AI Ads Don't Break Marketing Measurement. They Reinforce It.
Ads inside AI assistants may produce few clicks, but that doesn't make them unmeasurable. Spend, impressions and sales over time are enough to model impact.
Every time a new advertising channel appears, the same prediction follows: you won't be able to measure it.
We're hearing it again now about AI assistants and the advertising that may run inside conversational platforms. The assumption is that with no clicks, no pixels and no traditional web journey, marketers will lose sight of performance.
That misunderstands how modern measurement works. The channel may be new. The math isn't.
Measurement doesn't start with clicks
For years, marketers were taught that clicks are the foundation of attribution. They aren't. A click is one observable event in a much longer journey.
Long before someone clicks, they've seen impressions, messages and experiences that shaped their decision. That's why measurement has moved beyond click-based attribution. Today, marketers can measure incremental impact from signals that exist whether or not a click ever happens, starting with two simple questions:
- How much did you spend?
- How many people did you reach?
The data every advertiser already gets
Consider how advertising platforms work. If you buy media, the platform has to bill you, so you know exactly what you spent each day, down to the campaign.
Platforms also typically report impressions, because advertisers want to know how often their campaigns were delivered and how much of the available inventory they won. Platforms have their own reasons to share it: metrics like impression share show advertisers what they missed and often encourage more spending.
If AI advertising platforms develop along similar lines, marketers can reasonably expect the basic delivery data needed to evaluate performance.
Measurement no longer depends on pixels
Traditional attribution relied on cookies, pixels and user-level tracking. Privacy changes, including Apple's App Tracking Transparency and browser tracking protections, have already pushed the industry past that model.
Methods such as marketing mix modeling and incrementality analysis don't need every click tracked. They study how media delivery and business results move together over time. The inputs are straightforward:
- Daily spend
- Daily impressions
- Campaign-level delivery
- Business outcomes, such as sales or leads
From those, a model estimates what each channel added: the sales that would not have happened without it.
New channel, familiar questions
Whether a campaign runs on connected TV, streaming audio, digital out-of-home, retail media or an AI assistant, the measurement questions are the same:
- Did it build awareness?
- Did it change what customers did?
- Did it create sales that wouldn't have happened otherwise?
The delivery mechanism changes. The analytical framework doesn't.
We've been here before
Search changed the industry. So did social media, and so did streaming television. Each time, someone predicted measurement would become impossible, and each time marketers adapted by using better methods rather than clinging to old metrics.
AI-driven advertising will be no different. The real risk isn't that it can't be measured. It's that AI platforms report their own results inside their own walls, the way other large platforms do. We cover that side of the story in AI could become advertising's next walled garden.
The focus should be less on whether a channel looks different and more on whether you can measure its contribution to growth. If you know what you invested, how the media was delivered and how your results changed, you can measure it. The channel is new. The math is not.
Frequently asked questions
Can you measure ads that run inside AI assistants?
Yes, as long as you know what you spent, how much the ads were delivered, and how your sales or leads changed over time. Those are the inputs marketing mix models use for any channel, whether or not anyone clicks.
Do you need clicks or pixels to measure advertising?
No. Clicks and pixels are useful signals, but aggregate methods like marketing mix modeling estimate a channel's incremental effect from spend, impressions and outcomes. That is how TV and radio have been measured for decades.
What data should I ask an AI ad platform for?
At minimum, daily spend and daily impressions at the campaign level. With those, plus your own sales or lead data, an incrementality model can estimate what the channel added.