Great New Product Ideas – DON’T Ask the Users What They Want!

Everybody knows that successful companies are market driven rather than technology driven and everybody knows that a market driven company asks its customers what they want rather than asking its engineers what they want to build.  Everybody who knows all that is wrong.  The very best ideas, the ones that spawn industries and spread creative destruction, come from smart people realizing that a new use of a new technology can create a compelling new capability.

Marc Andreesen’s invention of the modern Internet browser is one example.

Tim Berner-Lee’s invention of the WorldWideWeb without which there would have been nothing to browse is another example.

But my favorite example is VisiCalc – in its time it was by far the most important computer program ever written.

VisiCalc was the first computer spreadsheet in case you are too young to remember.  Before VisiCalc, the few personal computers in businesses were used as replacements for dedicated word processing machines which, themselves, had just recently begun to replace typewriters and WhiteOut.

VisiCalc was invented and programmed by Dan Bricklin and Bob Frankston – two smart guys I’ve been lucky enough to run into from time to time.  They could’ve spent a hundred years asking users what they wanted their personal computers to do and no one would ever have said “I want a program which works like a big sheet of accounting paper but, when I make a change in one place, the change propagates through all the rows and columns.”  The people who were using personal computers for typewriters and playing what now seem very primitive games just wouldn’t have thought of that.

But Dan Bricklin thought what a pain it was in business planning to have to redo all those pages full of calculations for what-if scenarios. Of course, many what-ifs never got asked because it was too much work to calculate them.   The way I remember the story, he thought about it because he liked to do lots of what-ifs and he didn’t like to do lots of columns of numbers. His first day dreams about this, he says, were during business courses at Harvard B School: he visualized a mouse built into his calculator and a heads-up display like a fighter jet as he navigated through scenarios.  The result was a beautifully simple concept – a simulated spreadsheet with formulae and defined dependencies between cells.  Now, of course, when we say “spreadsheet” we mean the version on the computer and we’ve forgotten all about the huge sheets of paper with all the lines on them that were the original spreadsheet.

VisiCalc changed everything.  Of course, mainframes and minis had been used to crunch numbers; but, before you could crunch any numbers, you had to write a program.  Programs were written by MIS and there was always a backlog so it took forever to actually have a program available to do something.  Given the lead time, computers were rarely useful for one-time calculations, certainly not for churning out variations on a business plan.

In 1981 at about the time VisiCalc was invented, I was on a two year sabbatical from Solutions, my mainframe programming business, and was Secretary of Transportation for the State of Vermont.  We had a big problem: our roads were falling apart!  The Arab oil embargo and its aftermath in the 70s had driven up the price of oil causing a quadrupling of the price of asphalt.  Higher gasoline prices meant that people were driving less and buying more efficient cars so gas tax revenues were down.  Maintenance had been deferred for years.

If you’ve never been a road commissioner, you may not know that the structural purpose of the asphalt on top of the road is to be a roof over the foundation of the road which goes down several feet.  If water gets through – especially if it freezes as it is wont to do in Vermont, the expensive foundation will eventually be destroyed and now you have a reconstruction rather than a repaving project on your hands.  So, with deferred maintenance, we had not only politically unpopular potholes but also roads that needed complete rebuilding.

We had to go to the legislature and ask for a lot of money.  And, because it was clear that we could only get so much money per year, we needed to triage so that the money was aimed mainly at those roads which, if left unattended for one more year, would turn from repaving into rebuilding projects.  The ones that were already gone could be patched until we got to them.  The ones that could last another year should wait another year.

We needed to show the legislators, district-by-district, what we wanted to do.  We asked the State Computer Division to write us a program.  They said fine, it would take two years.  The legislature was meeting in two weeks!

VisiCalc to the rescue.  Who needs MIS?  We used our typewriter budget to buy a TRS-80 Model II at Radio Shack and a tractor-feed printer for green and white striped paper and soon we had more numbers than anyone could possibly read.  When a legislator asked what if we did this road or that road in his or her district instead of something else, we had an answer. We got our money.  The roads really did get fixed over the next five years.

My experience wasn’t unique.  That was the beginning of the end of the tyranny of MIS and their jealously guarded mainframes.  Departments could crunch their own numbers, make their own plans.  VisiCalc meant that computers could be used for new applications without programming.  The era of the personal computer in business had truly arrived.

Does all this mean that every over-engineered geegaw at the Consumer Electronics Show is going to be a hit?  No, of course not. Does it mean that you should ignore user input and let the engineers build what they want?  No, user input is crucial to tell you what you left out of version 1.0 so that you can put it into 1.1.  If you do “me-too” products, user input can tell you what the category leader forgot.  VisiCalc is, sadly, also a story of a product eventually overtaken by Lotus 1-2-3 with integrated graphics and, eventually, the story of the dominance of Excel.

The point of this story is that no survey or focus group will ever tell you what the next great thing is going to be.  That kind of idea, that kind of product, comes from visionaries who understand a new technology well enough to dream up an unintended use and who are stubborn and skillful enough to implement what nobody even knew to want. 

These are the products and services that change our lives.

More on product ideas is in a later post.