It's obviously true that vendors often fail to deliver what customer's want. It also true that vendors often fail to deliver what customer's need.
It's less obvious that "need" and "want" are sometimes only loosely related, and that, on occasion, wants can be opposed to needs. Ok, so that's obvious for you and I, but it's also true of businesses.
Sometimes vendors, by virtue of hard experience, know that what a customer wants will hurt them. But what do you do when they really insist?
Joel Spolsky, a justly famed software entrepreneur, gives us an example ...
Joel on SoftwareIn my own vastly smaller way, I used to lean towards Joel's former position. Don't indulge the customer's self-destructive tendencies. It might boost sales, but in the medium-run the customer will be unhappy. That will translate to lesser sales.
Mysteriously, about a week ago, Dan, the program manager designing most of the new features in FogBugz 7, came to ask me what features I thought should go in the timesheet reporting plug-in...
... We have a theory, here, that this is a bad idea. Using timesheets as a performance metric can lead to only one thing: bad data in timesheets.
The first time your boss comes into your office and gives you grief because it looks like you only did 7 hours of work yesterday, you’re going to make sure that never happens again. And then, suddenly, behold, the timesheets show everyone working 12 hour days, and all the data in the timesheets becomes instantly bogus. And EBS, our statistical technique for predicting ship dates, suddenly stops working, because you’re feeding it data that is meant to get your boss to stop bugging you, not accurate data.
Now, this theory may be completely off the wall, but it is our theory, and until we hear something better, that’s the one we’re going with.
So our policy has been that if you want to get the timesheet data, well, yes, you can, we’ll give you a way to get it in CSV format or XML format or something and then you can abuse it all you want... go ahead, hang yourself, but we’re sure as heck not going to make it easy for you with a pretty report all tied up with a bow that you might click on by accident, as you browse around, because thou shalt not put a stumbling block before the blind...
... Apparently nice people email us and ask for that exact feature and offer to give us little green rectangular things that can be exchanged for other goods and services if we do the feature...
Nowadays I lean towards trying hard to understand if there's an underlying problem that could be solved another way. If all else fails I'll accede to the customer's request, while trying to make their learning period as short as possible -- and to provide an alternative better path when they're ready.
It's like raising kids. Sometimes you just have to ride the train over the cliff.
Besides, things can be more complicated than they seem. Bogus numbers may seem worthless, but what if the project is capitalized? Maybe errors in ship dates will harm one part of the business, but better capitalization numbers will help another part of the business.
It's a messy world. It's nice to know even Spolsky struggles at least sometimes.