AI in the Advertising Agency Landscape

A few days ago, I read an article published by dircomfidencial that covered a round table with senior executives from some of Spain’s leading advertising agencies. The discussion focused on how AI is impacting the agency business.

I have to say that I agree with them: AI is not just about giving employees access to a very powerful generative technology. It is a cultural transformation — similar to, and perhaps even greater than, the shift agencies had to face at the beginning of the digital era.

From that point on, however, I honestly feel they start to dodge the real issue and do not fully say what is really happening. Let me explain:

For years now, the “traditional” advertising agency model has been under pressure and needs a serious rethink in terms of its long-term viability. Broadly speaking, the days of excessive client budgets, limited transparency around results, uncontrolled creative theatre and oversized agency structures are over. Advertisers have evolved: they are more knowledgeable, more focused on cost control and, above all, demand far greater visibility and accountability.

For years, we have been seeing major agencies constantly cutting teams, merging brands and shutting some of them down altogether. In other words, agencies have already been forced to tighten their belts for quite some time.

Now, with AI, things become even more complicated — another turn of the screw. The major shift is that clients themselves now have access to tools that allow them to take on projects which, until very recently, necessarily required professional or specialised expertise. In other words, clients now feel more capable.

Careful: I am not saying that the results they achieve will necessarily be exceptional. Having access to the tools is one thing; knowing how to get the most out of them is another. But it does mean that:

a) For many projects, that “good” rather than “excellent” approach is enough — especially when we are talking about tactical digital actions.

b) Clients have more knowledge, and therefore want more control.

c) Clients believe — not always correctly, and I would actually say often wrongly — that everything is now easier, faster and, as a result, cheaper.

In other words, having AI tools does not mean they will do everything, or that they will do it well. You need to know how to use them and how to give them the right inputs — the famous prompts — for that to happen.

That said, I would also say this is true: many mechanical tasks can be automated, and many activities can be made faster. Inevitably, this means that, over the medium term, some of these tasks can become cheaper. Not in the short term, though, because the automation process itself also comes at a cost.

That said, reading the article, some old ghosts start to reappear:

a) One option they put forward is to create new agency brands that they say will be specialised in AI. But this feels very similar to what happened when agencies started creating digital, performance or other specialist brands. In many cases, it became a way to dress things up, rather than being forced to explain costs clearly to clients and provide real transparency.

b) On the other hand, they say that the arrival of AI requires structural adjustments, but then quickly add that this does not mean “headcount reductions”. Instead, those resources — those people — will now be dedicated to other tasks in greater depth. Really? Are we sure? Hardly a day goes by without news of job cuts at major agencies, and even at technology companies — the very companies driving AI forward. And “in greater depth”? Does that mean these tasks were not being done with enough care before? It sounds a lot like bullshit.

c) Not hiring profiles specialised in AI is, in my view, a mistake — just as it was a mistake not to hire digital profiles years ago. Training, upskilling and reskilling are essential, of course. But a transformation of this scale also requires new knowledge and new types of talent. Agencies that do not invest in that will, quite honestly, be making a mistake. We have seen this happen across many sectors — for example, in the media industry during the digital shift — and most of them ended up crashing into that reality.

In short, agencies have been facing a business model challenge for years. It requires a different approach to client relationships and inevitably involves adjusting — or restructuring — both teams and cost breakdowns.

Simply dressing it up is just prolonging the agony — and I say this because we see it in our day-to-day work. AI in the advertising agency landscape is a major challenge, but also another clear signal that this transformation process is no longer optional.