My background as a product manager has been with startups: One was small, looking for a product/market fit; and the other was scaling quickly. Like every job, they both had their challenges and I've been looking for some sense on what worked well and what didn't. Many of the lessons in 37Signals' book "Getting Real" spoke to my experience. I had been through the waterfall-to-agile conversion and the light on documentation mantra certainly makes some sense. Which brings me to this post: 37Signals just posted a note in the Signals vs Noise blog: Product roadmaps are dangerous. I disagree.
Their post has the light but oh-so-convincing tone of Getting Real. Hell it's only 332 words! The main points:
- Roadmaps let the past drive the future.
- You should only look out a few weeks into the future.
- Roadmaps create expectations.
I get it - they are in the web applications business: build fast, release often, measure and learn; Repeat... But some of the points seem plain wrong. Here are my objections:
- Yes, focussing on now is great, but trying to predict based on past patterns is important, as well.
- Once you grow past the everyone-does-everything stage of a startup, organizational behaviour begins to take hold. Stakeholders will always try to protect their interests: Sales wants the next feature that will reel in their current prospect. Support wants the highest visibility bugs fixed. The CEO is after the next big thing. A product manager who doesn't lay out out a path that extends beyond the next few months, will simply be reacting to these forces.
- Dealing with expectations is part of the management game. Everyone wants a date: Operations needs time to roll out infrastructure or train customer support for that great new feature you want to release. Acquiring customer in a consumer web business is different than the SMB or enterprise world. There may be limited opportunities to interact with prospects. There is a reason that marketing is fixated about that feature that needs to be ready for the the big conference: everyone who buys will be there! Product managers owe these constituents more than: "I don't do roadmaps, they are out of date as soon as you publish them"
- A Product Manager is not just a feature-monger throwing customer requests onto the roadmap willy-nilly. I won't use that hoary Henry Ford quote, but you won't get quality customer feedback if you ask them only what feature they want in the next couple of sprints. You will get a minor delta from the current product iteration. Show them you have a vision that extends beyond the here and now.
I think a better analogy for product planning document comes from the boating world: the chartplotter: it shows you the ports-of-call (customers) but also the rocks (market barriers) to avoid. Most importantly, it automatically updates with your progress through the environment. I could take this analogy further, but I think you get the idea. Expressing a vision for your product and the intermediate steps to get there isn't dangerous - it just needs to be done in a living document so you don't get lost at sea!