Turn off the lights
⟵ Go back home

The Scientific Method for Business

29 November 2019 • Article

René Descartes stood over a basket of apples. ‘Suppose a man’ he said, ‘had a basket of apples. And he wanted to know which ones which were rotten. Wouldn't he start by tipping out all the apples and discarding the ones that were rotten? So should we tip out our thoughts and opinions and discard the ones that are rotten’. Later, the police arrested René for scrumping.

Descartes's rationalist philosophy states that we must observe the world with absolute skepticism. A lot of what we think, what we do, what we believe to be right is taken as absolute truth. Descartes laid the foundation of the ‘Scientific Method’. Starting with a hypothesis, testing and observing with radical skepticism what you see. Anything that doesn't stand up to the test gets discarded.

So too it is in business. Every business or product decision is basically a bet on human behavior. But how can Design help us get a head start understanding of human behavior? Design is often thought of, even in the disruptive tech world, as it's bi-product. Clean interfaces. Iconic logos. Comprehensive Design Systems. All those things are great and good craft is certainly important. At it's the broadest level, Design is a toolkit for solving problems. But when discussing our biggest business challenges - design is often the last thing from our minds.

All over your company are ideas on how to drive growth. ‘We should do X with our app’. ‘If only we stopped doing Y and did Z we'd be much more successful’. These ideas are small and loosely formed. It's hard for someone to express them to another person clearly. This causes a lot of friction between teams. Even if an idea is lucky enough to live within a road-map, strategy document or PRD - it's often open to a lot of interpretation. And when you're making your biggest bets, you want to be sure about the hand you're going to play.

In most tech companies, our approach to trying new ideas is to bite the bullet and ship an ‘MVP’. The Agile/Lean model of ‘Ship early, ship often’ tells you to pick an idea and go with it. Then you build that idea, launch it and measure it out in the wild. Even though I think this is a great way to build products - it's still quite risky. First of all, the idea you're picking isn't clearly formed or it's a safe bet so you won't fail too spectacularly. A lot of the team probably just went with it, after the umpteenth meeting they had. Second, when you build it, its almost always more complicated than you first thought. When you finally launch and measure it, you are only collecting large quantitative data. This takes a long time to get a good sample, and will only tell you what happened not why it happened.

What we really need is to validate an idea earlier. What if we could get in a lab and simulate, with near-perfect accuracy, what would happen with our idea? What if we could validate our product without writing a single line of code, planning a sprint or launching to the market?

This is where, I think, Design comes into its own. Design has all these tools at its disposal to create a realistic version of our idea and test it with the people who matter. In short, Design can show us if we're going to tank our business or not.

We live in the golden age of prototyping tools right now. Designers can make a realistic interface you can interact with, enter data and tap on in little to no time at all. To me, this is Design's real potential within a company. Think about every product meeting you've been in. The first thing almost every person does is put on their ‘I know the user’ hat and start to offer their critique of will and won't work with this idea. I know because I am an main offender of this. This makes no sense if you're not basing your opinions on any clear data. Instead, if you can just keep quiet and then go away to your desk to prototype it, your team will be much more successful.

Within a few hours, you can create what you heard in that meeting. It almost certainly isn't what everyone had in their head, but you've got a single source of truth to discuss.

To be really effective, prototyping should live in what Daniel Burka calls the ‘Goldilocks level of fidelity’. Where we're not showing sketches, paper prototypes or wire-frames. We're not showing a high-level coded prototype or MVP. The reason is that the former requires you to suspend disbelief. If you're drawing boxes where buttons should be or squiggles where the copy should go - you're not working in objective reality. You're asking someone to imagine what it will be like. The latter, however, invests too much time in an idea that could be scrapped. There's a base human psychology that attaches you to your work if you've spent a lot of time and effort on it. It doesn't matter how ‘professional’ you think you are - you find out how delicate your ego is when someone trashes something you spent months crafting.

Aligning your team and having a shared single source of truth for an idea can be extremely effective for a team. It's an exercise designer's can engage in on their own, in just a few hours. However, the real value comes when you get out there and show it to customers. Once you stop talking amongst yourselves and you see people who use and buy your product react to it - that's when you can strike gold.

User tests are commonplace in a lot of mature design teams. Yet, they are only utilized properly in larger organisations. Large companies set up dedicated labs, with Starsky and Hutch style two-way-mirrors, fancy eye-tracking software that would make authoritarian regimes jealous and dedicated researchers on hand to study participants. This is brilliant and it's great that companies with those resources use them effectively. But you don't need a fancy lab to do user testing. You can achieve similar results by sending over a webpage on Google Hangouts and recording your screen. It's all about asking the right questions, observing behaviors without biases and working with what you learn. This can cost you next to nothing and be done in a few days during your sprint. Processes like Lean UX give great guidance on how to achieve that.

The masterclass of putting this all together and driving real value is a Design Sprint. Design Sprints are growing in popularity, mostly due to their success and innovative way of working. The Design Sprint is a 4-day intense workshop. But what's great about them is that they give you a step-by-step process to use Design as a toolkit to solve big problems and test them. Think of them as a recipe. You can know how to cook but to make a great meal, you need to follow a recipe. Only when you know what you're doing you can start to experiment.

It's understandable why we only focus on the bi-product of design. It's understandable why designers are always asked to work on the same types of problems. But if design wants to be truly effective in organisations, it needs a ‘rebrand’. The dream to me is that companies of all shapes and sizes will think ‘Well obviously I need a Design team’. Regardless of what product or service they're making. Design is a flexible toolkit that can be applied anywhere, and designers should be the proselytizers of that toolkit. This is where Design needs to go.

Design can be more than pretty interfaces and nice logos. It can be a scientific method to build a business.