rss feed

blog - making technology instead of products

posted by witten on May 20, 2007

I once worked at a software startup where it was announced that the engineering team was to have a meeting across the street at the local bar. This was a little out of the ordinary, especially since we were only given a vague idea about the meeting's topic: Future directions after the current release is shipped. So at around 4 in the afternoon, a motley group of developers, testers, and managers filed into a nearly empty Cuban-themed bar with congo drums for bar stools. We sat uncomfortably around several wooden tables, sipping sugary mixed drinks and waiting for whatever announcement was coming.

The announcement, as delivered by one of our own venerable managers, was as follows. The company has awesome technology that's worth millions of dollars. That was assumed as a given. However, if some employee comes up with a killer app to showcase this technology, then the killer app would have a multiplier effect, making the millions of dollars of company value into tens of millions of dollars. The implication being that since we all have some stock options, coming up with a killer app will make us even richer when the eventual and inevitable huge payout happens. Names of various inspirational killer apps were bandied about, such as MySpace and dodgeball.com. It was left as an exercise to the listener as to what the hell these existing web sites had to do with our company's technology.

We squinted at each other in the afternoon sun, wondering why this announcement merited taking us all away from coding for two hours. Then, a challenge was put forth. The first person to make a killer app will be paid one thousand dollars! This proclamation was made with neither any hint of irony nor a Dr.-Evil-style pinky-to-mouth gesture. By busting our butts and producing a killer app to make the company's technology into something actually useful, and thereby lining the pockets of the owners and investors with tens of millions of dollars in value, one of us would be paid a whole thousand dollars for our trouble! And our infinitesimally small stock option grants would be worth slightly more. It sounded like quite a deal.

As the inspirational ramblings by various managers wound down and the meeting degenerated into general chatting and mojito-sipping, some of us wondered aloud what the hell was going on. And it soon occurred to us: Desperation. The management realized that all this time the company had been producing cool technology rather than products that actually meet customer needs. And while developing software simply because the technology it's built on happens to be cool is a noble goal, it also has the unique property of not producing any measurable revenue. In other words, being technologically cool counts for something, but if your goal involves producing software that people will voluntarily exchange for money, then it's more important to make something that customers actually want.

Management realized this, albeit a bit late in the game, and so they were desperate for something, anything that would convert their multi-year investment into an actual product that customers would pay good money for. But rather than revamp our product suite or cut out "cool" features that no customers actually wanted, they settled on the quick fix idea of a killer app. If only our developers invent something like MySpace running on top of our awesome software, our otherwise useless technology will become an actual product, and we'll be rich! And we can reward one of them with a thousand dollars as a motivational tool!

The company made the classic mistake of focusing on technology rather than products. Unless you're making products like high-end audio equipment for people with too much discretionary income, customers don't care about whiz-bang technology. They care about whether the product solves a problem and makes their life easier. Whiz-bang technology might make a product that's already useful even more appealing, but technology on its own is rarely sufficient.

Here's a simple one step process detailing how to make a product that's focused around the customer's needs rather than your need to make cool technology. Step one: Actually focus on the customer's needs! When coming up with a new feature, first step back and figure out which customer problem it solves. What's that? It doesn't solve, address, or have anything to do with an actual customer problem? Don't implement it!

Okay, this advice of listening to your customers might sound trite, and the quite legitimate objection can be raised that sometimes customers don't know what they want. This is certainly a concern, but that doesn't excuse foisting upon your customers software that doesn't meet any of their needs. Even if a particular customer doesn't know what he or she wants, it is your job as the developer and all-around problem solver to guess whether a feature you're developing has a non-zero chance of being remotely useful to that customer. If in your esteemed opinion the feature is technologically cool but not terribly useful, well, then you have your answer.

Why do so many companies fall into this trap of making technology rather than products? Simply put, because it's easy. Actually focusing on the customer and gathering requirements and distilling them down into actual features is just plain hard, especially for a group of logically minded coders. Developers by their nature deal in technology, and it takes another set of skills that many developers don't have to think in terms of what would satisfy their users. So many software companies take the easy route of focusing exclusively on the technology because it's what they know. Doing otherwise would not only require stepping outside of their comfort area but would require much more up-front effort.

One of the interesting claims in the Mythical Man-Month is that turning a piece of software into an actual product takes three times the development effort of just producing the software to begin with. It is this sort of transition from a simple technology-oriented program to a shippable, customer-oriented "solution" that many companies neglect. A company might get their software to market faster by focusing on the technology rather than the customer, but the software they do ship won't make any customers happy.

Fortunately, these problems have a way of correcting themselves. At the software startup I worked for with the "cool" technology, no one ever came up with the killer app and so no one claimed the thousand bucks of prize money. The company ended up missing payroll, nearly went out of business, and was eventually sold off at fire sale prices to a holding company. No one got rich, and all the stock options that were to be worth millions of dollars turned out to be more useful when printed out and used as toilet paper.

4 comments

Write a comment!