When the design process becomes a cargo cult

Email this to someoneShare on FacebookShare on Google+Share on LinkedInShare on RedditShare on StumbleUponTweet about this on Twitter

When we look at the output of a successful design process, it’s tempting to think that we can share its success by copying its external, cosmetic details. But in doing this we make the same mistake as the cargo cults of the south Pacific. 

On the remote south Pacific island of Tanna, in the middle of the 20th century, a new religion emerged. 

The religion’s followers did things that we might consider strange. They cut down trees to create space for landing strips. They built bamboo structures that looked like control towers. They fashioned radio headsets out of wood. Day after day, they stood by the landing strips waving sticks in the air, making signals that they hoped would summon planes from the sky.

Why did they do this? Because they’d seen it done before, when American soldiers had landed on the island and performed these rituals. Airstrips were laid, control towers built, mysterious hand signals made, and airplanes had come, bearing cargo that transformed the islander’s lives. So if they did the same things, they’d get the same rewards.

This religion, the “cargo cult” of John Frum, might seem almost comically misguided, but you’d be wrong to laugh. The mistake they made was common, understandable, even mundane.

Their mistake was to confuse the external effects of a process for the process itself. Watching the airstrips and the hand signals, John Frum’s followers thought they were seeing the entire process that brought planes to their island. The real process – World War 2 – was completely unknown to them, so it didn’t factor into their thinking.  

This is a universal human mistake, but it is particularly common in the field of design. A recent example is the Apple-Samsung trial, and the insights it provided into Samsung’s design process. 

Samsung’s executives wanted to understand the reasons behind the iPhone’s success so they could emulate it. But they weren’t curious enough. They looked exclusively at details of the iPhone’s execution – specific colours, shapes, icon designs, scrolling behaviours – and thought that adopting the same details as Apple would bring the same rewards. 

Like the John Frum cargo cult, Samsung mistook the external effects of a process, Apple’s design process, for the process itself. If it had probed more deeply to understand and emulate Apple’s design methodology rather than mere details of its execution, it might not have suffered such a calamitous loss in that Californian courtroom.

Samsung’s mistake was to treat design like a cargo cult, but this is not unique to Samsung. Design teams are in danger of doing this every day.

  • When we look at competitors’ products but forget that the design problems they solved are not the same ones we face, we are thinking like a cargo cult.
  • When we generate deliverables – personas, user journeys, site maps – but can’t explain in simple terms why they’re valuable, we’re thinking like a cargo cult.
  • When we plaster the walls with sticky notes without really understanding why, we’re thinking like a cargo cult. 
  • When we go through the motions of user testing before we’ve seriously considered what questions to ask, we’re thinking like a cargo cult.
  • And when we define success as nothing more than making our stakeholders happy, we’re thinking like a cargo cult.

Breaking free from the cargo cult approach can be painful, especially when some team members believe it is genuinely solving problems. But a good design team must be able to recognise when they, or their clients, are slipping into this kind of mindset, and to take steps to stop it. If you fail to break free of the cargo cult mentality you will be left waving sticks at the sky, looking at the horizon for planes that will never come.

Email this to someoneShare on FacebookShare on Google+Share on LinkedInShare on RedditShare on StumbleUponTweet about this on Twitter

Leave a Reply

Your email address will not be published. Required fields are marked *