Ideas and/or data

Andy Clarke’s post What man, laid on his back counting stars… is a call to recognise the role of ideas and intuition in design. Bravo: we need more of this stuff.

Our industry has reached sufficient maturity to sustain different styles and schools of thought. This can only be a good thing.

My design approach relies on strong fundamentals: signifiers, metaphor, language. I thrive on improving a product over time through prioritisation, and helping a whole team see user experience as a shared concern. My favourite raw materials are user insight, strategy, design theory, and strong team relationships. This approach makes me reliable, if occasionally conservative.

I get the impression that Andy is more of a flair designer, thriving off sparks of creativity and inspiration. His raw materials will be different: perhaps they’ll be time, exploration, synthesis. That’s great. I’m a little envious of that style myself, but it’s just not me. If you want a Creative Director for your agency, Andy’s your man. I’d end up installing a different team and culture.

Andy’s article sets out a spectrum of design styles:

I see things a bit differently:

(I know I’m a manager these days because I find myself drawing quadrants.)

I add an extra axis because I don’t think product design means data-led design any more than web design means idea-led design. To me, the axes are orthogonal.

I should recognise some controversy here. Some folks dispute my separation of product design and website design, arguing they’re the same thing. I’m convinced they’re not, although I do agree there’s a continuum, not two distinct buckets. I’ll write more about this in due course.


The approaches that succeed at company are largely a function of the company’s culture and lifecycle.

Data-based approaches tend to be highly valued in engineering cultures, sales cultures, mature product cultures, larger teams, and companies that have high volume and low margins.

Idea-based approaches tend to be more highly valued in entrepreneurial cultures, design cultures, companies with nascent products, and teams with lower volume and higher margins.

These aren’t fixed patterns, but I see them pretty often.

These valued approaches therefore change as a company evolves. Successful methods of a 50-person startup won’t work at a public company of 3000. As products and companies grow, they typically drift toward data-driven approaches. For that, we can thank capitalism and the nature of large systems: it’s easier to focus on reliable incremental improvement than risky reinvention. As we know, this can also blind a company, leaving it vulnerable to disruption from an idea-led competitor that attacks the problem in a new way.

What gets noticed?

Of the quadrants in my model, some get more attention, for sure.

It’s certainly true that the tech press is more interested in writing about apps than websites. Product companies are the ones talking about venture capital rounds, MAUs, hypergrowth: juicy American dream stuff. Website companies, less so.

But if anything, I think the industry is enthralled by apps built on ideas. The hot startups, the disruptors, the TechCrunch fodder are the companies building on an idea they hope will render their competitors irrelevant. Data-led product design is seen as something for the optimisers, the big boys, the unsexy. Adobe, Amazon, Facebook perhaps. These companies face a big challenge to retain their appetite for bolder bets. Some would argue the endeavour is likely doomed.

But it doesn’t matter what’s hot in the press. The four approaches in my model are all fundamentally sound. As with any discussion about beliefs, the danger lies in the extremes. It’s possible to become so invested in a data-only or idea-only approach that you become blind to the value of fitting your approach to the context.

Product design that’s driven entirely by data is horrible. It leads us down a familiar path: the 41 shades of blue, the death by 1000 cuts, the button whose only purpose is to make a metric arc upward. It’s soul-destroying for a designer. But its moderate counterpart, data-informed product design, is fine. It reduces risk, and encourages confidence and accountability.

Product design driven entirely by ideas is equally painful. The romantic notion of design genius and the Big Idea soon gets swamped by a culture of risk, favouritism, and blame.Idea-informed product design is fine. It provides agility, creativity, the power to see blindspots and seize opportunities.

I fully agree with Andy that we should never lose sight of the aspects of design that give soul to our work. These are just as prevalent in product design as they are in website design. Granted, they may manifest in different ways (mostly motion design for products, often art direction for websites), but design is a fantastic synthesis of ideas and outcomes, and will always be so.