Tag Archives: feature prioritization

Roadmaps Are Not A Popularity Contest

By Bruce McCarthy

I write a blog called User>Driven and my Twitter handle is @d8a_driven, so you would think I like the idea of setting roadmap priorities based primarily on customer requests. You would be wrong.

Dell Idea Storm

Back in 2007, Dell created an online customer forum where people could post “innovative” ideas for how Dell could improve their offering and others could vote on those ideas. I wrote at the time that I thought it was a very bad idea.

In my opinion, Dell had fundamentally misunderstood the uses of customer data and the best ways to collect it. The proof of that was that the voting had been hijacked by a small number of very vocal open source advocates. Checking back in this week, I see it mostly consists of >complaints and suggestions for minor tweaks to design and packaging. Interestingly, they have made it impossible to see which ideas have the most votes.

Tracking Customer Requests

Over 20 years as a product person, I’ve developed an objective scoring methodology for prioritizing ideas. (See my presentation on Slideshare.) It focuses on ranking proposed improvements or changes as to how much they contribute to strategic goals compared to their implementation cost.

So am I arguing for ignoring customer input? Not at all. If improving your customer retention rate is one of your strategic goals, it might make sense to include number of customer requests in your scoring. I know more than one product person who has tracked popularity as input to their planning and tools like UserVoice make this easier than ever.

Your Customers Can Hold You Back

Tracking customer requests can easily create a false sense of security for a product person, though. You’re collecting real, live market data, right? It’s all neatly tallied and summarized in your PowerPoint, right? Ok, fine, but bear these very real limitations in mind.

  • As Dell Idea Storm showed, you have to recognize that not all votes are equal. Vocal customers are not necessarily paying customers. And even if you could get evenly-distributed input, most customers are small. Should you be focused on their needs or on the needs of your largest customers? Tailor your data collection to your strategic goals and weight the input accordingly.
  • Expect incremental input only. Henry Ford famously said that if he’d asked customers what they wanted, they would have answered: “a faster horse.” Customers will give you ‘better, faster, cheaper’ advice, but are unlikely to propose a break-through idea that uses technology to solve problems in new ways.
  • Beware “featuritis.” We all complain about the lengthy list of features in Microsoft products that we never use but make things complex, crash-prone and slow. Where does all of this bloat come from? Customer requests. Beware of adding features that only a few will use but many will have to put up with.
  • Existing customers tell you nothing about non-buyers. People who are in your target market but don’t respond to your messaging or solution are your best source for information on how to expand your appeal. Reach out to people who evaluated your solution but decided to go with the competition or do nothing (this is called “win/loss analysis”) and ask them about their business, their problems, and why they went a different direction.
  • Existing customers won’t help you penetrate new markets. If you want to grow beyond your current market niche into other verticals or up to larger customers, engage with those people directly and, rather than talk about your solution, ask them about their needs and wants. Develop an understanding of the personas in that space and what they need and you will be on your way to designing a solution that will capture that market.

One Way To Use Popularity To Prioritize

Ok, yes, I did create a forum where people could vote on what a roadmapping tool for product people should do when setting out to design Reqqs, my forthcoming tool for product people. But that was different. Really. Let me tell you why.

Before putting up the forum, I spent a lot of time interviewing other product people, asking them about their jobs, what were they trying to achieve, and what were their main obstacles and frustrations. I did this until it was pretty easy to guess what the next person would say. That kind of qualitative data helped me define the essence of the product as a roadmapping tool for product people.

The next step was to figure out which of the 10 or so problems I had uncovered were the most common and most painful. UserVoice’s voting feature is perfect for that. I set up the forum, pre-populated it with my 10 things, set up each user with 10 votes to allocate however they liked, and began promoting it on product discussion boards like those on LinkedIn. This limited my audience to qualified prospects, all of whom were equal in my eyes since none were yet a customer.

With the hard work of using interview data to develop innovative ideas done, a survey within my target market provided the quantitative data needed to determine the top few of those ideas that would form the MVP. As it happens, potential customers highlighted prioritization as the number one thing they wanted help with, and that will be the core feature of the initial release of Reqqs.

Customer Requests Are But One Input

So, yes, customer input can be a powerful tool to aid in decision-making. But like all superpowers, this one must be used responsibly. Set your strategic goals first, then decide which items from your utility belt will help you toward those goals.


Bruce McCarthy is a serial entrepreneur, 20-year product person, and sought-after speaker on roadmapping and prioritization. By day, he is VP of Product for NetProspex, but only you know his secret identity as Chief Product Person for Reqqs, the smart roadmap tool for product people (forthcoming). He is available for advice on product management topics of all kinds.