Feature Prioritization: 7 Ways to Prioritize Features and Product Improvements
Few companies fail because of a lack of good ideas. But choosing the wrong idea? That can sink your ship faster than an iceberg on a dark night. In fact, according to CB Insights, the #1 reason startups fail is because they built the wrong thing and there’s no market need for it.
But it’s not like you can sit back and not choose a feature or product to build. Choosing the right features to build means prioritizing from your long list of good ideas for what’s most important, realistic, and urgent now. It’s no small feat. And most project managers will agree that the most difficult part of their job is deciding which features deserve the team’s limited time, resources, money, and energy.
In this guide, we’re going to run you through all the many factors that get in the way of feature prioritization, and then round up some of the best strategies for how to actually prioritize features in the best way possible.
Before we dive in, here's how this article is structured:
Feature prioritization starts with a shared vision and purpose
One of the hardest parts about prioritizing features is that they aren’t just product decisions. They’re personal ones. Every single feature, angle, approach, and idea reflects someone’s hard work and opinion. And this only gets even more complicated when you’re dealing with stakeholders with different levels of investment and control over a project.
It might not be that difficult to put off Jeff the UX designer’s idea for a different onboarding flow. But how can you de-prioritize Sally the CEO’s suggestion that you move to a different framework?
However, to be successful, prioritization can’t be personal. You’re not picking someone’s idea over someone else’s idea. You’re choosing the right feature for your company’s strategy and goals.
As Richard Banfield, author of “Product Leadership: How Top Product Managers Launch Awesome Products and Build Successful Teams”, writes:
If the team doesn’t agree on the big picture, then they certainly won’t agree on a single feature.
The big picture—strategy and company goals—needs to be clear before you can debate the merits of each feature. Otherwise it’s like arguing whether to take a car or a boat when you don’t even know where you’re travelling to.
However, while feature prioritization starts with a shared vision, that’s not where it ends.
Stalemates and decision deadlock happen even if you’re all agreed on a shared vision. When you’re prioritizing features, you need to act as a collaborative leader—a role the Harvard Business Review describes as “[leaders with] the capacity to engage people and groups outside one’s formal control and inspire them to work toward common goals — despite difference in convictions, cultural values and operating norms.”
As necessary and just plain right it is to treat everyone on your team as equals, they can’t have equal say in what features get prioritized. Because how can they? Yes, you’re working with good, smart people. But not everyone has the context needed to make major product decisions.
As Mina Radhakrishnan, the first Head of Product at Uber says:
A big part of product leadership is thinking about why are we doing this-and-that to set the basis for saying no, we shouldn’t do that.
A few other ways not to prioritize features
Alright. So we already know that prioritization requires a shared company vision. And we also know that it requires someone to make tough choices and stop the deadlock of “designing by committee.” But before we can get into the practical methods for feature prioritization, there are a few more red flags to look out for:
- Gut Reactions: We’ve all had people push their agenda because of some recent personal experience. Maybe your CEO logged in and saw something they didn’t like or got negative feedback from an investor. While it’s easy to jump on those as obvious priorities, those experiences should always be backed up by data or user research before given resources and time.
- Sales & Support requests (aka the loudest people at the table): When the people on the frontlines of your company are hearing complaints, they’ll tell you. Again, while this can be a powerful way to find features to prioritize, you want to make choices based on trends. Not individual requests. Do the due diligence to see if this is your ideal customer and that it’s a valid request before you let them interrupt your roadmap.
- Isolated feature ROI: It’s hard to say no to more revenue. Yet not every short-term, money-making feature is good for your company in the long term. Income doesn’t always equal a better user experience, and in the long run, happier customers are what will bring you the most success.
7 practical ways to prioritize features
With all the fear mongering out of the way, let’s get back into the fun part of feature prioritization.
Because while this is a big decision, it’s also an exciting one. You’re choosing the future path of your product and helping pick features you know your customers will love and that will help the company grow.
Let’s start with a quick recap: We know we need data and trends to back up our beliefs. We know we need to tie our priorities to larger company strategies and avoid personal bias. And we know we need to not fall victim to short-term thinking.
But in most cases, this will still leave you with a ton of great features to pick from. So how do you know which ones to focus on now?
Luckily, there are some incredibly smart and easy strategies for diving deep into product decisions and help you prioritize features, improvements, and ideas.
1. Place features into themes to avoid choice paralysis
Before you dive right into prioritizing individual features, you need to break them up into smaller groups. Choice paralysis is a real issue when it comes to picking projects to work on. And one of the easiest ways to avoid it is to use feature “themes”.
Themes are groups of features that align with a company goal, product vision, or overall strategy. They help you make sure you’re working on the types of features that are most important right now, while also avoiding the issues of too many options. (Remember, saying no to a feature or update doesn’t mean abandoning it forever. You’re simply choosing where to place your efforts right now.)
There are a few ways you can approach themes for your features:
- Product Roadmap themes: Your product roadmap probably is already broken up into high-level themes like “Reporting”, “Integrations”, “Communication”, “Workflow”, etc… One of the simplest ways to break up new features is to group them by these established themes. This way, you know they’re working towards specific parts of your product and your strategy.
- Metric movers, customer requests, and delight: Another option is to group features by where they came from and their potential impact. To do this, Greylock partner Adam Nash suggests using three specific buckets: Metric Movers for features tied to specific business needs; Customer requests for features your users actively ask for; and Customer delights for features that aren’t being asked for, but that you think your users would love. Features can fall into multiple buckets, but a healthy roadmap prioritizes options from all three.
- Specific, metric-tied themes: Lastly, if you’re super clear on what metrics need to be moved, it can be good to create themes that are specifically tied to them. This could mean categories like: “Reducing churn by increasing engagement” or “Increasing sign-up to purchase conversions”. This way, your priorities are tied to specific needs and you know what success will look like.
Whichever method you choose, being able to look at higher-level feature categories is a great first step. (Decision trees are also a popular decision making and prioritization framework you might want to try here.) Next, you need to prioritize the themes themselves, and then finally, the features within them.
2. Break down product features by feasibility, desirability, and viability
If personal biases and recency can lead us astray, then one of the first things you want to do is to look at your features through a more objective lens. Namely, that means looking at each one based on a few criteria and talking to specific members of your team:
- Feasibility: How technically possible is the feature given the resources and tools you currently have? Talk to your technical team members—back-end engineers, UI designers, and front-end developers—to understand what can be done (vs. what’s impossible or highly improbable).
- Desirability: Do your customers actually want it? Use every available tool to understand whether this is something your users desire. That means talking to researchers, UX designers, marketers, and support, as well as going through any users tests and validation you may have already completed.
- Viability: How does this feature relate to or support your overall strategy and the requirements of the market? Talk to relevant executives and other product managers to understand how this feature works in a bigger ecosystem—both your own (other features, strategies, and goals) and the industry as a whole (regulations, legal issues, financials).
While these criteria come from individuals’ opinions, cross-examining them through multiple lenses helps keep everything objective. And of course, bringing in any supporting or complementary data can keep you extra honest as your go through this exercise.
3. Score options on an Effort/Impact scale
With your features more or less mapped out and validated, it’s time to look at which ones are most important to work on first. A basic and commonly used way to do this is to plot them on a simple Effort/Impact matrix.
This is just a 2x2 grid, where each square represents a different level of effort to build the feature and the potential impact it will have:
The goal is to find the features that will have the highest impact with the lowest effort. However, it’s not always easy to know where a feature fits on the matrix.
Design agency AJ&Smart suggests doing this as a team exercise. Write down each feature idea on a sticky note and then draw your matrix on a whiteboard. Gather a diverse group of teammates and then one at a time, take each sticky note, explain it, and let the team vote on how much effort it will take and then the potential impact from it.
You’ll still have final say on what features get prioritized. But this exercise helps you quickly gather input from a diverse group of people on your team.
4. Go deeper with the RICE Method
Sometimes features are complicated and need to be prioritized with more detail than a simple grid can do. In this case, the RICE method is a great way to score priorities. As Sean McBride, product manager at Intercom, explains:
Systems designed to balance costs and benefits abound. But you can have a hard time finding one that allows you to usefully compare different ideas in a consistent way.
In response, Sean and his team defined four common factors to evaluate each feature when deciding which to prioritize:
- Reach: How many people will this feature affect in a given period? Reach is measured using real product metrics like “customers per quarter” or “transactions per month” to help avoid the bias of picking products or features that you personally want to build.
- Impact: How much will this project move the needle on your goals and strategy? To make this more uniform, Sean uses a multiple-choice scale: 3 for “massive impact”, 2 for “high”, 1 for “medium”, 0.5 for “low”, and finally 0.25 for “minimal”.
- Confidence: Based on what you know, how confident are you that this feature will be a success? Confidence helps back you up if you think a project will be impactful, but don’t have data to back it up. Again, using a simple multiple-choice scale makes this easy: 100% is “high confidence”, 80% is “medium”, 50% is “low”. (And anything below that is “total moonshot”).
- Effort: How much time will the project require from product, design, and engineering teams? You can measure this in “person-months” and stick to whole numbers (with a minimum of half a month).
Once you’ve got all your numbers for each feature, it’s time to put them into a simple equation:
The resulting score gives you “total impact per time worked”—a pretty powerful number for prioritizing features accurately.
5. Use a Priority Scorecard to score features by custom criteria
The RICE Method isn’t the only way to accurately score your features. Sometimes you need to customize the factors that you score by to make sure that all your stakeholders’ needs are being met. In this case, a simple Priority Scorecard might be a better option.
With a Priority Scorecard, you start with a proposed list of parameters and their “weights” (basically, what is their importance as a percentage of the whole project?) It’s important that you start by making this list on your own, but then get stakeholder feedback to finetune the numbers.
Here’s an example from project manager Daniel Elizalde:
|Category||Customer Engagement||User Experience||Sales Funnel||Operational Efficiency||Total|
Now, for each feature you’re prioritizing, assign a score from 1–100 for each of the Priority Scorecard categories. (100 means high impact on that category. 0 means no impact.)
So, if we’re trying to prioritize between a website redesign and a new checkout experience, it might look something like this:
|Category||Customer Engagement||User Experience||Sales Funnel||Operational Efficient||Total|
The total score for each feature is calculated by multiplying the score by the weight. So for our website redesign that’s: 90 x 20% + 90 x 10% + 60 x 30% + 50 x 40%.
What’s great about this method is that as long as the features are all under the same theme, it allows you to prioritize based on the specific needs of your different stakeholders. And while there are still some opinions and biases in the model (based on how you figure out weights), it still lends significant credibility to the priorities on your roadmap.
6. Use the Kano Method to prioritize features by delight
As you prioritize features, you don’t want to forget that the end goal is to create something your customers will love.
With the Kano model, you get to see each potential feature through the lens of customer delight. It’s a bit of a more complicated process compared to the other methods we’ve looked at, but it can bring you some amazing insights when you’re feeling stuck.
With the Kano model, each potential feature is broken down into different categories and their emotional responses:
- Attractive Needs: These features trigger feelings of satisfaction and delight, but users are not dissatisfied if the feature is not included.
- Performance Needs: These features result in delight if they’re present and dissatisfaction when they’re not (or are flawed). They’re very one-dimensional in nature and rely on great execution to be valued by users.
- Basic Needs: These are your must-haves—the features your customer expects to be there. Not including them is dissatisfying, but the ROI of improving them tapers off quickly.
Lastly, there’s also features that are simply undesired and take away the positive impact of your other features. Avoid these at all costs.
Now, here’s where things get tricky. To understand where each feature sits on the curve, you need to talk to a representative group of 12—24 users and ask them a few simple questions:
- How would they feel if the feature was present?
- How would the feel if the feature was not provided or not as fully present?
Each positive/negative question pair is answered with either “I like it,” “I expect it”, “I’m neutral”, “I can tolerate it”, or “I dislike it”. Based on the response, you can plot which emotional curve that feature sits on.
(Here’s a more in-depth explanation of implementing and using the Kano model from UX Magazine.)
7. Prioritize features by constraints
Lastly, if you’re still unsure of what features to prioritize based on the resources and needs you do have, it can be just as powerful to prioritize by what you don’t have.
Constraints like time, people, money, and process can be great ways to whittle down your options and focus in on features that are the most realistic and high-value.
Most constraints come down to two large buckets: People and processes.
For people: Start by asking if you have the right people for these projects. If yes, which of your right people would bring the best results? Are they available? If you answered “no” to the first question you might want to reconsider the features or look at hiring a freelancer or outside firm.
For processes: Look at all the non-human things that limit what you can do. This means time (i.e. Will this feature fit into your delivery cycle?) as well as dependencies (i.e. What else needs to be done now or later to make this feature work?) Every feature you prioritize adds complexity to your overall product. And eventually the dependencies can be a major factor in whether or not you should build something.
Final thoughts on feature prioritization
There’s always a ton of excitement in the air when you start talking about things to build. New features are exciting. You can picture all the amazing places your product could go, the results they could bring, and what the best case scenario is. But as a product manager, you need to be the voice of reality.
As you go through and prioritize features into your product roadmap, remember that your overall strategy and product roadmap always need to be front and center. Don’t lose focus on the bigger picture by some exciting idea. Long-term strategy always trumps short-term results.
Be conservative and try to live by the mantra of “less is more”. Big features can just as easily be big risks if you don’t have the data and user research to back them up. Whenever possible, use Agile development practices to launch early and often.
Lastly, make time to regularly re-prioritize. Business needs change. Markets change. Leadership changes. And despite all the work you put into prioritizing features, those priorities will change as well. Set aside time to go through your list and make sure everything is still aligned with the bigger picture.