Stop Burning Ad Budgets: What a 20-Year Market Research Veteran Wants Every Brand to Know About AI Consumer Intelligence

 There's a particular kind of panic that sets into a boardroom when someone finally says out loud what everyone's been thinking: the data we're using to make decisions is already old news by the time it reaches us.

That's roughly where my conversation with Ankesh Saxena started.

Ankesh isn't a hype merchant. He's spent two decades inside some of the most respected names in market research — Gallup, Lucid, Survey Sampling International — before co-founding Acutus AI Insights in August 2024. He's a certified ESOMAR and MRSI member, hosts The Collaborative Canvas Podcast, and runs The Research Pulse Wire. In other words, this is someone who has watched the research industry from the inside for a long time, and he's telling us the old playbook is done.

I wanted to unpack why — and what brands should actually do about it.

The Old Way Was Already Broken

Here's the uncomfortable truth Ankesh laid out early in our conversation: traditional consumer research typically takes six to ten weeks to complete. Along the way, more than 40% of respondents drop out before finishing. And once the findings finally land on a strategist's desk, the market has usually already moved.

Think about that for a second. You commission a study to understand what your consumers want right now. By the time you get an answer, "right now" is two and a half months in the past. In a market that changes as fast as India's does, that's not a small gap — it's a strategic liability.

This is the problem Acutus AI is trying to solve with two products: QMAPI (Qualification Mapping API) and ZCoded, a synthetic data prediction engine that builds digital-twin consumer personas. Instead of waiting weeks for a panel to respond, brands can run decisions against AI-modeled personas that reflect real consumer behavior patterns — compressing a process that used to take months into something closer to real time.

Synthetic Data Isn't a Buzzword — It's a Shortcut Around a Broken System

When people first hear "synthetic data," they assume it means fake data standing in for real people. Ankesh's take is more nuanced than that. It's about using AI to model and predict consumer behavior based on patterns that hold up across huge amounts of real signal — so a brand isn't just reacting to what happened last quarter, but getting a workable read on what's likely to happen next.

The bigger shift, though, isn't just speed. It's quality. Ankesh talked about how AI-driven qualification changes the game from the very first interaction with a respondent — filtering out bad-faith or low-effort responses before they ever pollute the dataset, rather than trying to clean up the mess after the fact.

GEO: The Marketing Emergency Nobody's Budgeting For Yet

If there's one idea from this conversation that should make every brand manager sit up, it's this one: Generative Engine Optimisation, or GEO.

For the last two decades, brands have obsessed over ranking on Google. Fair enough — that's where people searched. But increasingly, people aren't searching. They're asking. They're typing a question into ChatGPT, Gemini, or an AI Overview and taking whatever answer comes back as gospel.

Here's the number that should worry you: 80% of consumers say they don't fully trust AI-generated answers. And yet those same consumers are running roughly 40% of their searches through AI tools anyway. That contradiction — mistrust paired with heavy reliance — is exactly the space where brand perception is quietly being rewritten. If an AI assistant doesn't mention your brand when someone asks "what's the best option for X," you don't just lose a click. You lose the conversation entirely, and the consumer never even knows you existed as an option.

Ankesh's advice here is blunt: audit whether your brand is actually being surfaced in AI search results today. Most marketing teams have never even checked.

Agentic AI Buyers: When the Customer Isn't a Human Anymore

This is the part of the conversation that genuinely reframed how I think about the funnel.

We're moving into a world where some consumers set up AI agents to shop for them — comparing prices, checking reviews, and completing purchases without ever seeing an ad, browsing a website, or being swayed by a campaign. The agent makes the call based on data, not brand affinity.

If that sounds like a small niche behavior today, Ankesh would tell you it won't stay that way. And it raises an uncomfortable question for every marketing team: if your entire strategy is built around influencing a human's emotions and attention, what happens when the buyer on the other end is a piece of software that only cares about verifiable facts, pricing, and specs?

The First-Party Data War Is Already Underway

Third-party cookies are dying — that part isn't news to anyone in marketing anymore. What Ankesh framed differently is why this matters more than most teams realize: zero-party data — information consumers deliberately and willingly hand over — is becoming the real competitive moat.

The brands that win in this next phase won't be the ones with the biggest ad budgets. They'll be the ones that gave consumers a good enough reason to voluntarily share what they want, what they're deciding between, and what would actually make them buy.

The Trust Paradox Runs Deeper Than AI

There's a psychological thread running through all of this that Ankesh calls "treatonomics" — essentially, the study of how trust (or the lack of it) shapes economic and purchasing behavior in an AI-saturated environment. Consumers are more skeptical than ever, but they're also more dependent than ever on tools they don't fully trust. Brands that understand this tension — and design their communication to work with it instead of ignoring it — are the ones that will hold onto relevance.

So What Should a Brand Actually Do This Week?

This is where Ankesh stopped being theoretical and got practical. His playbook, boiled down:

1. Check your AI visibility. Ask ChatGPT, Gemini, and Perplexity questions a real customer would ask, and see if your brand shows up at all.
2. Stop treating short-form video as a production problem. It's a consumer intelligence problem — you need to know what your audience actually wants to see before you shoot anything.
3. Build a real zero-party data strategy. Give consumers a reason to tell you things directly, instead of trying to infer everything from behavior.
4. Move from lagging to predictive data. Stop measuring what already happened and start modeling what's likely to happen next.
5. Treat GEO like paid search circa 2010. The brands that take it seriously early will have a multi-year head start on everyone who waits.

The Bigger Picture

What struck me most about this conversation wasn't any single statistic — it was the underlying message: the tools have changed faster than most marketing organizations have. Research that used to take two months can now happen in days. Discovery that used to run entirely through search engines is now increasingly mediated by AI assistants making judgment calls on a brand's behalf. And the consumer at the other end of it all is more skeptical, more automated, and more powerful than ever.

Ankesh's point isn't that everything brands know is wrong. It's that the shelf life of "what worked before" just got a lot shorter — and the brands willing to rebuild their consumer intelligence approach now will be the ones still standing when this becomes table stakes instead of an edge.

🎥 Watch the full conversation with Ankesh Saxena on synthetic data, GEO, and the future of consumer intelligence: https://youtu.be/cUi0hT0J50U