How to Find Customers of Any SaaS Company

Most SaaS companies list their customers somewhere on their website. Logos on the homepage, a "Customers" page, case studies, testimonials, "Trusted by" sections. It's right there, publicly available.

The problem is every site does it differently. One company shows a logo wall. Another has a dedicated /customers page with detailed case studies. A third buries customer names inside blog posts or press releases. You can't write one scraper that works for all of them.

WebCrawler Agent solves this. You point it at a website, tell it to find the customers, and it figures out where to look. It browses the site, follows relevant links, and comes back with a list.

One thing to be clear about: this only works if the company actually lists their customers on their website. If they don't mention customers anywhere, there's nothing to extract. The agent can't guess or pull data from private sources.

How it works

curl --request POST \
  --url https://api.webcrawlerapi.com/v1/agent \
  --header 'Authorization: Bearer <YOUR_API_KEY>' \
  --data '{
    "prompt": "Find all customers listed on this website. Check the homepage, customers page, case studies, and testimonials. Return a JSON object with a customers array containing company names.",
    "urls": ["https://example-saas.com"],
    "max_spend_usd": 0.5
  }'

The agent starts at the URL you provide. It reads the page, looks for customer mentions, and follows links to pages like /customers, /case-studies, or /testimonials if they exist. When it's done, you get JSON.

Send a POST with your prompt, seed URL, and budget.
The agent browses the site and collects customer names from wherever they appear.
Poll for the result. When status is done, you get your list.

Where companies hide their customers

I say "hide" loosely. They're not hiding them, they just put them in different spots.

Homepage logo wall. The most common one. A row of logos under "Trusted by" or "Used by." Sometimes the logos don't have alt text, which makes traditional scraping annoying. The agent reads the page visually.
Dedicated customers page. /customers, /case-studies, /wall-of-love. Clean and easy when it exists.
Testimonial sections. Customer names and companies embedded in quotes. Could be on the homepage, a landing page, or scattered across multiple pages.
Blog posts and press releases. "We're excited to announce that Company X chose our platform." These are harder to catch, and the agent might miss some if they're buried deep. Setting a higher max_spend_usd gives it more room to browse.
Integration partner pages. Some companies list customers on their integrations or partners page.

The agent checks all of these. You don't need to tell it which pages to visit (though you can narrow it down with seed URLs if you want faster results).

Writing a good prompt

Tell the agent what you want and what format to return it in. Being specific about where to look helps too.

Find all customers, clients, or users listed on this website.
Check: homepage, customers page, case studies, testimonials, about page, and any "trusted by" or "used by" sections.

For each customer return:
- "company_name": the customer's company name
- "source": where on the site you found them (e.g. "homepage logo wall", "case study page", "testimonial")

If the same company appears in multiple places, include it once and list all sources.

The source field is useful. When you're building a competitive analysis, knowing whether a customer was featured in a full case study vs just a logo on the homepage tells you how deep that relationship probably is.

Prompt template

Find all customers listed on {company_name}'s website at {website_url}.

Look at the homepage, any customers or case studies page, testimonials, about page, and partner pages.

For each customer return:
- "company_name": the customer's company name
- "source": where you found them on the site
- "context": any additional detail (e.g. quote, use case description, industry)

Only include companies that are clearly identified as customers, clients, or users. Don't include partners, investors, or integrations unless they're also listed as customers.

That last instruction matters. Without it, the agent might lump in technology partners, investors, or integration logos. Being explicit about what counts as a "customer" keeps the data clean.

What you won't get

Worth repeating: the agent can only find customers that are publicly listed on the website. If a SaaS company doesn't show customer logos, doesn't publish case studies, and doesn't have testimonials, you'll get an empty result.

Some companies are intentionally quiet about their customer base. Enterprise B2B products often don't list clients publicly. The agent will tell you it didn't find any, and that's an honest answer.

Wrapping up

Customer extraction is a common competitive intelligence task. Run it across 20 competitor websites and you'll have a pretty good map of who uses what in your market.

The same API works for anything else you need from those sites: pricing, feature lists, team pages, job postings. The prompt is the interface.

The output is JSON, so it fits into your CRM, a research spreadsheet, or an AI agent that's mapping your competitors automatically.

Grab an API key and try it on one competitor first. Full Agent API docs are here.

The same agent also handles other use cases: finding a company website by name, extracting SaaS pricing, or pulling contact emails from any website.

API to Find Customers of Any SaaS Company