What Can Designers Learn From Marketers?

I've always argued that product marketers and industrial designers in essence have the same job, albeit with a wildly different approach. I also feel we're sleeping on a lot of cool stuff from the seemingly boring world of market research.

I might be kind of an exception from the "typical" industrial designer. Most people will imagine it as a creative hipster that makes dazzling sketches and has some mystical way of designing awesome products using fluffy language and "feeling", and there are definitely a lot of those out there.

I on the other hand am a very pragmatic person (or so I tell myself) and am rather allergic to these types, especially the pseudo-intellectual ones with a god-complex, but let's not get too personal, shall we?

I firmly believe strategy and design can be extremely rational and fact-based and can be brought down to a science. Now, as it relates to our superior-feeling designers, I'm of course jesting to an extent, but we often tend to design from "the gut" and wear this as a badge of honor, priding ourselves in our user-centered design process.

When defining which user needs to address, what features should make it in your MVP and creating user personas, we prefer and sometimes even glorify qualitative methods such as workshops, interviews, user-centered design sprints, … And most of those are really valid, but there often seems to be a lack of self-awareness and I've seen quite some of these that are too fluffy or self-engrossed and unable to cut to the core of things.

All these tools serve a purpose and are useful, but it is one of many ways to come to try and achieve the same thing, and by far not the only way. If we look at our market research colleagues, there are certain nuts they've cracked decades ago while we're all trying to reinvent the wheel and come up with fancy frameworks and workshops, some things are just solved, period.

For instance, if it comes down to defining persona's or customer segments, a great designer will tell you that you shouldn't rely on demographics and psychographics, but on people's needs, values and willingness to pay…

And they're right, because a great market researcher will tell you the same! But where the designer would create an elaborate collection of sprints, workshops and qualitative tools, the market researcher would go to town with a MaxDiff Analysis and Latent Class Clustering.

What the hell is that, you ask?

To put it plainly, these are statistical tools that allow you to identify which factors users value most when considering a (new) product and to create clusters (or segments) of people that have the most similar set of needs.

You can literally start from a 100+ features you *might* want to put in your product and figure out which are more important to who and how large these groups are. This can help you define (a set of) products taylored to specific niches or segment of our market.

Take the below example of a paper company (Source: the amazing book Monetizing Innovation). Where they initially had the same "product" for all their customers. After applying the above techniques, they discovered certain segment based only on values, needs and willingness to pay. Which allowed them to more effectively capture value from these different groups by offering taylored products. This plays nicely into my Smallest Viable Audience theories as well.

Actual needs when grouped per cluster of similar-responding people

Product offerings based on cluster

A next level of awesome is if you evolve this toward a Conjoint Analysis, when you start to factor in willingness to pay and trade-offs. You can build data-backed "what-if simulators" allowing you to build different product variations to figure out the volume, revenue and profit of each variant you create and defining the ideal set of features for your product(s).

Biggest drawback is that this works really well for mass products, but is much more difficult to pull off for low(er) volume and more niche products. Mainly related to the cost of research tools and sample size (although this is changing rapidly by affordable upcoming AI research tools). When your developing for a specific type of surgeon or niche professional, these are hard to find audiences when you need a significant sample size for these tools.

To conclude, this is not to say that the qualitative models are faulty, it's even essential to start from to create proper quantitative experiments. It's simply about the right tools for the right job and only very few designers (or industrial design agencies) realize or even know that there are extremely suited, proven solutions out there for certain types of cases. Even more so, a lot of them tend to thrive on their "secret sauce" methodology, but I'm fairly certain, depending on the agency, it's 50% fluff and 50% actually useful stuff.

So cut the fluff, don't get bullshitted and focus on what matters and gets results!