Reach the moment, not the bucket. Business intelligence will outgrow advertising. And the best leaders get out of the way.

Mobsta used to be a mobile organisation, targeting people on mobile using device IDs. Over the last five years that has changed. Now it is about combining as many data sets as possible to figure out where people are most likely to be, on any channel: digital, outdoor, audio, connected TV, mobile display.

We just talk about the perfect moment. There is not a single person we can reach at any time. It is about being seen when someone is in their most perceptive state, because brands all around you hit different emotions, and you have to find the moment to tap into that so the recipient is most likely to engage.

Audiences are very different throughout the day. I am a CEO at Cannes now, but this time tomorrow I will be home taking my kids to cricket, and the brands I want to engage with have changed completely within 24 hours. You cannot just put people into buckets and hope they stay there forever. That is what we are trying to solve.

Three and a half years ago I asked our head of product what happens if a big client pulls out, and he said we would be in serious trouble. So we changed it. We built our own activation team and tech platform, went from 30 people to 60 plus, and grew revenue three times. I try to instil that there are no boundaries on what we can do.

I describe it like being on a yacht. You head for the horizon knowing you will never reach it, but you might see a nice little port you were not planning to stop in, with the best local restaurant you have ever been to. You do not plan for those moments, they just happen. That is what we do at work: get people on the journey and stop worrying about targets quite so much. Targets are a measure, not why we go to work.

I did a degree in archaeology, then searched bags for drugs as a customs officer for three and a half years. The shift work pushed me out, so I applied for every civil service job going and got one in marketing. I ran the Electoral Commission's marketing team after four years, moved to GroupM's government agency M4C, went to Dentsu with the account and built a government team from nine people to 80 in six months. It ended badly after the Brexit campaign. I ran the Camelot National Lottery account at Vizeum for two years, then joined Mobsta seven years ago as Client Director, becoming MD within a year and CEO two years ago.

My best boss coached me to leave, running mock interviews even though it lost her me. That taught me the leader I wanted to be. I will always help people move on if we cannot keep them, and I tell new joiners they should do the best work of their careers while they are here. My job is to remove barriers and get out of the way.

Last year we took our own rooftop spot so we could host 20 people for breakfast, lunch and dinner. It feels far more professional than my first Cannes in 2019. There is no other event where this many people in the industry are within a square mile for three or four days, so you can have short, sharp conversations that would take months to arrange in London. What I love is the idea of possibility, conversations with no endpoint.

A lot of media agencies have not realised they are true consultants, talking to finance, procurement, retail and customer success, not just marketing. Their job now is far more of a KPMG or McKinsey style job. If business intelligence is not a bigger revenue driver for us in five years than advertising, we have done the wrong thing. Media buying will largely be automated.

A retailer asked if we could match a research supply that had gone to their CEO every week for five years. It was three charts. We pulled 40 clickable data points in 90 minutes: footfall, average basket size, whether people moved between stores. On its own that is just data. It is the conversations off the back of it that seed the briefs of three months' time. I would have paid them to take that data.

It might be helping a retailer increase basket size, or noticing they are serving the wrong message where baskets are 12 pounds of after-work drop-ins versus a 200 pound trolley shop. It changes street by street. People north of a road may be loyal to one shop, people south of it to another, and they need different messages. You can buy nationally but activate completely dynamically, down to the minute, based on where people are likely to be.

We have become far less risk averse. If ten things fail fast but one works, we get there quicker. When our COO thought he had found a new media channel, I said find a client, run it for free and prove it now. That comes from our B Corp ethos. Everything we do puts people and planet first, which gives us freedom to test without maximising every deal. We charge a CPM, but I still need to figure out how to charge for the intelligence.

The audience is still the same. They are still humans, led by emotions and affinity towards brands. We will never say we will reach this one person at this time and they will buy your product, because that does not exist. But people are predictable in large numbers. At the Electoral Commission, voter registration ran almost identical year on year, within 40 or 50 people out of 41 million voters. You never know what stage of life someone is at, which is what makes us all individual.

Not much anymore. I have learnt that when I panic, it always ends up okay, so if you ride the waves you can deal with it. Now it is my brain whirring with ideas. Last year I woke at 3am, read an article about publisher numbers falling because of Gemini, and spent eight days locked in a study. I came out with a paper that set our agenda around persuasion architecture: how chat agents help people choose brands through the user's eyes, not what a brand thinks they want.

It's all gonna be okay. There were times I really did not think it would be. Talk to people, be open about it, do not keep it in. I would have liked to have started therapy earlier, it would have got me here a lot quicker. You've got this.