The ad tech industry built the most precise measurement infrastructure in the history of media. Every impression counted. Every click tracked. Every conversion attributed. The ability to measure what happened after an ad was served, across devices, across platforms, across sessions, became the industry’s defining capability and its core value proposition.
But that infrastructure doesn’t measure influence. And as B2B buying behavior shifts toward channels that ad tech can’t see (AI-assisted research, private peer conversations, dark social sharing), the gap between what measurement captures and where decisions actually happen is becoming the industry’s most consequential blind spot.
Impressions are measured. Decisions aren’t.
The B2B buyer who sees a display ad, clicks through, visits a landing page and fills out a form is fully measured. Every touch point is tracked, timestamped and attributed.
What about the B2B buyer who asks an AI assistant for a vendor recommendation, receives an answer that doesn’t mention your brand, forms a shortlist without visiting any website and eventually fills out a competitor’s form? That buyer is invisible to every measurement system the industry has built.
There’s also the B2B buyer who receives a forwarded email from a colleague (“This is the company I mentioned”), clicks through from that personal message, visits the site and converts. The measurement system sees a direct visit with no campaign attribution. The reality is that the most influential touch point, the peer recommendation, happened in a channel no ad tech platform can observe.
Each of these scenarios represents a real buying decision. And the proportion of buying decisions happening in invisible channels is growing, while the proportion happening in measurable channels is shrinking.
Why incrementality matters more than attribution
Attribution can only credit what it can see. If a significant portion of the buyer’s journey happens in channels attribution can’t track (AI conversations, private messages, dark social, peer recommendations), the model is distributing full credit across a partial journey.
The visible touch points get overcredited. The invisible touch points get zero credit. The resulting picture of “what’s working” is systematically biased toward the measurable channels regardless of whether they’re the influential ones.
This bias compounds over time and shapes the budget.
If attribution measures the visible journey, incrementality measures the actual impact: the difference between what happened with the marketing activity and what would have happened without it.
Incrementality testing asks the question attribution can’t: Did this activity change behavior, or did it reach someone who was going to convert anyway? A campaign that reaches 1,000 buyers who were already in-market and attributes 50 conversions looks strong in the attribution model. An incrementality test might reveal that 45 of those 50 would have converted without the campaign, meaning the actual contribution was five conversions, not 50.
The ad tech industry has the infrastructure to run incrementality tests (holdout groups, geographic splits, matched market analysis). The tools exist. The industrywide commitment to using them as the primary measure doesn’t, because incrementality almost always produces lower numbers than attribution, and lower numbers are harder to report.
The problem is this: The incentive structure rewards the metric that makes performance look best, rather than the metric that reflects reality.
AI-mediated buying widens the gap
The measurement gap is about to significantly grow. AI is inserting itself into the B2B buying process in ways that make traditional measurement increasingly incomplete.
When a buyer uses an AI assistant to research a purchase, the research happens inside a conversation that no ad tech platform can observe. The AI synthesizes information from across the web, forms a recommendation and delivers it in a chat window. No impression was served. No click happened. No cookie was set. The buyer received a recommendation that shaped their shortlist, and it was based on content and data the buyer never directly visited.
As AI purchasing agents begin evaluating B2B vendors, the top of the funnel moves into a channel that current measurement can’t capture. The industry that built its value proposition on measurability is facing a future where the most consequential buyer interactions are unmeasurable by the infrastructure that exists today.
What decision measurement looks like
The buy side needs a measurement paradigm built around decisions rather than impressions.
- Self-reported attribution at every high-intent conversion point, a free-text field asking “How did you hear about us?” captures the channels technology can’t.
- Incrementality as the primary metric for paid activity. Did the activity change behavior that wouldn’t have happened otherwise?
- Pipeline and revenue as anchor metrics. Did the activity produce a pipeline? Did that pipeline close?
- Buyer research as a measurement input. Post-deal interviews and win/loss research reveal the actual decision journey in ways no technology can replicate. The buyer tells you what influenced them. That qualitative data is the highest-fidelity measurement available and the most underused.
The industry that solved measurement for impressions needs to solve it for decisions. That’s the harder problem. It’s also the only one that matters, as the buyer’s journey moves beyond the reach of the infrastructure built to measure it.
“Data-Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.
Follow Sojourn Solutions and AdExchanger on LinkedIn.
For more articles featuring Margarita Savytska, click here.
