This article was published on June 19, 2015

How to prioritize feature building on your startup

How to prioritize feature building on your startup
Itai Raveh
Story by

Itai Raveh

Itai is a Product Manager at Yotpo and merman who starts every morning with a surf or a swim. Itai is a Product Manager at Yotpo and merman who starts every morning with a surf or a swim.

The scenario is all too familiar for product managers: there’s a million feature requests and ideas for development and just as many stakeholders with conflicting agendas about which should be prioritized. The solution – objectively determining what’s most important – isn’t always so easy to achieve.

For startups, the wrong call during product development doesn’t just waste valuable company resources, it puts the entire product (and, by default, the company), in jeopardy. There’s a reason they call it a product roadmap: it guides the future of the entire company.

For this reason, prioritization of development makes or breaks a startup’s success. Like many product managers at quickly growing companies, I wanted a magic formula to hack the prioritization process.

I developed a three-step strategy that could not just help our company, but others who are scaling quickly. It may not be magic fairy dust, but it’s an effective method for startup product managers to process feature development based on objective business values and criteria, while taking multiple stakeholders’ needs into concern and keeping the process transparent.

Step 1: Curate great ideas

Every feature begins with an idea. There are many ways to generate ideas for product features, such as analysis of core metrics, customer feedback through usability tests, interviews and more.

In addition to these methods, we are using a feedback portal where the entire Yotpo community – both customers and employees – can see what features we have on our mind and suggest their own.

We use Uservoice, a feedback, helpdesk, and knowledge base management software. It functions as our Feature Request Portal, where anyone can submit their ideas and the entire community can view and vote them up.

Screen Shot 2015-06-18 at 13.01.36

This is enormously helpful to build transparency among our users and create community-powered development.

Additionally, we add CTAs to request features in the Client Success Managers’ email signatures, as we believe that good ideas can come from anyone at any time.

Many times, users have brilliant ideas – but it’s important to understand not just the “what,” but also the “why.” It’s necessary to understand the issue that caused users to request a feature so you can identify weak spots and find more comprehensive solutions.

Often, many different feature requests can all point out one major problem and lead us to develop a much more robust solution.

In the end, the customer’s idea may be really valid and can lead to a future feature development, but it’s important to always start with the core problem rather than taking every solution that is suggested to heart.

When deciding which ideas to add to the roadmap, use an organization method like Trello to store a backlog of all ideas. This can then be used to complement your ranking system. Organizing this backlog according to the themes you will use in your ranking system allows you to easily keep track of them.

All features that appear under the organized backlog will eventually make their way into the ranking system – we use the ranking system to determine the order in which they get sent to sprints.

Trello can be used to track which features are in the oven and which are still waiting for development, while the actual ranking determines the order in which they will be developed.

Step 2: Figure what to develop first

In product development, it’s not just determining what to develop, but figuring out what to develop first. After determining which ideas are valid features, it’s time to begin ranking them.

For feature ranking, we developed a system to prioritize features based on objective business value – choosing the features that will have the biggest business impact. This scoring system takes into consideration many different perspectives, needs, and factors in order to enable the best ideas to climb to the top. (click the image to see it full-size).

Screen Shot 2015-06-18 at 13.02.37

It doesn’t just provide you with an insight as to what you should consider developing, but more importantly, what you shouldn’t. Product strategy means saying no, a lot.

In order to understand the full impact a feature will have on business drivers, we ask key stakeholders, including product management, marketing, and client-facing teams, rank each feature based on the objective value it would generate for them.

Screen Shot 2015-06-18 at 13.03.29

Each stakeholder can click on a link within the document to view the features’ Trello card and story points for more information. The rank is from zero to five and we have a clear scoring system to show what each rank represents.

For example, a rank of five coming from Support might mean that the feature has consistently reoccurring issues, and a feature enhancement would solve the issue and reduce ticket creation.  A rank of five coming from Product may mean that a feature enhancement for a highly used feature is expected to reduce friction, improve a core metric or increase traffic and conversion for our customers’ stores.

The key to ranking features is making it as objective as possible: clear outlines let us really measure the true impact of each feature.

Step 3: Assess Associated Costs

After ranking features, consider the cost associated with each feature, which for most startups would be the cost associated with R&D efforts. Here R&D can use different methods to predict with reasonable accuracy how difficult it will be for them to execute the tasks.

We use the T-shirt sizing method for estimating development efforts. Teams estimate whether they think a feature requires extra-small, small, medium, large, or extra-large effort. We allocate a weight of X2 to the R&D score (cost), which is then subtracted from the sum of all the scores relating to business drivers’ impact in order to arrive at the final score.

Best practice tip: Use a weighting system to accommodate for uncertainty and an ever evolving market.

We found the most effective way to combat changing circumstances was to assign each pillar within the ranking a weight. So, if we decide that revenue is the most important metric for us this quarter, we have the flexibility to increase the weight for the Sales pillar from one to two.

Also, we use sorting to narrow down the results, so if we decide that we want to focus on improving a specific theme/metric, we can sort the features to show only features that are related.

Screen Shot 2015-06-18 at 13.03.58

If you have multiple features in a tier, each can be differentiated based on theme or metric we want to focus on.

Bear in mind that you can always use more complex, binary scoring equations to map the prioritization, but for us, keeping it simple was the most effective way to quickly reach the answers that matter.

We chose to use our own solution as we didn’t want to have our methodology and flows pre-determined by a software (however flexible the software might be). However, there are many softwares that assist product managers and improve the methodologies for building products and prioritizing features. Aha! is a product roadmap software that provides a suite of solutions to product managers and can help scale decision making.

Another great benefit of the ranking document is that each internal stakeholder can run through the features and write commentary notes, ask questions, and point out elements to consider from their point of view when developing a feature. This ensures we communicate with all stakeholders clearly and concisely, which ensures that every aspect is considered before requirements are written, and we determine how to build the feature.

A word of caution – this method should not hinder innovation and ownership of the product by adhering to the ranking, but complement it. A product manager should not be bound by the ranking, the ranking should merely create a clearer a picture. This, together with a solid basis for decision making, can help the product manager see the full picture for curating the right features.

By following these steps, you will be able to cultivate a decision making culture which clearly defines and prioritizes the what by communicating and collaborating with all stakeholders. Feel free to post your comments below and I will do my best to answer each one in turn.

Read Next: How to use data to plan your startup budget

Image credit: Shutterstock

Get the TNW newsletter

Get the most important tech news in your inbox each week.