I get asked some version of "why not just use Hunter?" fairly often, usually by people who already pay for it. It's a fair question. Hunter is a good product and for a lot of searches it's the right tool. But it works in a particular way, and that way suits some companies much better than others.
So here's an honest split: where a database like Hunter wins, where reading the company's own website wins, and how I'd combine them.
How each approach works
A database (Hunter, Apollo, Snov and others) collects addresses from across the web and other sources over time, stores them, and works out each domain's pattern. When you search a domain, you get what they've already collected plus a guess for anyone else, with a confidence score and a verification check.
Reading the website means visiting the company's own pages at the moment you ask: home, about, team, contact, the legal notice. You get whatever the company itself publishes today, and if there's a named owner but no personal address, you guess from the name.
Both end up guessing sometimes. The difference is where the names and examples come from.
Where a database wins
Companies with more than a few dozen staff. Big companies have many addresses floating around the web: press releases, conference talks, GitHub, old PDFs. A database that has crawled for years will have seen a lot of them and knows the pattern with confidence. A company website for a 2,000-person firm doesn't list staff emails at all.
Searching for a specific person. If you already know you want "Priya Shah, Head of Procurement at Example Corp", a database can return her likely address in one call. Reading the company website won't find a head of procurement; she isn't on it.
Role-based searches. "Show me marketing people at these 50 companies" is a database query. The website approach has no way to do it beyond the founders and leaders a site chooses to show.
Verification built in. Hunter checks addresses against mail servers and tells you the result. That's convenient, and it's a lot of plumbing you'd otherwise need to build or buy.
Polished tools. Browser extension, CRM integrations, a Google Sheets add-on. If your team lives in Gmail and HubSpot, that matters.
Where the website wins
Small, owner-run businesses. A plumber, a dental practice, a 6-person agency. These rarely appear in sales databases in any depth because their staff don't speak at conferences or commit to GitHub. But their own site often says exactly who runs it: "Founded by Maria in 2014", a team page with two photos, a contact page with maria@.
Outside the US. Database coverage is strongest for US tech and weaker as you move away from it. Meanwhile some countries make the website approach unusually good. German, Austrian and Swiss sites must publish a legal notice (Impressum) naming the owner or managing director. In a sample of 48 DACH sites, the owner was in the Impressum on 44.
Freshness. A database record might be two years old. The website is what the company says today. For small businesses that change hands or rebrand, that matters.
Context beyond the email. Reading the site also gives you phone numbers, social links and the tech stack, which help you decide whether to email at all and what to say.
Cost on big lists of small companies. Hunter's public pricing lists the Starter plan at $49/month for 2,000 credits. That's reasonable for targeted work. If you're working through 20,000 Google Maps results for dentists, a per-site approach is usually much cheaper.
How good is the website approach, honestly?
Some numbers from my own testing:
- On 63 small-business sites, a website scraper found emails for 57 of them.
- When the owner's address wasn't published and I guessed
firstname@, checked against 189 real people, the guess was exactly right 70% of the time, 76% when the site showed its email format somewhere. - Among 80 hiring startups, 43 founders had published their address somewhere, but only around 10 on the company site itself; 13 were on Y Combinator pages or Hacker News.
That last point is a real limit. For venture-backed startups, the company website is often a polished marketing page with no names. A database that has picked up addresses from elsewhere may well do better there.
The junk problem
One thing that trips up website scraping, and that a curated database avoids: rubbish addresses. Small-business sites contain the web designer's email in the footer, template placeholders like example@ or your@email.com, and addresses belonging to the site builder or a tracking service. Testimonial sections and investor logos get mistaken for team members. A naive scraper hands you all of it. In my 63-site test, the most-used contact scraper on Apify returned 22 emails belonging to other organisations; mine returned 5. Any website approach you use needs to filter that out.
How I'd combine them
- Selling to companies over ~50 staff, or to a specific role: start with a database.
- Selling to small local businesses, or outside the US: start with the website.
- Mixed list: run the website pass first because it's cheap, then send only the companies with no named owner to a database.
Whichever you use, treat a guessed address as a guess, keep send volumes low, and stop when people ask. CAN-SPAM in the US and GDPR/PECR in the EU and UK both care about how you got the address and whether you honour opt-outs.
The website side, packaged
For the website pass I use Founder Email Finder (https://founders.intropi.in, on Apify). It reads home, about, team, contact and imprint pages, returns the owner's name and title, and labels each email as published on the site (high) or a firstname@ guess on a domain with MX records (low). It's $1.50 per 1,000 websites with contacts, and failed sites aren't charged.
Disclosure: I built it. It doesn't replace Hunter for large companies or people search, and I wouldn't claim it does.
Written with AI assistance; the numbers come from my own test runs.

