Product Roadmaps, in general, are confusing. The truth? Even the most experienced Product Managers still don’t have them fully figured out.
Three years ago I wrote a blog post about Rethinking the Product Roadmap, in which I advocated for a focus on solving customer problems instead of listing out features with deadlines. I received lots of feedback on the post, much of it praise, but also some criticism. The biggest takeaway comment for me was, “This is great for new products in the Discovery Phase, but what about accounting for products that we have already validated and need to build?”
In the time since, I’ve been working with clients to find what works best in real life situations. I want to share with you the framework that I have found, which allows for a balance of discovery and delivery and takes us out of that laundry list of features. But first, let’s talk about Roadmaps in general, and how they got us on the wrong track to begin with.
Think back, before Google Maps, to what a real roadmap was- those things with tiny, tiny city names and thousands of squiggly lines. The ones that made it so you could never drive alone because you needed a copilot, someone to wrangle the giant map and dictate to you how to navigate from point A to point B. Yeah THOSE things.
Well that’s where the term Roadmap actually comes from. When using a roadmap, you always knew where you were and where you wanted to go, and you had all these different routes to choose from. Some routes were shorter, some were longer but more scenic. It was ultimately up to you to choose your optimal path. Product Roadmaps should work just like their namesake, but their intended purpose was lost somewhere along the way. Now, Roadmaps function more like Gantt Charts.
Gantt Charts are a project management tool used to keep track of timelines and due dates. While Gantt Charts are fine for large manufacturing projects, they do not work well for most software situations because of the uncertainty around development.
When we’re first deciding what software product we should be exploring, we don’t know what our users want, and thus we are not exactly sure what the finished product will look like. This makes it nearly impossible to estimate how long the product will take to build. Yet, we still use this Gantt Chart model to plan our work, which locks us into delivering solutions by certain dates. Teams get frustrated that there is no perceived flexibility in these plans, and management gets mad when teams don’t deliver on time. The result is a bunch of feature releases that don’t make a difference for our users.
Instead, we need a model that helps us take into account uncertainty and allows us the flexibility for both discovery and delivery. To do this, we focus on Product Roadmaps that are made up of outcomes, a theme, and hypotheses, and we leave out any unvalidated solutions or features.
Let’s look at an example I created of a completed Product Roadmap for one team. I’m using the New York Health Exchange’s website as an example of a product as I can easily pin down things that need to be fixed- like, lots of things. I’ve never worked with them, and I don’t claim to have knowledge about how they function internally, but I am a customer of the website. This is how I would approach improving the product if I were working with them.
The top section of our Roadmap outlines our Vision, Challenge, and Target Condition, which reflect the goals outlined in my post on Product Strategy. A Product Roadmap should be aligned around these same goals and help communicate the tactics to reach them.
According to this Roadmap, Team X is working towards the overarching goal (or Challenge) of getting 99% of people who create an account on NYS of Health to sign up for insurance.
Now that we have our Challenge set, the team analyzes the customer problems standing in the way of them reaching that goal. This is the same process we outlined in the original Roadmap post. Let’s say that after user research, we find that the biggest problem for NYS of Health’s website users is that it takes a very long time to search for and find an ideal health insurance plan. It’s a confusing process and incredibly frustrating. It causes people to give up instead of sign up.
So Team X decides to tackle this problem, and sets a Target Condition that will tell them when they have reached a successful outcome. “Decrease the average time to find and purchase health insurance from 3 weeks to 2 days in Q2.” While users can only sign up once a year for health insurance, there are plenty of changes that could be implemented to make it easier once that enrollment period opens.
From there, Team X would figure out the top areas they want to explore or deliver on in Q2 to reach that Target Condition. We call these Themes. For example, the “Search Enhancements” Theme will include all the UX work needed to simplify the search for health insurance options.
In order to answer the question, “Why are we exploring this theme?” we write a Hypothesis or Problem Statement. When products are still in the Discovery Phase, we are still validating the problem and solution, so we want to state what we know and what we don’t know. Our Hypothesis should support that we believe our solution will solve the biggest problem.
So with our Search Enhancements Theme we are in Delivery Phase, and our Hypothesis is “We believe that by enhancing the filter options to reflect the top deciding factors that shoppers make when choosing a health insurance, we will make it easier for them to narrow down their choices and find the right health insurance.” We have already validated that people have problems narrowing down their choices to find the right health insurance, and we think that offering some enhancements to the filter options can solve that problem.
Our last theme, Subsidy Communication, is an example of a problem that is still very early in Discovery Phase. We have identified that there is a very low conversion rate for people who qualify for Subsidies, and we want to investigate why that is. We have a few ideas but we have not yet validated them. So we need to clearly communicate our Hypothesis, highlighting what we know (“the people who would get Health Insurance for almost free are converting at a very low rate”) and what we don’t know (“we are not sure if people leave before they know if they whether or not they qualify for subsidies, or if they just don’t understand our explanation of subsidies”).
After the Hypothesis section comes the “Outcomes” – the most important part of our Roadmap. Here, we state what we hope to achieve by solving this problem or proving this Hypothesis. Each Outcome should relate back to the Target Condition in some way. All of our work is going to help us reach the Target Condition, so it would be a pretty boring Roadmap if every Outcome for each theme is just a repetition of that goal. Instead, think about how each of those changes or improvements signals a move towards your goal.
For example, with the Search Enhancements, we can tell if we are making it easier for people to search by measuring the qualitative components of frustration and the quantitative components of decreasing the time to identify an option. You want to make sure you are focusing on a good combination of both business and user outcomes in these sections, and make them measurable.
The last two sections deal with status and priority. The Status should state where you are in the Discovery versus Delivery phase. If you are in the Discovery Phase, you’re delivering to learn, so your focus is on experimentation over releasing features. In the Delivery phase, you’ll be delivering valuable features and solutions to users. Sit down with your team and define what this means within your company, and then share these definitions with the rest of the company so there is mutual understanding.
Notice here that our Roadmap does not include any specifics on how we plan on tackling these problems. This is because we are still experimenting with options- we haven’t laid out a plan to implement a set of features or solution components.
The above example is only one team’s Roadmap. There might be many other teams working towards this same Challenge, but they would all have different Target Conditions. These Roadmaps could then be combined into a Portfolio Roadmap that could be communicated across different levels of the business, like so:
It’s important to remember that a Roadmap is a communication tool first and foremost. It’s meant to align teams across time horizons and goals. Management should be goal and Vision setting with a timeline of 6 months to a few years, while teams should be looking at a smaller timeline, from quarters up to 6 months. There should be an accompanying strategy document that addresses how these goals help the company reach its Vision, but the Roadmap itself does not serve that purpose.
Roadmaps are a living document. If you invalidate a theme, update it. If you’ve changed a goal, update it. There is no reason you should be struggling to work around a Roadmap that does not work for you. When you’re communicating with customers, focus on how you are going to solve their problems, not how you are going to deliver their features. Your customers care first and foremost about the value they are receiving.
We teach Product Roadmaps in our 10 week online course at
The next class will start in March 2017.
Sign up for our mailing list on the top, right hand side of this page to be notified when we launch our next class.