Mariana Peneva interviewed Johan Hallin for IRG’s research on AI and growth in December 2024. He opens by placing himself: he has not been acting CMO for five years and now works more with strategy, while remaining involved in what the company does.
His overall read is that the company is at the start of something large, with first trials in a few areas rather than a completed program.
Key findings
- The first application is a copilot for the store network, built with Microsoft, so that staff can ask a question and get an answer instead of searching internal manuals.
- The questions it handles are practical and wide-ranging, from visual merchandising to customer policy.
- The selection criterion was time. The team looked at what consumes time in the store and asked whether AI could reduce it.
- The second application is in design. Designers prompt for ideas rather than final work, and start from something rather than from scratch.
- Customer service is identified as the next candidate but has not been implemented.
Why progress is slower than the ambition
Ownership is unresolved. Asked who owns AI, he calls it a tricky question with no easy answer, and says technology and business should work closely together, which they are only starting to do.
His candid view is that the business side cannot yet see what AI can do, so the pull is not coming from where it eventually should. The head of technology is credited as the person looking for opportunities and driving them internally.
The constraint is prioritization rather than conviction. The company is in the middle of implementing a new distribution center, the biggest investment in its history, and has relaunched its point of sale, both of which absorb technology capacity.
He recognizes the contradiction and names it. Knowing this could save so much time, it should be prioritized much more than it currently is, but some projects have to be finished first, which he says is a struggle for many companies.
Training is the other gap. He says the company does not have the right platforms for internal education in place and has no good method yet for learning internally, which matters because everyone using AI must know how to stay within brand limits.
What he expects marketing to get out of it
On creative work he expects AI to become a natural partner, including on big ideas, working the way brainstorming works. But he insists the human creators still have to be involved, because otherwise you keep finding things that are already out there.
The clearest opportunity he sees is scale. Working across many countries, the company could create one master version and use AI to produce versions for every country, medium and format.
He makes the same point about CRM, where the historic obstacle was the production cost of making different units for different needs and target groups.
The brand argument
His central concern is control. A brand has to take control over AI, which he frames as a prompting discipline grounded in the brand platform, its values, the feelings it wants to create and its keywords.
The risk he describes is convergence. If everyone asks for what most people find attractive, everyone gets the same answer, and distinctiveness disappears. He concludes that working with your brand becomes more important in the age of AI, not less.
His framing of the payoff is two-sided. AI is not only about being more efficient, it can help a company be more distinct, provided human capability is used alongside it.
The Coca-Cola Christmas commercial is his example. It was produced with AI and reduced cost, and it worked because the brand assets already existed and consumer testing rated it highly. He is less sure the same approach would produce something entirely new.
Stakeholders, and what he wants from the study
Asked whether AI can serve stakeholders beyond customers and employees, he answers by starting from purpose and the four stakeholder groups the company uses internally, which he says is a good way to organize the work.
The community example he gives is unresolved rather than delivered: whether AI could help find new ways to engage in fighting period poverty, tied to a headline about meaningful change for women. He says it is definitely possible but has not been explored.
What he wants back from the research is other people’s answers on the same question. Efficiency he considers settled, limited only by imagination. Keeping the brand together, or possibly making it more distinct, is the open question he calls the biggest challenge.
Scope and limitations
This is an automated transcript of a single conversation, and the speaker notes at the start that he is no longer the acting CMO. Company and product names are rendered inconsistently by the transcription.
Source
This page summarises Transcript: Johan Hallin (Lindex) Interview on AI and Growth, by Johan Hallin, Mariana Peneva, December 3, 2024.