An engineer’s take on the theory of dominant design

Main illustration: Ally Reeves

As an engineer, things that don’t make sense keep me up at night.

I need to find some underlying order in things. Lately, the thing keeping me awake at night have been tailfins. Tailfins don’t make sense. Somewhere in the history of the American car industry, people started buying cars with things like tailfins that served no functional purpose and actually reduced speed due to increased drag.

As an engineer this make no sense to me. Why did something that adds no real value make people buy more cars? What does this mean for the products we create and buy today? Should they all have tailfins too?

The answer to these questions is related to the idea of dominant design.

What is dominant design?

Dominant design is the idea that one product wins in every industry and becomes the de-facto standard. But here’s the kicker – the best design doesn’t always win. (To an engineer this is like saying that gravity doesn’t work in certain corners of this office ? )

Understanding dominant design

Let’s take Microsoft Windows, which became the dominant design in PC operating systems. Before Windows came along, there was no dominant design in the industry, and so there was really no clear way to compare products. Consumers didn’t have a set framework in their heads about what an operating system should look like. It was all new to them.

When consumers don’t have a clear image in their mind of what the product should look like, it’s difficult to differentiate between good and bad versions. As Steve Jobs famously said, “People don’t know what they want until you show it to them”.

It makes sense when you think about it. How can the “best” win if there is no standard to compare it to? For example, before a dominant design in the car industry emerged people had no idea what a standard car should look like. Once it got people from A to B better than a horse it was an improvement. Henry Ford didn’t need the best engine or the fastest car. He needed a car that was good enough to be able to say it was better than a horse. The key here is that before a dominant design emerges, people are not comparing like for like.

Good enough is not always enough

The problem is that when a dominant design becomes standardized, people are no longer comparing you to horses. People have a clear picture of what the product should look like and consumers need a way to differentiate between your product and the next. This is where branding and advertising comes in.

In car advertisements up until the 1950s the manufacturers simply listed the cars many features. That is, until Harley Earl, one of the first car designers and head of design at General Motors realized that instead of making the car go faster they could simply make it look faster.

The result was a tailfin design that gave the illusion of speed, but added no functional value. GM could have squeezed out a few more MPH but that would not have significantly differentiated their product. Instead, abstracting their design from the purely functional gave them a competitive advantage.

What this means for your product

So what can this teach of about building products today? Should you try to add functional value to your product (an engineer’s first choice) or should you add a tailfin?

Firstly, understand whether a dominant design exists

This sounds easy but companies get this wrong all the time. Take the Amazon Fire smartphone. Even an engineering-driven company like Amazon could not successfully compete in the smartphone market after a dominant design emerged. They produced a phone that was “good enough” but it was too late. Yet not long after this they launched the Amazon Echo, an incredibly successful product. A dominant design did not yet exist in market so the Echo only had to be good enough to be successful.

Find a niche that nobody else is serving

If a dominant design does not yet exist, competing head-to-head on sets of features won’t get you far. You are really trying to define a different way of doing something e.g my product does X and Y well.

For example, in the videotape format war of the 1980s, VHS won because it advertised around its two-hour recording time, whereas Betamax only had a one-hour capacity. They could market around the fact that you could record an entire movie on VHS, something that wasn’t possible with Betamax. So even though many people felt Betamax was a technologically more advanced competitor, VHS sold a better way of doing things than any product before it.

It’s not always about being the best

It seems counter-intuitive (especially to an engineer) to not try build the best product we can. But the key is that before a dominant design emerges, there is no standard by which to measure “best” against. It is only after a new standard emerges that consumers start thinking about the “best”. Clay Christensen uses the excellent example of phone cameras. When they first appeared, they took terrible pictures, but they were so convenient that people used them anyway, and over time they got better. Instead of trying to improve an existing technology, they created a lo-fi solution to a entirely new one.

To an engineer, non-functional design does not seem to make sense. What I’ve realized is that things like tailfins do have value, but that value depends on the frame of reference. Tailfins make no sense if you think of a car from a purely engineering perspective. But if you think of it from other perspectives – such as dominant design, marketing, and Jobs-to-be-Done – you can begin to see some method in the tailfin madness. And then this engineer can finally get some sleep.