Mindset beats audience. Relevancy and receptivity are the two R's. And if you can't deliver results for brands, there is no reason to be in business.

The origin of the name is highly debated, and probably only a couple of the founders know the truth. GumGum is almost 20 years old and, at its heart, an advertising technology company. The heritage is computer vision and AI-driven contextual understanding of the moment an ad is served. Over the years the company built a huge amount of technology and business with customers, and today Startzel frames it as the mindset company, on a mission to understand mindset in a world where digital advertising has not always worked on brands' behalf. He was appointed CEO just before Cannes last year and is coming up on one year in the role.

The industry has talked about artificial intelligence for the better part of 20 years, machine learning and algorithmic bidding. What changed was the expansion of ChatGPT and the progression of large language models. Twenty years ago, selling performance advertising, Startzel was already saying right person, right ad, right time. Now, he argues, there is a real opportunity to transform the way technology impacts ads and to understand more about a person. Before he joined, the company had built the Mindset Graph, which lets customers see what is happening inside the ads, where they are working, who is responding and how a brand is mentioned against competitors.

A person today is a collection of the audiences they belong to online, as measured by platforms and algorithms. Startzel quotes an agency CEO friend: the algorithms are flattening the brands, so it takes more and more work for a brand to pop. He is a father, a CEO and a classic car restorer, and he is not in any one of those mindsets all day. That, he says, is the big opportunity.

He distils it into two R words, relevancy and receptivity. He is part of some audience 24 hours a day, but he is interested in different things at 7am than at 7pm and responds to different ads. Given the signals in context, in the environment and from the user, there is the chance to do better and really engage.

The company acquired the attention measurement business Playground XYZ and integrated it into the platform, so for every ad opportunity GumGum predicts whether someone will pay attention. For a brand or agency that seems simple, but it is not trivial. Beyond that, Startzel cites a 3 or 4 times increase in relevancy around brand impact, and the ability to measure the attention required to drive that lift. Customers love brand impact and engagement, but they want outcomes, so the platform feeds vector embeddings to get smarter and find target audiences in the moments that matter.

The journey ends with a customer who achieves their goals, so that is where you start: deploy data scientists and technology to understand mindset, then apply it omnichannel to get the outcomes. GumGum may have the smartest technology and the best theoretical understanding of mindset, but if it cannot deliver results for brands there is no reason to be in business.

Coming from Amazon and its obsession with the customer, Startzel let direction come from customers. He heard GumGum was well-liked, with strong NPS. The gap was focus. The business was heavily concentrated in display advertising, a small fraction of where brands actually spend. With streaming and the fracturing of attention, being omnichannel matters more than ever, and that is where the growth is coming from.

The biggest change is the choice consumers and buyers now have. When Startzel was growing up there were three US networks, NBC, CBS and ABC, choosing what America watched in primetime. Today there are creators, women's sports, as his customer e.l.f. Cosmetics described, and endless streaming services.

A lot has not changed. Platforms that deliver results for advertisers create gravity, from AOL, Microsoft and Yahoo to Amazon, Meta, Google, Netflix and Snapchat, with OpenAI on its way to being the next big platform. The names change, but the oldest tenet of advertising is constant: you have to advertise where the people are, where the eyeballs are.

Trust comes back to choosing the customer. Build optimisation systems that drive conversions, differentiation and brand lift and you are doing right by brands. Build optimisations just to claim credit for conversions that were going to happen anyway and you are not building trust. With near endless choice for where people spend money, you have to build with integrity and be transparent, starting internally. GumGum's high NPS is the proof point: a 9 or 10 means you would recommend the company, and you do not answer 9 or 10 if you do not trust it.

From the Naval Academy onward, the saying that stuck is ship, shipmate, self. Every decision should be what is best for the ship first, then shipmates, then self. Shipmates do not always like a decision, but they recognise the intent. The submarine, contrary to the Hollywood picture, was flat and inclusive: the chain of command was real, but everyone from captain to enlisted men could share an opinion, because hundreds of feet down you count on each person doing their part.

Decision-making is one of the most important things a leader does, and it needs speed. Some choices are two-way doors you can reverse; others are one-way doors you cannot. But there are moments, especially around people, to slow down: everyone carries bias from everything they have heard in life, so listen first and learn before making changes, while staying impatient about product, sales and delivering for customers. Calm leadership, and telling your team you are not always right, creates trust and space for real debate.

AI is an incredible opportunity, and the winners will be the companies that deploy it to better understand mindset and to make sure ads hit the moment with true incrementality and outcomes for brands. GumGum tests new products with customers like e.l.f. and Archer Food snacks, and keeps innovating because it is never satisfied, with cycles moving much faster and new questions around agents, data usage and privacy. Startzel believes GumGum can transform advertising through mindset. It is not a certainty, but with a headstart of decades of contextual computer vision, he is excited about the opportunity.