AI in marketing needs less hype — and more business sense

by Heidi Byskata | October 25, 2025

Nike’s data chief Linda Cereda told at MarTech Nordic that true AI advantage comes not from models, but from wiring them into people and processes.

The AI paradox in marketing

Companies have never invested more in artificial intelligence — yet few see the returns.

At MarTech Nordic in Helsinki, Nike’s Global and EMEA Head of Marketing Data Linda Cereda described the situation bluntly: “It’s not the best AI tool that drives ROI. It’s how you wire it into the business.”

According to MIT and Gartner, up to 70 percent of AI initiatives fail to meet expectations. The main reasons are not technical: siloed data, missing skills, and poor change management.
Cereda summarized the right balance as “10 percent models, 20 percent data & tech, and 70 percent people and process.”

Nike’s SNKRS case: from fan frustration to data-driven loyalty

Nike’s SNKRS app had built the world’s largest sneaker community — and a wave of disappointment when limited drops sold out instantly. Cereda’s team used machine learning to combine retail and member data, creating a “member engagement score” that identified loyal users most likely to buy and stay active.

The result: a 44 percent reduction in forecast error within a year and 90 percent redemption on targeted offers.

It was, Cereda said, “a shift from rule-based marketing to learning loops — systems that improve every day.”

What’s next: from segmentation to signals

Cereda outlined three structural changes shaping modern marketing:

From membership to meaning. Loyalty will rest on shared values, not just points.
From segmentation to signals. The most responsive brand — not the loudest — wins.
From CRM to learning loops. Data systems must learn in real time, not only report the past.

AI, she argued, allows marketers to move from reacting to anticipating — if they invest in data quality and organizational readiness first.

A clickless, agentic future

Search behaviour is shifting fast. Voice and AI assistants already handle up to 40 percent of queries. Brands must ensure their content is readable not only by people but also by AI engines.
Cereda advised marketers to adopt “hybrid AEO/SEO strategies” — maintaining traditional optimization while writing conversational, machine-friendly answers to user intent.

In such a world, brand trust and emotional storytelling become scarce assets. “AI compresses stories into recommendations,” she said. “Brands need to build memory before the algorithm.”

Watch the full keynote here: