Data vendors sell a snapshot
You get a name and an email. You cannot see where it came from, or when. So you cannot tell if it is still true.
Enter your website. We draft your ideal customer profile, find companies that match, and crawl their public sites. Every cell carries its source, time, confidence and license, and blanks stay honest.
See how it works ↓Free plan includes 200 reveals a month. No card needed. No website yet? Describe what you sell instead.
| Company | Fit | Contact | Signals | Public revenue |
|---|---|---|---|---|
| Harbor & Pine Buildersharborpine.example | — | not looked yet | not looked yet | not looked yet |
| Northwind Accounting LLCnorthwind.example | — | not looked yet | not looked yet | not looked yet |
| Cedar Row Logisticscedarrow.example | — | not looked yet | not looked yet | not looked yet |
Fictional companies and people, shown for illustration. The real thing looks exactly like this, with your data.
They just added 40 new products. They posted a job for the exact role you sell to. None of that is in a contact database. All of it is on their website, right now.
You get a name and an email. You cannot see where it came from, or when. So you cannot tell if it is still true.
Ask an AI tool a question and it always answers. Even when it does not know. That is the problem.
Copy, paste, check, repeat. Every week, by hand.
We only write down what we actually saw, on a page we can link to. Four rules come from that.
Click any answer. See the page it came from and the day we read it. Your export has it too.
Sometimes the answer is not out there. We say so, instead of guessing.
We show revenue only when the company says it out loud. A guess is labelled a guess.
We read a company once and share what is public. You never pay for the same page twice.
An evidence-grade GTM warehouse is a lead database where every value is stored with the URL it was read from, the time it was seen, a confidence and a license — so a sales team can defend any row in it. VisibleTrust is one: you describe who you sell to, it finds companies that match, reads what they publish about themselves, pulls the signals that say they are buying, scores them against your profile, and lets you reach the right person from your own accounts.
Blanks stay honest. Estimates are labelled. Every cell can answer “where did this come from, and are we allowed to have it”.
Data processor, not a data brokerYour accounts send, we never dorobots.txt respected, always
We read four places, in this order, and stop as soon as we have the answer.
first party · free
Their about page, team page, careers page and contact page. Free, and first-hand.
first party · free
If they run a shop, it tells us their city, their currency, and how much they sell.
CC0 · statutory · free
Company registers, job boards and news. Open data, free to use, easy to cite.
gated · optional
Only if the first three came up empty. Paid sources wait behind a paywall until you say yes.
No scraping of LinkedIn, Google Maps or bought lists. Those sources are declined in the registry, with the reason attached.
Messages go out from your own LinkedIn, your own inbox, your own X. Never ours. When someone replies, everything stops.
sends from your connected account
Best for a first hello to founders. We keep to safe daily limits and stop the moment LinkedIn says slow down.
sends from your own mailbox
Best for detail and follow-ups. We check every address works before we use it.
sends from your connected account
For brands that live on X. Same limits, same evidence.
Instagram and Reddit are parked, not scraped. They arrive the same way: through an account you own.
Paste your website. We read it and write down who you sell to, quoting your own words back.
They're hiring. They just raised money. They launched something. Each one dated, with a link.
Ask one question about every company at once. The answer must quote the page, or it says not found.
“Hi, I came across your profile and thought you might be interested in our solution for growing brands…”
“You listed 38 new SKUs in the last quarter and you're hiring an ops lead — the two things that usually break returns handling first. We fix that for brands your size in Bengaluru.”
Every fact in the draft is one the book can show you the source of.
Six steps. They run once, then keep running every night.
Paste your website. We read your homepage and quote it back to you.
We write down who you sell to. You edit it until it is right.
We find companies that match. Every one is checked before it reaches you.
We read their site, their shop, then open data — until the blanks are filled.
We draft messages using only facts we can prove. You approve them, then they send.
We re-check the ones you care about every night, and tell you when something moves.
Watch the accounts that matter. Every night they are re-checked for new signals, re-enriched when a collector has improved, and re-ranked. When something moves, you get a digest — or a webhook.
Every source you rely on records what it found on real traffic, so the platform's own coverage numbers correct themselves as it runs.
Describe what you sell, get a book, reach the first fifty yourself.
Catalogue size, price band and launch cadence are the signals you actually qualify on.
One workspace per client, evidence they can forward, exports with the sources on.
Consent basis per cell, a deletion path per person, no bought lists anywhere in the pipeline.
Research columns, webhooks, inbound ingest, API keys. Sit in the middle of your stack.
If a vendor ever handed you a number you could not defend, this is the opposite of that.
| Contact database | AI prospecting tool | VisibleTrust | |
|---|---|---|---|
| Where a value came from | Provider name | Not shown | URL, time, license, consent |
| When it cannot find something | Stale row | Plausible guess | “Not found on public web” |
| Revenue | Estimated as fact | Estimated as fact | Public statement, or a labelled estimate with its method |
| Research questions | — | Answer | Answer + verbatim quote, verified on the page |
| Who sends | You, elsewhere | Their infrastructure | Your accounts, your caps |
| Erasure request | Ticket | — | One call, receipt logged, address suppressed |
$0
$49/seat/mo
$99/seat/mo
$249/mo + $49/seat
Allowances are per workspace. Viewers are free. Research columns and vendor lookups spend enrich steps. Prices in USD, taxes may apply.
Public company websites and storefront endpoints crawled by our bot (which respects robots.txt); open data such as Wikidata, statutory registers, public job boards and news feeds; a licensed search API for public pages about a company; and, if you connect them, your own LinkedIn and X accounts. Every source sits in a registry with its access rule, license and consent basis, and every cell shows which one it came from. We don't scrape LinkedIn profiles or buy contact lists.
It drafts your ICP from your own homepage, suggests companies to look at, drafts outreach that may only cite facts the book holds, classifies replies, and answers research questions — with a quote we then verify verbatim on the page. Every suggested company is verified by crawling its homepage; unreachable, parked or mismatched domains are dropped. Nothing the model says is stored as a fact.
Opening a company's contacts, evidence and dossier inside one campaign. It costs one reveal the first time, even if we already have fresh data. Candidates — name, domain, country, fit — are always free.
Only from your own accounts. Connect your mailbox, LinkedIn or X and VisibleTrust drafts, schedules, caps and pauses sequences that go out as you. We are never the sender of record and we run no shared mail infrastructure.
No. LinkedIn's robots.txt disallows crawlers, so we do not crawl it; what we read there we read through your own connected account, under your agreement with LinkedIn. Bought or scraped lists of unclear origin are declined in the source registry with the reason attached.
A question you ask of every account — “do they ship internationally?”, “who runs operations?”. The answer must come with the URL and an exact quote from that page, which we check. Otherwise the cell says not found. One enrich step per account, found or not.
A rubric with stated weights — industry, geography, tech, pain, intent, contact, size — and every point has a reason. Once your own sequences have produced enough replies, the weights are re-fitted to what actually replied and shrunk toward the rubric by sample size. The explanation never disappears.
No. We are a data processor. Personal data never crosses workspaces, paid partner data is bought by you through us, and one call erases what we hold about a person, suppresses their identifiers on every channel and logs a receipt.
Yes: upload a CSV, sync a CRM, or register a source you have rights to along with its consent basis. Rows imported under it stay private to your workspace and carry the source in the lineage drawer.
Yes. Deleting a campaign removes your seller profile, ICP, book, scores and notes. Contact privacy@getvisibletrust.com to close a workspace.
Two minutes later you have an ICP, a book of matching companies, and the evidence to act on it.
200 reveals a month, free. No card. Nothing invented.