How to See Which Companies Visit Your Website (Step by Step)

If you’ve ever wondered how to see which companies visit your website, the frustrating answer inside Google Analytics is: you can’t. GA4 shows you sessions, channels and events, but it deliberately anonymizes who is behind them. It even removed the old network/service-provider report that used to give a rough hint. For a B2B business, that means the most valuable signal you have — a real buyer researching you — is invisible.

The math makes this painful. Many of those anonymous visitors work at exactly the companies you want to sell to. They read your pricing page, compared you to a competitor, and left without a trace — at least in GA4.

Company-level visitor identification closes that gap. It doesn’t tell you the individual person (more on why in our guide on how visitor identification works), but it tells you the firm: “Someone at Acme Logistics viewed your pricing page three times this week.” For a sales-led or hybrid B2B motion, that is usually the single highest-intent signal you’re not acting on today.

There are two layers, and being honest about both matters. The first is IP-to-company resolution. Businesses often route traffic through networks registered to the company itself, so when a visitor arrives from one of those ranges, the IP can be matched against business databases to reveal the employer. This works without cookies and without identifying any individual — it’s firmographic data about an organization, which is also what makes it defensible under GDPR (we cover that in depth in our visitor identification GDPR guide).

The second layer is email-domain stitching. When a visitor eventually does identify themselves — a form fill, a clicked link in an email, a signup — their email domain (say, [email protected]) confirms the company with certainty, and the tracker retroactively stitches their entire anonymous journey to that contact and firm. This layer is exact, not probabilistic, and it compounds: every identified person makes the picture of that account sharper.

Set expectations now: IP resolution typically identifies roughly 20–35% of B2B traffic at the company level. Remote workers on home ISPs, mobile connections and VPNs resolve to providers, not employers, so they stay anonymous. That’s normal, and any vendor telling you otherwise is overpromising. The identified quarter-to-third of your traffic is still a goldmine most teams never touch.

With BusinessMCP, the entire setup is a single cookieless script tag. Create a free account (no credit card), and the dashboard hands you a snippet during onboarding. Paste it into your site’s HTML just before the closing head tag — that’s the whole install. If your site runs on WordPress, Webflow, Framer, Next.js or anything else that lets you add custom code to the head, you’re done in under five minutes.

• Direct install: paste the snippet before the closing `head` tag on every template.
• Google Tag Manager: add a Custom HTML tag containing the snippet, trigger it on All Pages, publish.
• Verify: open your own site in a new tab — the dashboard’s live view should show a visitor within seconds.
• Enrichment: company identification is part of the paid plans (from $19/mo); the free plan proves tracking works first.

If your team manages tags through Google Tag Manager, that works too — and because the script is cookieless by design, you’re not adding another consent-banner dependency just to get analytics. That’s a meaningful difference from cookie-based tools, which we unpack in the GDPR guide.

Once traffic flows, the Users view stops being a wall of “Anonymous” rows. Identified visitors show a company name, industry, size and location pulled from business databases — enrichment happens automatically on paid plans, with no extra configuration. You’ll see which firms are visiting, how often they return, and which of them match the profile you actually sell to.

Click into any company or visitor and you get the full journey: the landing page, the referral source or campaign that brought them, every page viewed, and every goal fired — across sessions, not just the last one. This is where intent becomes readable. A firm that landed once from a blog post is a curiosity; a firm that has hit your pricing page in three separate sessions this week is a live evaluation.

Because BusinessMCP includes a CRM in the same product, identified companies and any stitched contacts flow straight into your pipeline view rather than living in a separate tool you have to export from. Analytics, the company feed and the CRM read from the same identity graph, so a visitor who later fills a form doesn’t become a duplicate — they become the same record with more history.

The raw company feed will contain noise — competitors checking your pricing, students, vendors trying to sell to you. The workflow that works is aggressive filtering: narrow to what matches your ideal customer profile. A 200-person logistics company that viewed your pricing page beats fifty one-off blog readers, and your filters should encode that.

• Industry — the verticals you actually close, not the ones you aspire to.
• Company size — a band around your real customer base.
• Behavior — pricing, integrations or docs pages beat blog-only visits.
• Recency and frequency — returned this week beats seen once last month.

Then take the humans out of the loop for detection. Connect Slack in BusinessMCP and route alerts to a channel your sales team actually reads, so a high-fit company hitting a high-intent page pings you the same day — not in next Monday’s report. Speed matters here: the visit is the moment of interest, and the value of the signal decays fast.

Identification without action is just voyeurism, so decide upfront what happens when a target account shows up. The honest playbook: don’t email saying “we saw you on our website” — that reads as surveillance and burns trust. Instead, treat the visit as timing intelligence. You now know the account is in-market; reach out with relevant value as if you’d picked them off a well-researched list.

A concrete play: an ICP-fit company reads your integrations page and your pricing twice. Find the two or three likely buyers at that company on LinkedIn, and open with something specific to their firm and the problem your product solves — never referencing the visit itself. The signal doesn’t change your message; it changes who you message and when.

If you want this automated end to end, our AI SDR (on the Scale plan, from $199/mo) can take identified in-market companies, find decision-makers, research them and draft the outreach for your approval — the full loop from anonymous visit to personalized first touch. We cover the sequencing and messaging in our warm outbound playbook.

Mistake one: expecting 100% match rates. No tool identifies everyone, because the underlying data doesn’t exist — home ISPs, mobile carriers and VPNs mask the employer. If a vendor demo implies near-total coverage, they are counting ISP names as “companies” or padding with stale data. Plan your workflow around the realistic 20–35% of B2B traffic, and judge tools on the quality of those matches, not the quantity.

Mistake two: ignoring repeat-visit signals. Teams often treat the company feed as a one-shot list, work it once, and conclude “this doesn’t convert.” The compounding value is behavioral: which accounts keep coming back, which pages they escalate to, and how visits cluster before a deal. One visit is weak evidence; a pattern is a forecast.

Mistake three: treating ISP hits as companies. When a visitor can’t be matched to an employer, some datasets return the network operator — “Comcast,” “Vodafone,” a hosting provider. Those aren’t prospects, and outreach built on them wastes everyone’s time.

You can validate the whole idea cheaply. BusinessMCP’s free plan (no credit card) gets the cookieless script live and proves tracking works on your stack; paid plans start at $19/mo and add visitor identification and company enrichment. Because the results depend entirely on your traffic mix, the only evaluation that means anything is running it on your own site for a week or two — not a vendor’s demo data. For context, established competitors in this category typically start much higher: Leadfeeder’s paid plans and Warmly’s pricing are an order of magnitude up from that entry point.

Start small and honest: install the script today, let a week of traffic accumulate, filter to ICP-fit companies with repeat visits, and hand your sales team the top five accounts with their full journeys. If those five conversations aren’t warmer than your average cold list, you’ve lost almost nothing. In our experience, that first handoff is the moment the anonymous 98% stops feeling invisible.