Most B2B teams wire website visitor identification into one place: sales alerts. A rep gets pinged, a lead lands in the CRM, and that's the end of the workflow. But the same identified-account data answers a question your paid media budget has been guessing at for years: which of your ads are actually reaching your ICP, and which accounts should you be retargeting instead of writing off as "anonymous traffic"?

Why Google and Meta Can't Target Like LinkedIn Does

LinkedIn has always had one structural advantage over every other paid channel: it knows who works where. You can build a Matched Audience off a company list or a title, and the platform will actually hold you to it. Google and Meta can't do that natively. One ABM lead building a program from scratch for a fast-growing SaaS company put the frustration plainly on a recent call with our team: outside of LinkedIn, there isn't a channel that lets you target "at that degree" — you can run Google and Meta campaigns at your ICP, but you can't confirm the platform is actually reaching it. You're buying impressions and hoping the targeting holds.

That gap is exactly where identified website traffic earns a second job. If you already know which companies are hitting your site, you don't need Google or Meta to solve account-based targeting for you. You can solve it yourself, on the back end, with data you're already collecting.

Two Plays: Attribution First, Audiences Second

There are two distinct ways to point visitor identification at your paid media problem. Most teams should start with the first one — it's lower lift, and it tells you whether the second is worth building at all.

Play 1: Prove Which Campaigns Are Actually Reaching Your ICP

This is an attribution exercise, not a new data-capture project. It works with what you already have:

1. Tag every paid campaign with UTMs — source, medium, and campaign at minimum, content if you're running more than one creative per campaign.
2. Join UTM sessions against your identified-account data. Instead of reporting clicks and CPC in a vacuum, you now know which sessions belonged to companies that actually match your ICP criteria (industry, size, tech stack, whatever you score on).
3. Build a weekly ICP-match-rate report per campaign, not just per channel. Two campaigns with identical CTR can have wildly different ICP match rates once you can actually see who clicked.
4. Reallocate based on match rate, not just cost-per-click. A campaign with a mediocre CPC but a 40% ICP match rate is outperforming a cheap campaign matching at 8%, even if the raw numbers say otherwise.

On a walkthrough call with a customer mid-onboarding, our own team ended up demonstrating this live almost by accident — asking whether they were running paid campaigns, then pulling up which ones were actually driving ICP-matched sessions using the UTMs already in place. Their reaction was some version of "wait, we can see that?" It's a genuinely underused report, and it's the one every marketing team should build before touching the second play.

Play 2: Build Company-Level Retargeting Lists LinkedIn, Google, and Meta Can All Use

Once attribution tells you which campaigns are working, the next move is building your own account-based retargeting audiences instead of waiting for the ad platform to do it for you. The mechanics are straightforward and, importantly, stay at the company level rather than the individual level:

• Pull the list of identified companies that match your ICP criteria and hit a high-intent page — pricing, a demo request that didn't convert, a specific product page.
• Export it as a company-domain list, not a list of named individuals.
• Upload the domain list to LinkedIn Matched Audiences, Google Ads' Customer Match (company-level lists via your CRM integration), or Meta's Custom Audiences.
• Refresh the list on a set cadence — weekly is usually tight enough to matter, monthly is usually too slow for a fast sales cycle.

Keep this play account-based, not person-based. You're retargeting companies that showed real intent, not individuals matched off personal identifiers. That distinction isn't just a compliance nicety — it's also the version of this play that's actually defensible when someone on your team asks how the list got built.

Which Ad Channels Can Actually Run This Today

Not every channel is equally ready for company-list-based targeting. Here's how the major B2B ad channels stack up on maturity for this specific play:

• 🔴 LinkedIn Matched Audiences - native company and title-level targeting, matches company lists directly, still the most mature ABM ad product on the market.
• 🟠 Google Ads (Customer Match / Display & Video) - can match against company-domain lists through your CRM integration, but the list has to be built and refreshed manually; works well once volume is high enough to matter.
• 🟡 Meta Custom Audiences - accepts company-domain lists, but Meta's targeting stack is consumer-first at its core, so treat it as brand-awareness retargeting rather than an account-precision play.
• 🟢 Programmatic display (The Trade Desk, StackAdapt) - strong account-based reach and IP-based targeting, but needs dedicated setup and a real budget behind it to be worth the lift.
• ⚪ X/TikTok Ads - minimal native support for account-based B2B targeting today; skip unless your ICP genuinely lives there.

How Much Traffic Do You Need Before This Is Worth Building

This play is gated by identification volume, not creativity. Knock2's published identification benchmarks are 93%* account/company-level and 62%* person-level (name, email, title) on US traffic, both measured against engaged sessions.* If your paid campaigns are only driving a trickle of sessions, even a 93% match rate produces a retargeting list too small to be useful. As a rough gut check: if your paid campaigns are driving fewer than a few hundred sessions a month, build the attribution report first (Play 1) and hold off on retargeting audiences (Play 2) until volume catches up. Once you're in the thousands of monthly paid sessions, the company-list approach starts producing audiences big enough for LinkedIn and Google to actually optimize against.

*Identification rates measured against engaged sessions. Results may vary by traffic profile, geography, and industry.

The Weekly Habit That Makes It Stick

Neither play survives as a one-time report. The teams getting real value out of this treat it as a standing weekly check: ICP match rate by campaign, refreshed retargeting list size, and which campaigns get more or less budget as a result. That's the same operating rhythm we've written about for routing identified visitors to reps and for proving ROI on visitor identification more broadly — the pattern holds here too. A signal nobody reviews on a cadence is a signal that quietly stops mattering.

This is also where Knock2's Ad Reveal earns its keep: it ties identified visitors directly back to the campaign that brought them, so the attribution report in Play 1 doesn't require you to hand-build a join between your ad platform's UTMs and your identification data every week. If you're weighing where paid media fits into a broader GTM stack built around website visitor identification, this is one of the higher-leverage additions precisely because most teams haven't built it yet.

FAQ

Can I upload identified visitors directly as ad audiences?

Stick to company-level lists, not individual identifiers. Uploading personal identifiers (emails, names) to an ad platform is a different and riskier practice than uploading a list of company domains that match your ICP. The plays in this piece work at the account level for a reason — it's the version that's both effective and easy to defend.

Does this work on Meta and Google, or just LinkedIn?

All three can accept company-domain lists, but LinkedIn is the only one with native company/title targeting built in. Google and Meta require more manual list-building and refresh work, and Meta in particular should be treated as a brand-awareness channel rather than a precision account-based one.

How much traffic do I need before this pays off?

Enough paid session volume that a 93%* account-level match rate still produces a list big enough for an ad platform to optimize against — in practice, that's usually a few thousand monthly paid sessions. Below that, run the attribution report (Play 1) and hold off on building retargeting audiences (Play 2).

How is this different from a standard retargeting pixel?

A retargeting pixel re-shows ads to anyone who visited, with no sense of whether they're actually in your ICP. This play filters first — you're only retargeting identified companies that already match your ideal customer profile, not every visitor who bounced.

Do I need a dedicated ABM platform to do this?

No. Both plays run on identification data, UTM tracking, and the native audience-upload tools each ad platform already provides. A dedicated ABM platform can make list refresh more automated, but it isn't a prerequisite to start.

Want to see which campaigns are actually reaching your ICP? See how Knock2 works for marketing teams and start with the attribution report before you build a single retargeting list.