That question started during my PhD at Alliance Manchester Business School, where I spent four research cycles studying how people form emotional connections with music and with each other. The first cycle was pure inquiry: a music memory survey exploring whether other people experienced the same emotional recall with music that I did. The findings were small in sample but striking in what they revealed. Two respondents, a young Taiwanese man and a middle-aged British man on opposite sides of the world, with no connection to each other whatsoever, had submitted the same song with the same emotional core: a father, a family, a moment that mattered.
No algorithm would have grouped those two people together. But at the level of human meaning, they were saying the same thing.
That finding shaped everything that followed. The survey made clear there was something worth building, so across the next three cycles I built three distinct platforms, each one a direct response to what the previous version taught me. The first, My Oopus, started as a simple mobile community for capturing music memories. The second, Your Life With Music, expanded that into a richer web experience combining social features, contextual listening data, and a more personalized sense of each user's relationship with music. The third, Instatune, was the most evolved version of the idea: a platform where social context, behavioural signals, personalized recommendations, and community participation all worked together in one place.
Each iteration got closer to something I was reaching for but couldn’t yet fully name. The idea that a system could understand the person behind the behaviour, not just the behaviour itself.
The research wasn’t just theoretical. I was designing, launching, gathering users, seeing what didn’t work, and rebuilding. I didn’t know it at the time, but I was running a version of the Lean Startup methodology inside an academic framework: building, measuring, learning, and going again.
One of those platforms, Your Life With Music, was later used in peer-reviewed physiological research examining how personalized, mood-matched music affects metabolic rate and cortisol levels. The findings were striking. Music chosen to match a person's emotional state produced measurably stronger effects than randomly selected music. It was another early confirmation of something I had been circling for years: context and meaning change everything. Data without understanding the human behind it is just noise.
After my PhD I co-founded two venture-backed companies in Manchester. Moment.Us was a consumer music platform that integrated behavioural and environmental signals to personalize the listening experience. It was selected into the Collider Accelerator Programme in London and raised early-stage funding including investment from Unilever Ventures. Personalyze was a behavioural intelligence platform serving enterprise clients across the UK, Europe, Asia, and North America, raising approximately £1.6 million in a round led by OpenOcean. Between them they taught me more about human behaviour, data, and the gap between technical accuracy and genuine understanding than anything else I’ve done.
At Personalyze I used to remind my team that every data point is a person. It was the reason to handle what we collected with care and empathy, and it’s the sentence The Data Human is built on.
I now teach at Centennial College in Toronto, where I run courses in entrepreneurship, digital marketing, analytics, and AI. I founded the Digital Insights Group, a student-led consulting initiative that brings real client work into the classroom. I also teach as a visiting lecturer at universities in the UK and France, most recently at Liverpool John Moores University Business School and INSEEC Grande École in Paris.
The Data Human grew out of all of this. It’s not a rebranding of ideas I read somewhere. It’s the framework that emerged from fifteen years of building things, watching them succeed and fail, teaching what I learned, and asking the same question from every possible angle: are our data systems designed to understand people, or just to measure them?
I’m also building Siftbox, which is the same argument from the other side of the exchange. It gives people a dedicated address and lets them decide, brand by brand, what marketing email actually reaches them. You should know I build it, because the Index scores companies that send that mail. It cuts the other way too: Siftbox measures what brands actually do rather than what they say, how often they send, how hard they push, whether an unsubscribe is honoured, and that is harder evidence than anything I can gather from the outside with a browser.
I should be clear about where I stand, because it isn’t where people usually expect. I don’t think collecting data is the problem. Extraction is: value taken without the person understanding it, agreeing to it, or getting anything back. Brands can’t take everything, and people can’t expect to be known while giving nothing. There’s a fair deal in between, and it only works when both sides can see it.
That argument is the subject of my book, The Consumer Hypocrisy, which is about the half of this nobody likes discussing: we say we want privacy and then choose the experience that knows us. I’ve been on both sides of the exchange, building the systems and then asking what they felt like to be on the end of, and I don’t think either side gets to be the villain.
The Data Human Index is one practical answer to that question. This site is where I think in public about the rest.