Long live the roadmap.
When you are working within product management there are a few principles that you need to accept and all of them point in a direction where the typical roadmap is dead.
I am a pretty good product guy, for the last 10+ years I have been doing this product stuff in one way or another, but I still don't know what features are going to be successful. I would estimate that I am wrong about the likelihood of a feature being successful 70% of the time. An average person is maybe wrong 75% of the time. 10+ years of experience have given me a 5% higher precision, it will take me another 140 years to get to a 0% failure rate - I am not going to wait around for that!
Why don't I know if a feature is going to be successful? Well, there are so many factors that come into play when building and designing a feature (or a product), will the customer use it? Does it solve their real problem and need? Can they use it? Can they pay for it? Can we support it? Can we build it? etc.
The only thing we can do is to stop thinking about a feature roadmap and change to a process where we are not planning and committing features, but are talking about the business impact that we want to achieve and then running discovery (basically testing and iterating) to achieve the wanted impact. If we test and validate I would guesstimate that our chances of failure go down to around 40%, this is a lot better then where we were before!
Have you ever tried to estimate a delivery time on a tech project? Have you ever heard the joke that you always need to multiply with π to get a correct time estimation of a developer's tasks?
If you and your team have no chance of guesstimating the delivery date, do you think its fair to make them accountable for a deadline?
If you are working with a roadmap you are not only committing features where you don't have a clue if they are going to work, you are also committing deadlines where you have absolutely no chance of knowing when your team is done.
Would you stand in front of your companies board and commit on features and deadlines knowingly that they are all doomed to fail?
Stop lying to yourself, your team and your clients and start working on a proper process that sadly will not give you the answers that a roadmaps promise you, but will give you a successful product that drives a successful business.
We are a global company with 550+ people, 9 product teams, and not a single roadmap, if we can do it, you can too! Go and save yourself from failure.