On this page
ABM retargeting is paid media pointed at a defined set of accounts, and the people inside them, after they have shown a signal. The version that produces pipeline builds the audience from what your revenue system already knows about the account: fit, buying stage, near-term propensity, whether there is an open opportunity, what product they looked at. It suppresses the accounts that should not see the ad. It syncs to the ad platforms on a schedule instead of once. And it gets measured on pipeline lift against a holdout, not on clicks.
Most programmes we see do none of those four things, and then wonder why the CFO is unmoved.
Why the pixel version underperforms
A pixel fires on a visit. That is all it knows. It cannot tell whether the visitor was a procurement analyst at a target account, a student, a competitor, or an existing customer hunting for the login page. It does not know the account is in a live negotiation where the rep would very much like the buying committee to stop seeing ads about the product they are haggling over. It does not know the account was closed-lost four months ago and the cooling-off period has another two to run.
So the report comes back the way it always does: lots of impressions, an acceptable click-through rate, and no way to show anyone that pipeline moved.
Gartner's June 2026 survey of 401 CMOs, most at companies above $1 billion, found awareness and conversion now take 62.6% of media spend, up more than ten points since 2024, with digital at over 67% of the media budget. Only 30% of marketing organisations in the same survey had mature AI capabilities. The biggest and fastest-growing line in the budget is, for most teams, being targeted with the audience logic they had in 2023.
A demand generation leader at an enterprise search company told us LinkedIn was, for the first time in his career, beating Google Ads, and that he would keep spending there "even if it was telling me diminishing returns, because the volume it's getting is a lot better." He could say that because he could see the pipeline the channel touched. Without that view the same sentence is a hunch.
Building the audience the other way round
Start from everything you know about the account
ABM intent and account identification from your platform. Website visits resolved to accounts and, where you are allowed, to people. Marketing automation activity. CRM stage and opportunity data. Sales activity from email and calendar. Product usage if you have a trial or freemium motion. The ABM platform supplies the first of these and stays exactly where it is. The rest is what it was never built to hold.
Score it
For each account, four values: fit against your ICP, buying stage inferred from behaviour rather than declared by a form, near-term propensity to create pipeline, and, for open opportunities, risk of slipping. Those four are what an audience definition should reference. Our predictive models produce them continuously on the Full Funnel Data Graph; the Hot Accounts agent is one consumer, retargeting audiences are another.
Define the audiences in revenue terms
High-fit accounts that are anonymous on the site but surging on intent. This is the awareness play, and its only job is to get one buying-committee member to identify themselves.
Accounts with rising three-month propensity and at least two known contacts. The acceleration play, aimed at the rest of the buying group.
Open opportunities showing risk, such as a pushed close date and falling engagement. The re-engagement play, aimed at the stakeholders who went quiet, with creative the rep has actually agreed to.
Customers with usage signals in an adjacent module. The expansion play.
Every one of those carries its own suppression list. Customers are out of awareness and acceleration. Closed-lost inside the cooling-off window is out. Accounts already in an active opportunity with a rep engaged are out of awareness and only enter re-engagement if risk fires. Low-fit is out of everything.
A marketer at a cloud security company running a proof of concept with us flagged a campaign the model had credited with pipeline that, in his judgement, should never have generated any. He was right. The fix was an exclusion list, and the same logic belongs in audience definitions: the human decides which accounts the machine never targets. We would rather have that argument up front than after the spend.
Keep the sync alive
Push the segments to LinkedIn, Google, Meta and whatever else you run, and keep pushing. This is where most ABM retargeting quietly dies. An audience uploaded once is stale in two weeks: accounts convert, opportunities close, stakeholders leave. The LinkedIn Audience Updater agent exists for this one reason. It re-syncs matched audiences as accounts change stage, so the spend follows the pipeline instead of trailing it by a fortnight.
Scores and stage also go back to the ABM platform, which makes its own segments sharper. That return trip is what the platform was bought for and it is the part most integrations skip.
Measure on pipeline
Hold out a matched set of eligible accounts from each audience and compare pipeline created and influenced. Run attribution on the retargeting campaigns alongside every other touch so the credit is proportionate rather than last-click. Where the budget allows, run incrementality tests by region or segment. Then hand the results to the Campaign Reallocation agent, which is where retargeting money moves toward the segments that produced pipeline and away from the ones that produced impressions.
What the numbers might look like
We invented these to show the shape.
A company with 2,400 target accounts runs the four audiences above. The awareness audience is 900 anonymous, high-fit, surging accounts; over a quarter, 310 of them produce an identified visitor, against 140 in the holdout. The acceleration audience of 350 accounts creates 41 opportunities against 27 in the holdout. Of 60 at-risk opportunities in the re-engagement audience, 22 see a stakeholder return within three weeks, against 9. Quarterly retargeting spend is $180,000. Attributed incremental pipeline is $2.1 million.
Those are numbers a CFO will read. "4.2 million impressions, 38,000 clicks and 412 form fills" is a report about the ad platform's health, not yours.
Where the ABM platform sits
At the front, where it always was. It finds the in-market accounts, supplies the intent, runs the account-level advertising. Everything above is what happens to its signal once it meets first-party context the platform never had: the web, the CRM, the conversations, the product. The scores and segments go back to it. The pipeline measurement sits underneath both. One of our customers reallocated $100,000 across five verticals on that kind of evidence and added 105 net new opportunities. That is the bar we would hold a retargeting programme to.
Frequently asked questions
What is ABM retargeting?
Paid media aimed at a defined list of target accounts, and the people in their buying groups, after they show a signal. Done well, the list comes from fit, buying stage, near-term propensity and opportunity status rather than from a cookie that saw a page.
How is ABM retargeting different from standard B2B retargeting?
Standard retargeting follows anyone who visited a page. ABM retargeting follows accounts that clear a fit threshold and a stage or propensity condition, suppresses the ones that should not see the ad, and gets measured on pipeline rather than clicks.
Which accounts should be suppressed from retargeting?
Customers, unless the play is expansion. Closed-lost accounts still inside their cooling-off window. Accounts below the fit threshold. Accounts already in an open opportunity with a rep engaged. Anyone who has shown a negative signal such as an unsubscribe or a competitor win.
How do you measure ABM retargeting?
Pipeline lift against a matched holdout of eligible accounts you deliberately did not target, attribution of created and influenced pipeline to the retargeting campaigns, and incrementality tests where the budget allows. Impressions and click-through rate report delivery, not effect.
How does RevSure build retargeting audiences?
Segments are defined on the Full Funnel Data Graph using fit, buying stage, propensity, opportunity status and web behaviour, then synced to LinkedIn, Google and other paid platforms by agents like the LinkedIn Audience Updater, which keeps them current as accounts move stage.
Related reading: LinkedIn matched audiences that stay current, why ABM needs next-gen attribution.