Causality Is the New Attribution: How Alembic Distills Data

Hitesh Wadhwani, newly appointed GM of Alembic, arrived with a confidence booster from a deep background in measurement including 12 years at Google, where he helped lead development of global measurement products including Meridian. As part of a series I recorded during the POSSIBLE Conference in Miami for “Insider Interviews,” he explained causality, how to escape attribution frustration, and how better measurement could help CMOs make a more convincing business case internally.

A favorite line from Wadhwani’s “new job post” on LinkedIn candidly cautioned: “you cannot hallucinate your way through a multimillion-dollar strategic budgeting decision.” Now, when generative AI is being inserted into nearly every marketing workflow imaginable, that distinction matters. Large language models are remarkably effective at language prediction and summarization. But, he says, enterprise measurement requires something fundamentally different: numerical reasoning, causal inference, and an understanding of how multiple business variables influence one another over time.

What Causal AI Actually Does Differently

Then there’s correlation versus causation. Just because two things happen at the same time does not necessarily mean one caused the other. In marketing, that becomes incredibly difficult to untangle once media spend, promotions, pricing, seasonality, consumer behavior, and outside events are all moving simultaneously.

Alembic’s approach is designed to look across those variables together rather than measuring channels separately. The goal is to help companies better understand what actually appears to be influencing business outcomes and by how much, and exactly when, or the “moments” that matter. Their aim is to show the relationships between variables more directly so teams can better understand what may actually be driving performance, what may not be contributing much at all, and where assumptions may simply be wrong.

That becomes really relevant inside large organizations, where marketing, finance, analytics, and operations teams are often working from different methodologies and different definitions of success. That’s where one of the more practical parts of our discussion could pay off: solving the disconnect between marketing and finance.

Why the CMO-CFO Conversation Gets Complicated

Brand marketers, performance teams, analytics groups, and finance departments can walk into the same meeting with entirely different ways of measuring impact. Each set of numbers may make sense independently, but together they often fail to create a clear investment story. With tighter budgets CFOs want greater accountability tied to business outcomes, without multiple methodologies. Wadhwani says causal AI can be there to help close some of that gap by evaluating brand, performance, and operational variables within the same framework rather than through disconnected reporting systems.

The Winning Olympics Example

The winning example Wadhwani shared involved a major airline running campaigns tied to the Paris Olympics. The business question was straightforward enough: did the investment drive bookings?

Using its broader modeling approach, Alembic analyzed booking behavior against multiple business and marketing variables happening simultaneously throughout the campaign. Surprisingly, even for the client, the strongest driver of bookings was not a traditional ad placement or search activation. It was the airline’s logo visibility during medal ceremonies, right on the ribbons. Those moments of pride and appreciation carry unusually high emotional engagement and attention. Even casual viewers tend to stop what they’re doing and watch. It ‘s very different from a commercial break.

They saw the biggest uptick in purchases of flights to Paris thanks to the ceremony!

Wadhwani also noted that the same type of modeling can identify the opposite effect and offer insights for sales strategies based on moments or environments that may hurt performance instead of helping it. Alembic can identify nodes that reveal a negative impact.

“You might have changes in your pricing strategy, as an example,” Wadhwani says, “and create a negative effect suddenly on your sales. Maybe you started adding discounts and over time that impacted on the value of your brand itself from a consumer perspective because in their mind they’re always expecting a discounted rate. That might lead to lower revenue in the long term. That’s a very common example which we see through data.”

Companies can begin to understand not only what drove results, but what may have worked against them, whether that involves timing, placement, surrounding content, or even promotional discounting that unintentionally weakens long-term value.

“That level of granularity does not come through from any other measurement tool I’ve seen in the market,” Wadhwani told me.

The example reinforced something marketers sometimes forget amid all the conversations about automation and AI: context and emotion still matter enormously in advertising.

Beyond Marketing

Wadhwani described organizations using similar approaches to think about inventory management, demand forecasting, and operational planning. He also said that creator and influencer marketing can be incorporated into the same framework when the data exists to support it, which could become increasingly important as more marketers shift spending toward creators while facing continued pressure to prove business impact.

With Wadhwani on board, NVIDIA as a compute partner and continued investment in infrastructure, Alembic is clearly betting that causal reasoning will play a larger role as organizations try to make better sense of increasingly complicated data environments.

Posted at MediaVillage through the Thought Leadership self-publishing platform.

Click the social buttons to share this story with colleagues and friends.
The opinions expressed here are the author's views and do not necessarily represent the views of MediaVillage.org/MyersBizNet.