeen fields: name, rating, review count, the categories it is listed under, claimed, closed and temporarily-closed flags, its website, profile URL and logo image, plus email, phone and a split address of street, city, postal code and country. The output is identical to the lookup service, which means a pipeline built for either reads the other without modification.
The distinction between the two Trustpilot services is worth being precise about, because it determines which one answers your question. The Trustpilot Scraper is a lookup: you hand it a domain, a profile link or a category URL for a business you have already identified, and it resolves them. This is a search: you hand it a term and it finds the matches for you. In practice teams use both in sequence, searching to discover a market and then using the lookup service to re-check specific businesses on a schedule once they know which ones matter.
One number dominates how this service behaves and what it costs. Results per term default to a hundred, so a search term is not a row - ten terms at the default is potentially a thousand rows rather than ten, and the arithmetic is terms multiplied by results-per-term. Lowering the per-term setting is the main cost lever available, and it is the right move when you only want the leaders in each category rather than the full tail. A skip setting complements it by starting a run further down the same result set, which lets you work through a large category in stages or resume where a previous job finished. Because it is easy to queue far more than intended, the estimator matters more on this service than on any one-row-per-query tool, and spending a few of the one-time 500 free rows on a single term is the cheapest way to learn how large a typical category is.
Three format details carry over from the lookup service and bite the same way. The logo URL is protocol-relative, beginning with a double slash rather than a scheme, so it must have https prepended before a spreadsheet or server-side fetch touches it. Missing values arrive in two different shapes, since the documented response shows the profile URL as null and the phone as an empty string within the same record - check for both. And the categories field holds an array rather than a single value, so flatten it deliberately when exporting. Search also accepts optional enrichments covering contact and lead discovery, email validation and company website finding, which can turn one job into a contactable, checked list; the timing caveat is that a mailbox check is only as good as the day it was taken, so if the list will sit for weeks before you send, run verification separately just before sending. On the legal side this sits at the lighter end of the catalogue: business profiles are published so customers can find them, there is no personal data in the seventeen fields, and because only a review count is returned no reviewer is named or quoted. The platform's terms restrict automated access as they do everywhere, nothing behind a login is touched, and exports auto-delete after 30 days. Start free: your first 500 rows cost nothing and need no credit card.