I hate when I have to admit that someone else was right, especially when I held such a strong position for so long.
For millennia (internet millennia, anyway, or at least since the dawn of software made for sale) there’s been a strong divide in many companies between sales/marketing and engineering. Marketing would seemingly promise things that they thought would be “cool” without knowing what it would take to make it, and engineering would stubbornly resist these whimsical changes to the precious algorithms.
See Dilbert for a zillion examples.
Having been on both sides of that battle now (often with the arguments being with myself as both packager and implementor) I see it a bit differently now.
There isn’t a single product or service that I’ve designed this year that wouldn’t have been better if I’d started with the marketing documentation first. Think of it as marketing-driven design Thankfully, most of my work for hire projects start this way via some kind of proposal, so I’ve managed to deliver value and excellence to my clients, but my internally developed projects, that people seem curiously resistant to purchasing? Not so good.
Here’s the crux of it: describing what is simply isn’t going to be as compelling as describing what could be. If a project has been completely built before the description is even started (via a gross misapplication of some extreme programming doctrine,) the marketing is going to be limited by what’s there. The best you can manage is to talk about the promise of the next version, to which people will generally respond by waiting. Marketing-driven design has not such constraint.
If you can describe what will be, and not commit any development effort beyond “is this even possible” prototyping, until you’ve got something that actually gets people excited, your chances are much better.
I’d write more on the topic, but I’ve got to go refactor a system to meet the “what I should have written in the first place” requirements that are based on solid marketing-driven design principles.
Image based on one by Sean McEntee