Mistake-Based Marketing

"The great thing about fact-based decisions is that they overrule the hierarchy."
Jeff Bezos, Founder, Amazon.com, quoted in Fast Company.

previous     next           up

Mistake-Based Marketing™

TechTransform, July 28, 2004: I'm often asked if I conform to some formal process or other for product research and development.
The answer is, yes and no. Yes, I strongly adhere to quantitative market research process and the proven positioning principles of Trout and Ries.
But that alone doesn't do it. Here's the rest of the story, as I see it.
Domain Knowledge
The number one requirement for what follows is domain-specific expertise. In other words, you have to recognize that software product marketing is as highly specific as directing a feature film. Sure, you can try to direct a feature film without knowing how, but you will make a lot of expensive mistakes, and it will show in the final result. That's not the same as being a producer, which you can cope through by knowing how to manage. I hope I've made the difference clear. In my opinion, many high tech companies suffer terribly from not having this expertise in house.
Having said that, let's review the survey process in product development.
Successful Surveying
The first thing you have to do is accept that you won't get all your answers from a single survey. That's the main reason why I don't do those huge surveys-in-a-binder that you see. You blow your budget and spend huge periods of time on just one survey - then you're still left with big questions. In the case of one survey I saw about a new product, the surveyors said - well, guess it's not very popular! To me, that was only the beginning of the story.
What I like to do is iterative surveying. ("Iterative" means you just do numerous cycles of the same thing, each one closer to the goal.)
We brought this to a high art at Quarterdeck in the mid-nineties. My brother Alexander Eckelberry and I worked with Michael Bach, Survey.com's CEO. Michael provided us with lightning-fast turnaround. And we ran a very tight analysis and supervision process.
Acquiring an Existing Product
Product manager Jeff Hyman (now at Allume) was releasing WinProbe 95, a new, 32-bit version of a product Quarterdeck had purchased. It needed a new positioning, image, and why-to-buy messages. Only the product name had been changed to add the "95" extension.
To get a positioning, we did three rapid phone surveys, 100 each of our target audience, one after the other. Each time we refined the questions.
Now the existing WinProbe had been packaged by Mike Siewruk's Landmark Research, its publisher, as a toolbox - literally. This breakthrough packaging, done by Jaffe/Soeder, got the original publisher onto the shelves; but it proved unwieldy in wide distribution.
So, after acquiring WinProbe, Quarterdeck had gone back to a regular box with a picture of a toolbox on it. That's where we started testing, with the idea of a toolbox object or image - but this time it got us nowhere.
Unexpected Contender
Eventually, we got a strong "read" on a "Swiss Knife" concept. Ironic, since you think of a Swiss Knife-type software product as being unfocused, doing too many things poorly; but it worked in this case, beautifully.
We also developed the positioning tagline “The Windows95 Troubleshooting Toolbox”, which tied back to the original idea.
Finally, we tested our way to the why-to-buy messages "Solve Windows95 Problems” and “Run at Peak Performance”.
All this was done with brilliant artistic support from the prolific David Brier. For each new design generation we would impression-test overnight in stores, and then whip out a new set of designs the next day, test again, etc. Not all designs madeit into the store; all in all David must have whipped out 100 separate design concepts.
In the end, the winner was what you see here. It took us about three weeks to complete the whole process.
Unassisted Launch
This box was rolled out silently, without a new version or promotion of any kind. But from one month to the next, sales of the product worldwide doubled.
It was the cleanest validation of research-driven product design that I've ever seen.
All this to say that you have to do quick surveying, evaluate, survey with a new set of questions, survey, evaluate etc. It's kind of like having a conversation with the public.
Mistake-Based Marketing
Now Quarterdeck had a budget. Where you don't have one, what do you do? That's where I do what I call mistake-based or mistake-driven marketing.
Literally, you have to test-market your way to success. It requires really spotting those micro-trends so that you can cut your losses and focus on what works.
Also, you have to be very, very close to the marketplace. Say, sitting on a sales room floor or pitching the product yourself.
But it can be done.
Strategic Product Development
First of all, in product development you have to choose your segment. Is it consumer? B2B (Business To Business)? SOHO (Small Office-Home Office)? Are you doing utilities, productivity, workflow, transactional applications?
That really requires understanding your core business and planning based on domain expertise.
In my opinion, many software companies have never really done this - kinda just backed into products, or used the opinions of people who weren't really knowledgeable about their space.
What It Takes
But then assuming you have a space you're in,
Understand who's who out there and what's needed.
Know your industry and your partners.
Know your market metrics.
So, now you intersect these data points. One key factor is: what are you going to use to get your product out there and adopted? Part of it is about what is Free. (See my “Free Marketing” articles here.)
The final step is, of course, end user research. This is pretty easy to do on the cheap when you have an inhouse newsletter audience; but if not, you can try to get an editorial mention in an industry newsletter like W2Knews, which will return a couple of hundred survey responses, and you're off.
Now - how do you get a firestorm of product adoption? More on this soon.
Until then,
Updated August 11, 2004

A product marketer we've worked with says:
...thanks for your 'mistake-driven marketing' terminology. Having joined my company 2 months ago I have converted the development cycle from 7 to 3 month iterations with a heavy focus on early demonstrator/alpha programs to test the concepts with the market.
Previously they wanted to hold on to the technology until 'they' thought that it was ready for market. I was struggling with a name that I could use - now I have it!
Lawrence Poynter - September 13, 2001

Copyright © 1996-2011 Riggs Eckelberry Marketing Group, Inc. All Rights Reserved
TechTransform and the interlocked "T" are trademarks of REMG, Inc.