Everyone has an AI story now, so the story is worthless. The data is the differentiator. Trust is doing what you said and proving it.

I like to take the historic view, because the hype around AI is almost overwhelming. The industry has been here before: the big data era, the cloud era, the big data and cloud era, addressable TV as a big advertising story for years until it became a reality, then machine learning as the fuddy-duddy beginnings of all this in the late 20-teens. Three years ago AI was very much hype cycle. Last year it became real. This year it is not only real, it has moved on to agentic. It has had far more staying power than the other next big things that have come along.

What is different is the existential fretting around it. Those earlier trends brought growth and positive implications. This one carries the possibility of being disintermediated from the ecosystem altogether if the true promise of agentic AI becomes a reality. It is fascinating, and there is a lot to sort through every single day.

Because AI is so pervasive and the stakes cut both ways, it has become a commoditised story. Not everybody needed or wanted a big data story or an addressable TV story, but everyone has to have an AI story, sink or swim, and that has flooded the zone for any messaging you are trying to deliver. So the marketing question is where the differentiator sits. The next part of the conversation answers it: if agentic AI collapses the funnel and agents conduct it, what is left is the data. Just as in the big data and cloud eras, the data is still the differentiator, and each player's data decides how they make money, succeed and survive. At Seedtag we have 10 years of contextual data and have been in AI for a long time, so we are equipped for the story and we have a genuinely differentiated data set.

The key thing I noticed when I joined is that we do not do enough client storytelling. The company is doing well, it is growing, there is energy and the talent is there, so you look for the gaps and the low-hanging fruit. Everyone at the company acknowledged this one. There are not enough opportunities where our clients are helping to tell the story, and there are plenty of happy clients out there who could. Without them it is just somebody banging on about contextual data that has been AI-driven for years.

It becomes real when a client says you were able to detect the interest, emotion and intent of the people targeted with our ads, and therefore beat a standard contextual play. This week we are issuing a case study on work Lenovo did with us, a bake off between standard contextual and our neuro-contextual methodology that shows superior results, verified through measurement partners. You have to show it in action and bring it to life through real examples.

I came in with a dream setup in terms of talent, and it is a smaller company and a smaller team, which I enjoy. At Criteo, after five years, I had got away from doing the work, because if you are doing your job as a leader people rise up around you. You lean in and they say, no, I have got it, and they do have it. Here we are nimble, so I get to be hands-on. There is a ton of work, so everyone says, I will do it, I will stuff tote bags, I will do marketing from every angle. After Edelman, Nielsen and Criteo, which are big organisations with big teams, it is good to be somewhere more tangible. The fundamentals of structure have not changed much: a team on brand, a team on digital channels, a team on comms and reputation. What matters now is that everyone has the same tools and the same LLM access, because AI has become a horizontal layer and the disciplines bring it together to make something original on top.

We are dotted line with go-to-market and with product marketing, and because it is a small company we stay very close to them. The goal is simply how we make the company bigger. No one needs to be famous and no one needs to put themselves in front of the next person. One of our values is one team, one dream. Sales cannot go out without the right materials, pitch and positioning, and marketing cannot do its work unless we have fed into that and understood who sales are trying to reach, so our programs serve their interests too.

With third-party cookies you are targeting on a 13-month cycle at the very longest, more likely six months from a shopper journey. You do not know within that cycle whether the person already bought the shoes or the suitcase that is still following them around the web. We have all shouted at the screen that we bought that last week. Cookies originally gave a broad addressable audience, but that has shrunk substantially through browser restrictions, privacy concerns and people denying cookies on the page. Some say it is down to 11 per cent. It is probably not that low, but it is substantially different.

What Seedtag does is detect, through embeddings on the page and inferences from video, where someone is in their content journey and where they are emotionally, then serve the ad that connects with that moment in real time. AI plus contextual gives you the scale back, and precision is where the work is. Cookies gave you precision because you knew that was an in-market shopper, even if they had bought four months ago. As AI improves you get closer and closer to the purchase decision at the bottom of the funnel. That is neuro-contextual: embeddings creating predictive inferences, a high probability based on maths and correlation that a person is in that state of mind, then targeting on that basis.

Prospects and clients are open and interested, because the AI message is everywhere and they want to know what the differentiator is. At the centre of our stack is our AI, Liz, and across the top is Neuro X, which helps scale it. We have real working AI and good demos, so we build a media plan live. The nerdy fun part for me is inventing a ridiculous campaign: I am a middle-aged mum in Connecticut, I have created a deodorant brand, I want to reach like-minded mums. You type that into chat conversationally and it produces an incredibly detailed plan in seconds. With clients we use their actual objectives, they write the prompt, and out comes a plan that could be shipped and executed. Seeing that it is real opens minds.

Our presence here rolls it up: one-on-one client meetings, thought leadership, podcasts and olfactory workshops. We have two very successful podcasts, some of the top-ranked in ad tech, which surprised me when I found out. The workshops make the brand experiential, so when we say the new funnel is interest, emotion and intent, people feel it. Day to day it is the same mix. Clients come to Madrid for experiential workshops and regular downloads on the product roadmap, which is as much marketing as client solutions, because they learn about our culture and get closer to us. Cannes matters because it is hard to argue it is not the most important conference the advertising industry has, which is funny given it began as an awards show for creatives and now carries a massive business development side.

Trust is won and lost on people doing what they say they will do. You commit to a number, a delivery, a change from A to B, you do it, then you go back and remind people: remember when we said we would change, and we did. Those little pieces build up. Whether your company is in a bad place and needs to get to a good place, or you are launching something long-awaited, much of it comes down to communications, integrated across executive comms, product storytelling and marketing that pulls through exactly what you did.

Everything has to be transparent and verifiable. Two and a half months in I have been asking who our measurement partners are, and how we make sure everything we say we will deliver can be verified. We use a dozen or so, and the Lenovo case study was verified through them. Measurement is the thing that has not changed. It was top of the list at Nielsen, when digital advertising was starting to blow up and standards had to change for it, and at Criteo, where retail and media measurement was core to the go to market. It can become more innovative and nimble, but it remains the foundation for trust, transparency and legitimacy. And it has to be repeated until the market accepts the technology as industrial grade rather than a single test.

You can say one is earned and the other is paid, but the earned piece often does not work unless you pay to support it, so it gets murky. Communications still comes back to reputation, and the pyramid of authority has fundamentally changed. You used to publish a byline in The Guardian and it carried the imprimatur of a professional journalist in a free independent outlet, and people took it as proof. Over the past 20 years that has changed, and it no longer cuts it. So the communicator's tool set is much broader: influencers, creators and press at the same time, and reporters have taken on all those hats too. None of it works unless it is connected to marketing. If your objective is solely a communications objective, that is not real.

People give me guff about a long title, but it is important and robust. There was a brief trend of questioning whether CMO reflects the full job: chief growth officer, marketing and growth, or marketing and communications supporting growth. Maybe I got caught in the buzz, but it represents what we can all do as a team. It means my very important communications team does not feel like an appendage of a marketing organisation, or the other way round.

It all comes down to localisation: where you launch, what stories you tell and how you tell them. Even if you are moving your global brand, you defer to what the local market will say in response and stay respectful of local stakeholders. At Criteo, where we helped develop the commerce media category, commerce is an amorphous and odd word in Japan, a key market, so we had to rethink how we talked about it. At Edelman in India we supported Starbucks entering the country. India was understandably wary of multinationals and had policies requiring joint ventures, so you enter with a partner that benefits the country as well as the company. And a coffee company entering a country known for tea, with indigenous coffee brands already there, has to reflect what the country actually prefers. Starbucks has been wildly successful in the 15 years or so since, because they thought about who the customers are and were respectful as they grew.

This is the first time Seedtag has done a built environment, and even though it is going well we have kept a running document from day one on how to improve it. On the current trajectory, agentic will have taken a step forward and data will again be the key differentiator. Seedtag will keep growing. A lot was invested in the brand before I joined, and the leadership landed with Brian Gleason as CEO, well-liked and respected across the industry, has built on the impressive foundation the founders laid. Hopefully by then we will have moved past being so pearl clutching and out of breath about AI.