Your ABM Platform Has a Data Problem Coming
The ABM platforms that transformed B2B marketing over the last decade were built on a simple and powerful premise: track what your target accounts are searching for and reading across the web, and use that behaviour to infer where they are in a buying cycle. It worked because search behaviour was observable, consistent, and abundant. Billions of queries, millions of content interactions, all leaving a traceable signal trail that intent platforms could hoover up, model, and sell back to marketing teams as buying signals.
That premise is now under pressure in a way most ABM platform users haven’t fully recognised yet.
The signal layer is thinning
Industry leading ABM platforms such as 6sense or Demandbase derive their intent intelligence from two primary sources: keyword-level search data aggregated across publisher networks, and behavioural data from content consumption across third-party sites. Both of those inputs depend on one thing above everything else: people clicking through to content.
Google’s AI Overviews are systematically reducing that behaviour. When a senior buyer at a target account searches for “best enterprise CDP solutions” or “how to evaluate ABM vendors” and gets a synthesised answer inside Google’s interface, they don’t click through to the comparison page, the analyst report, or the category hub. The research happens. The intent exists. But the signal never fires because the click never happens.
This isn’t a marginal effect. For informational and consideration-stage queries, the ones most valuable for early intent detection, zero-click rates are already climbing steeply. The searches that used to generate the behavioural data underpinning your surge scores are increasingly being answered before anyone leaves Google.
What breaks first inside your ABM platform
Account surge scores will become less reliable. Accounts that are genuinely in-market will show lower intent scores because the search and content consumption behaviour that would have triggered a surge is no longer happening in observable channels. You’ll see fewer accounts lighting up at the top of the funnel, not because buying activity has slowed, but because the activity has moved somewhere you can’t see.
The accounts that do appear in your intent data will be further along in their research than the signal suggests. They’ve already done the AI-assisted discovery phase. They’re arriving at your intent platform’s radar already mid-consideration, with a compressed timeline and pre-formed views. Your SDR sequence built around early-stage nurture is now landing at the wrong moment.
Keyword-level intent data will also start to lose its granularity. If fewer people are clicking through on category-level searches, the topic clustering models that platforms use to infer buying stages become less accurate. You’ll see topic surge data that is technically correct but contextually misleading because the underlying query volume doesn’t reflect the same research depth it once did.
The platform problem specifically
Industry leading ABM platforms have invested heavily in AI-driven predictive models that layer firmographic data, CRM signals, and third-party intent together to produce buying stage predictions. Those models were trained on historical patterns of observable behaviour. As AI search strips out a growing portion of that observable behaviour, the training data those models relied on becomes less representative of how buyers actually research today.
Platforms with Revenue AI capabilities are particularly dependent on keyword intent signals to move accounts through their buying stage models. If the keyword signals thin out, accounts will cluster in earlier buying stages not because they’re early, but because the model can’t see the signals that would move them forward. Your sales team acts on a cold account that’s actually warm. The conversation lands wrong. The deal takes longer or doesn’t happen.
Account identification and qualification logic that uses content consumption patterns to build a picture of where an account sits in the consideration journey becomes less complete as AI Overviews reduce traffic to the category content and review sites that feed those signals. Accounts that should be in the active pipeline stage look dormant.
These platforms are not standing still, they are investing in first-party data integrations, CRM signal weighting, and proprietary publisher networks to compensate. But the gap between where their models are calibrated and where buyer behaviour is heading is real, and it’s widening faster than their product roadmaps are closing it.
What agencies need to tell their clients right now
The first conversation is about expectation setting. Intent data is not going away. But its reliability as a standalone signal for identifying in-market accounts is declining, and any agency still presenting surge scores as confident buying signals without caveating the methodology is doing their clients a disservice.
The honest framing is this, intent data is now better understood as a corroborating signal than a primary one. When it fires, it still means something. But the absence of a signal no longer means absence of intent. That distinction matters enormously for how sales teams prioritise outreach and how marketing teams build their account lists.
The second conversation is about first-party data architecture. Clients who have invested in robust CRM hygiene, consistent product interaction tracking, event and webinar attendance data, and owned community engagement are significantly better positioned than those who relied on third-party intent as their primary account intelligence layer. Agencies should be auditing that first-party data infrastructure now and identifying the gaps before the intent signal degradation makes those gaps visible in pipeline numbers.
What to do inside the platforms
The configuration changes that matter most are the ones that shift weighting away from keyword intent and toward signals that don’t depend on search click behaviour.
Review your model weighting and increase the influence of CRM activity signals, website engagement from known accounts, and engagement with owned content assets. If you have product usage data or customer success interaction data flowing into the platform, make sure it is properly weighted. Reduce your reliance on the keyword intent surge score as a primary trigger for sales alerts.
For account-based platforms with intent topic libraries, audit what you have. Topics that are heavily dependent on informational search queries are the most exposed. Refocus your topic selection toward commercial and decision-stage intent signals, and increase the weight given to direct site engagement, ad engagement from target accounts, and first-party form interactions.
Across all platforms, tighten your ICP definition. When intent signals were abundant, broad account lists were manageable because the surge data helped filter who was active. As that filter weakens, working from a tighter, better-qualified account list becomes more important. Quality of target account selection compensates for some of the signal loss.
The broader strategic shift
The intent data conversation is really a proxy for a bigger question about where B2B marketers go to understand buyer behaviour when the traditional signal layer degrades.
The answer is a combination of things that individually aren’t new but together represent a different operating model. More weight on owned channels and first-party behavioural data. Deeper investment in communities, events, and direct category engagement where your brand controls the environment and can observe interactions directly. Closer alignment between marketing and sales so that qualitative signals from conversations are fed back into account prioritisation. And a more honest relationship with the metrics, accepting that some portion of buying behaviour is now happening in environments that will never be observable and building models that account for that rather than pretending the signal is complete.
The platforms will catch up, the category is too important and the vendors too well-resourced for the gap to persist indefinitely. But the catch-up will take time, and the businesses who adapt their operating model now will be the ones who maintain pipeline predictability while their competitors are still wondering why their intent scores stopped making sense.
The practical checklist for the next 90 days
- Audit your current intent signal sources and identify what percentage of your account intelligence is derived from keyword and third-party content signals versus first-party data. If that ratio is heavily skewed toward third-party, you have exposure.
- Run a retrospective on your last two quarters of the pipeline. Look at the accounts that converted and work backwards through their intent signal history. If you start seeing accounts that convert with minimal preceding intent activity, the signal degradation is already happening.
- Have an honest conversation with your ABM platform account team about how their models are being updated to account for AI search behaviour changes. Ask specifically about keyword intent signal weighting and what compensating signals they are introducing. The quality of that conversation will tell you a lot about how seriously they are taking the problem.
- Start building or strengthening the first-party data infrastructure that will matter most when third-party intent becomes less reliable. That means CRM discipline, event and webinar tracking, owned content engagement measurement, and wherever possible, product interaction data.
- And reset expectations with your sales team now, before the pipeline numbers make the conversation harder. Intent data is a useful signal. It has never been a guarantee. In the world AI search is creating, that distinction is more important than ever.
- And if you’re not sure where to start, talk to us. We’ve been working with B2B marketing teams navigating exactly this transition and can help you identify where your intent strategy has exposure and what to do about it. Get in touch.
If your surge scores are holding but your pipeline predictability is slipping, the gap between those two numbers is worth understanding. Let’s talk about how your ABM platform is configured and what signals you should be weighting differently in an AI search world.
