We Said No.

It was 1999. I was working at Razorfish, one of the world's first digital design agencies, at a time when the internet was being invented on a daily basis and the roles we held were being invented alongside it. My own title, Information Architect, barely existed before we started using it. It described something nobody had quite needed to name before: the discipline of understanding how information moves through an organisation, and what happens when you try to change that flow.

We had been asked to meet with John Browne, the Group Chief Executive of BP, later Lord Browne, to help him think through what digital could mean for his business.

This was serious territory. BP was one of the largest organisations on the planet. Browne was already regarded as one of the most forward-thinking CEOs of his generation, known for pushing his industry further than it wanted to go. He didn't want a surface-level briefing on the internet. He wanted to understand digital at the level of strategy and operations - what it could genuinely do, and how.

We explored a lot of ground together. But one project stood out.

He wanted to digitise the boardroom.

Not in any simple sense. The vision was ambitious, and genuinely extraordinary for its time.

The idea was to build a system capable of ingesting data from across BP's global operations - production figures, logistics, safety metrics, financial performance across dozens of countries and business units - and feeding it into a supercomputer used at the time to analysis geological data, enabling zero-footprint drilling. It could analyse, synthesise and render BP’s data as an immersive, navigable environment. Board members, geographically dispersed across the world, would put on headsets and walk through the company's data together. They would move through it spatially. They would interact with it, smash different data sets together, run scenarios in real time.

As a design and technology challenge, it was extraordinary. As a vision of what digital could eventually become, it was ahead of almost everything being imagined at the time.

We were genuinely excited.

Then we started our research; and we started to see something.

Buried inside BP's existing process was a team of people whose job was to do something that looked, on the surface, quite modest. They took the vast, complex, noisy output of a global energy company; billions of data points from operations on every continent; and they filtered it. They normalised it. They translated it into language that was precise, relevant and actionable for the people who needed to make decisions.

The eventual output of this process was a twelve-page booklet.

That booklet represented the distilled judgement of people who, between them, carried what I can only describe as collective centuries of expertise - in the category, in BP specifically, in the particular pressures and priorities of each function and each market. They knew not just what the numbers said, but what the numbers meant. Which anomalies were noise and which were signals. What needed to be elevated to the board and what could wait.

And the booklet was portable. It could be read in the back of a car, on a plane, in a helicopter. A board member could pick it up, work through it, return to it, annotate it. It was exactly the right size for the decision it was designed to support.

The supercomputer system we were being asked to build would have replaced all of that.

We went back to John Browne with our findings.

Our advice was not to proceed.

The technology would have cost enormous resources to build and maintain. It would have replaced the human judgement that was, we now understood, the most valuable part of the entire process - the filtering, the contextualising, the institutional knowledge that no algorithm could replicate at the time. It would have made insight harder to access, not easier. And it would have destroyed one of the most elegant solutions I have encountered in thirty years of working with organisations: twelve pages. Enough. No more.

It would have been an astonishing thing to design and build. We told him not to do it.

I have been telling this story at conferences for twenty-five years.

Not as a war story about a famous client. But because it crystallised something I have never been able to un-see. The moment you start with the technology, with what is possible rather than the problem you are actually trying to solve, you are already in trouble. And the more impressive the technology, the deeper the trouble.

I think about that boardroom every time a client tells me they are implementing a new AI system, a new data platform, a new customer intelligence layer — before they have answered the question of what they are actually trying to understand, and who needs to understand it.

We are living through the most significant moment of technological capability in human history. Generative AI, machine learning, real-time data architecture ; the tools available today make what we were imagining for John Browne look quaint.

And yet the same pattern is playing out in boardrooms and marketing departments everywhere. Organisations buying capability before they have defined purpose. Replacing human judgement: messy, slow, expensive, irreplaceable with systems that are faster and cheaper and less able to know what the numbers mean.

Just because you can, doesn't mean you should.

Now, this is not a cautionary tale about technology. Technology is extraordinary. I have spent my career working at the intersection of data, digital and human experience precisely because I believe in what it can do.

But the BP story taught me that the most sophisticated solution is not always the most valuable one. That human curation is not a bottleneck to be automated away - it is often the thing that makes information usable at all. That portability, legibility and judgement matter as much as power and scale.

That lesson is the foundation of everything I do.

It is why I believe organisations must start with clarity before capability. It’s why I think the question 'what do we already have that we are not using?' is more important than 'what should we buy next?' Why I am sceptical of transformation programmes that begin with technology and work backwards.

And it is why, when someone asks me where Intentional Simplicity came from; why I believe so strongly that organisations need to master fundamentals before chasing complexity that I tell them about a twelve-page booklet, a virtual world we didn’t build, and the advice we gave to one of the most ambitious CEOs of his generation.

He heard us out.

He nodded.

And he didn't build it.