The ingredients for a great product roadmapPosted: January 8, 2014
Right now a lot of our companies are going through their product roadmaps for 2014 and it’s my favourite discussion to have at a company. I wish all board and team meetings were product meetings. I am probably not going to be very helpful in actually making product decisions but hopefully on how to make good decisions and by transferring a few best practices on the way.
Coming up with a good product roadmap is usually a combination of art and science. The best companies will have the right gut feeling and be very close to their users; yet use plenty of data and a process to cross-check their thinking.
Bad product roadmaps are usually an outcome of a pure “negotiation” process between the product, engineering, marketing and business teams of a company, with little real underlying structure to tie it all together.
I was recently at a product roadmap workshop at one of our (consumer) companies where we applied the following, really simple, framework to help us get the right perspective on what we should be building and in what sequence:
1) Step back, forget what you think your roadmap should be and get all feature requests / product ideas out on to the table (or better on to the whiteboard)
Most likely you will have already have a product roadmap that is updated / kept agile constantly (how it should be done). Don’t bring it to that meeting. Step back, forget about what you thought you should be building next months and just make a long list of feature requests, product ideas – and especially don’t forget to include the crazy ones, dig-out some old gems, etc.
2) Sort features by degree of impact it has on user and company goals
It’s very easy for teams to spend a lot of time on maintenance, small fixes, “we always wanted to do this”, etc. And these things are important. But it can really distract you from the big things you should be tackling – so it’s key to really narrow down product roadmap to things that will have a major impact.
But how do you measure impact? In this case we decided to measure impact by how features would enhance user goals (we chose 2) and company goals (we also chose 2):
User goals: i) wants better utility and /or ii) increase his social status / have fun
Company goals: i) user growth and / or ii) user engagement
So we basically then mapped the features on to the 2 charts below. This looks easy but of course it can be hours of discussion and moving stuff around:
Map 1: Features that have a high impact on user experience
Map 2: Features that have a high impact on company goals
We then took all features that made it into the top right corner of either chart (some features could be on both of course) and some that were just very high in a category on to our next list; we threw the rest out.
So now you should have a much leaner list of features that will have a high impact on user experience, company goals or – ideally – both.
3) Priorities by effort vs. impact
Of course you want to build low effort & high impact features first – and it’s great to have a detailed discussion with the key folks around which features fall into that category. Things can be easier or harder than they look.
>> The best discussions are how to make hard things easy. This is where a lot of energy and time should go to. <<
So we mapped product features again by their relative impact and relative effort and roughly (with a few exceptions) agreed we would build them in the sequence 1-4:
Map 3: Effort vs. Impact
4) Users first
What happens when features are pretty equal on the effort vs. impact scale? In nearly all cases the answer will be to prioritize whatever delights users the most.
There are probably dozens of ways to do this and there will always be a reason why a low-impact & high effort feature may need to be prioritized because it’s an important enabler for other features etc.
My only advice would be to put yourself through a process and structure that works for you. At the end of it you may be surprised how much high impact & low effort things you have found.