Applying market research to product strategy


One of the toughest tasks in the tech industry is using market research to help you make product feature decisions. A lot of people in tech feel you shouldn't do it at all – that customers aren't capable of envisioning what could be built, and if you ask them you'll just distract yourself from your vision.

On the other hand, there are a lot of companies that use market research very heavily in their product decision-making, so much so that they're reluctant to make any decisions without tangible evidence to back them up.

I think there's a middle ground, in which you can use research to help inform your decisions but not to make them for you. If you're interested in the details, you're welcome to visit the latest chapter in my online book-in-progress, Stop Flying Blind.

4 comments:

Douglass Turner said...

Hi Mike,

I actually think this period in histoyr has the potential to be the golden era of product/service development. I say this because we have entered the MySpace/YouTube era in which a significant portion of peoples behaviour takes place in public on the Web where savvy companies can sift, analyze, and use as a guide in development of new product/services.

We are way beyond focus groups here.

We are living side by side with peoples experiences with a level of transparencey and insight unheard of just a few short years ago.

Never before have we had such an opportunity on a global scale 24/7 in realtime. The more I think about it the more stunning this development seems to me.

Perhaps you can say a bit about this in your new book Mike.

Regards,
Doug

Dean Bubley said...

Hi Mike

I think your post (and new book chapter) make a huge amount of sense. I suspect over-simplistic approaches to designing surveys and interpreting their results are the main culprits of both poor product design, and those publicised "research studies" that have results that seem outrageously counter-intuitive.

Perhaps another way of slicing it is to say that most products are designed for (& bought by) either "early adopters" or "the massmarket", but not both. Lumping both groups together in market research will therefore yield a meaningless average.

On the smartphone example I'd possibly go a bit further than just 3 segments (eg dividing entertainment-centric between music vs image/video), and perhaps adding a category of "smartphones bought by companies for employees" (IT-centric?)

Michael Mace said...

Douglass wrote:

>>a significant portion of peoples behaviour takes place in public on the Web where savvy companies can sift, analyze, and use as a guide in development of new product/services

It'll be interesting to see how much of that information people will be willing to share with companies to help them design products. Based on what I've seen so far, many people are willing to trade a lot of personal privacy in order to get even small amounts of customization.


>>The more I think about it the more stunning this development seems to me.

Good point.

The most stunning thing to me is how many major changes seem to be brewing at the same time, all catalyzed by web apps and related things happening in the tech world.

The changes are fast-moving and complex enough that it's hard to tell what all this change will add up to.


Dean wrote:

>>most products are designed for (& bought by) either "early adopters" or "the massmarket", but not both. Lumping both groups together in market research will therefore yield a meaningless average.

Yes, I agree completely.

I'm assuming that by "early adopters" you mean the bleeding edge technophiles like me who hang around on discussion boards. It's kind of interesting to note that in Everett Rogers' original literature on the adoption curve, "early adopters" meant more mainstream people who actually are a good proxy for mainstream users.

The people we call "early adopters" today were labeled "innovators" in Rogers' work, and he recognized that they're not a good proxy for anyone else.

Most engineers in Silicon Valley fit in the Innovators category, meaning they don't have the right instincts to design for normal people unless they get some outside data. That's the best use of properly-designed market research, at least in high tech.


>>On the smartphone example I'd possibly go a bit further than just 3 segments (eg dividing entertainment-centric between music vs image/video), and perhaps adding a category of "smartphones bought by companies for employees" (IT-centric?)

I think you're right -- and I bet there are even more segments than that.

The number of segments you can identiy is limited by the size of your survey sample. In the research we did at Palm, the sample size of ~2,000 people was enough to pick out only three smartphone segments.

Douglass Turner said...

Dean and Mike,

Regarding this early adopter / mass market issue I'd like to ask a question? Were the teenagers in Copenhagen/Stockholm/Reykjavik/Helsinki that stumbled upon SMS as a way to keep tabs on their "tribe" early adopters?

I am specifically thinking of the armies thirteen year old girls swarming the malls after school.

Man, that is about as far as one can get from the classic Silicon Vally mental model of an "early adopter".

For context, back in the day, when I was at Apple in the '90s nobody had a clue this phenomenon even existed.

I think this phenomenon dovetails nicely with the MySpace explosion that errupted out of nowhere and is clearly a similar development: seriously non-geek types appropriating technology for their own (unpredicted) needs.

I think in the modern era we will see this pattern repeating over and over again. It will be *the* dominant theme for how digital technology gets used and provides us with clues for what a successful products shoule look and feel like.

-Doug