My 8 predictions for our industry in 2025

(including a bet that Omnicom’s acquisition of IPG won’t complete.... here we go!)

Bye AI PoC, hello Scale.?

2024 was the ‘Year of the POC’ with clients nervously testing the AI waters. Early adopters are now wading in and, finding that the water is indeed lovely, in 2025 most will finally dive in. One of the biggest impacts of this shift will be a dramatic compression of the content supply chain. Compression of speed, cost and errors, and expansion of personalisation and quality. AI is all about crushing management’s ‘iron triangle’ and just doing things better, faster and cheaper.?

Play the Orchestra

In the new AI-at-scale world, the most valuable agency skill will be Orchestration. Marketing industry adoption has not been a technology challenge, the tech is already ahead of the marketing use cases that it will solve. Instead it’s an orchestration challenge - how to make the work work, how to de-fragment the constellation of agency scopes and remove wastage to fundamentally change the economics of your marketing. And with orchestration you can…

Set Creativity Free

We are going to see a continued shift from integrated vertical brand assignments with holdcos, where mediocre ideas are the price of efficiency and consistency, to a horizontal model with orchestration layers liberating creativity. Indy creative hotshops will flourish again with the lowest barriers to entry and can scale with technology like never before. One only has to look at the model that GM instigated this year, with Monks as the Orchestration layer across all brands, so as to liberate the origination roles for creative shops like Mother, Preacher and 72andSunny. And all the holdcos are out. This model will accelerate - brilliant ideas from wherever you find them, then executed perfectly by an Orchestration Partner that gets the right version of that idea, to the right person at the right place and time at the lowest cost, highest speed and utmost brand safety.?

Search becomes Find

Once they master their hallucinations, LLMs will be our first place to find anything and it will feel like a new internet of borderless discovery and learning and inspiration. Enjoy it while it lasts, since an early prediction of 2026 is that the race to monetise LLMs and agents to pay off the tech debt that is currently being accumulated will worsen the experience with pay-to-play solutions for advertisers like Perplexity is currently testing. ?

Glass boxes, not black ones

The platforms will innovate to disintermediate media agencies for direct to client partnerships. P-Max and Advantage Plus are still the early early beginnings of this, for the SME marketers who are the lab rats before it goes full Enterprise mode. Bringing visibility to what runs where and why, and what performed and why, in real time, will be a glass box approach to media in this black box world. This will accelerate interest of clients in the more opaque ways that large media buyers make money, which is not often in the interests of their brands, and they won’t like what they learn.?

Repeat after me: Multimodal AI Agents

The Cannes 2025 buzzword will be ‘Multimodal’, replacing ‘Generative’ from 2024. It’s all about agentic tools and workflows, using AI to stitch together the unnecessarily complex work of setting a marketing objective then manually lining up messaging, channels and data to execute against that brief. Agents will be born, they will learn and grow, and they will let us be more human as a result if we harness them properly.?

Blockchain’s Back, Baby

Now that the distraction that was the Metaverse has faded (how are all you Chief Metaverse Officers out there faring by the way?) we can get back to the magic that is blockchain technology. Blockchain will underpin agentic workflows, but so much more - media impression attribution, rights management, glass box media - now that the rising L2 blockchains are fast enough and scalable we can know what happened when and where, and trace resulting actions like never before.?

Lastly, While Rome Burns

WPP will do the opposite of Omnicom and sell some things instead of buy them, after finding out that their current minority stake in Kantar is worth more than the whole thing when they sold most of it to Bain to fix. They will do this again. Omnicom’s acquisition of IPG will pass regulatory muster, and even cultural clash - but it will fail for investors, who may still express themselves strongly enough to stop it. Why? There is rust under the panels of the business that were buffed for sale, without time and openness to look under the hood properly to date, the two growth strategies that are both smoke and mirrors will be apparent. Talent will have departed with energy and drive to do their own thing, and clients will ask for more for less because they have to tolerate distracted leadership teams looking at integration (and worried for their own jobs) and will be asking why these $750m in annualised synergies are gains being passed to shareholders, and not to the clients who pay for everything around here. The breaks in the deal terms will kick in when the value of IPG drops through them, and the desire for Acxiom to save Omnicom’s media capability will not be enough to see the deal through.?

That’s 8 from me. 8 that add up to a fascinating year in this eternally surprising, challenging and inspiring industry where we make people pay more than they should, more often, for things they don’t need. But it’s fun, right? Onwards!