 Hi everyone. Hope you're well and thank you for joining me to talk a bit about feature strategy. Today, I'd like to share some thoughts around how to prioritize features in products, how to launch them, and evaluate the impact to be able to enhance the product over the long term. But before we start, I'd like to share a little bit about myself. I'm Dan. I'm a product manager for more than 12 years now. I came from business where I was keen to improve processes and make our teams work more efficient. Due to my curiosity, I slowly transitioned to product management, which back then was writing huge specifications and praying for six months that the feature will be delivered as expected. But what shall I say? They never did. We quickly realized that we need to change our approach and apply agile principles to deliver incremental value quickly and get insights about most pressing problem spaces. I then joined SoundCloud, where we built insights and analytics platform from scratch with the mission to empower emerging artists through actionable recommendation to understand, build, and connect with their audience and grow their careers. In addition, we launched fan powered royalties, also known as the user centric model, a new way for artists to earn money from streaming services. I recently joined Spotify to help achieving the mission to unlock the potential of human creativity by giving a million of creative artists the opportunity to live of their art and billions of fans the opportunity to enjoy and be inspired by it. But enough about me. Let's jump into the topic. The feature strategy is one pillar of the wider product strategy work. The feature strategy focuses on improving our ability to create and capture value. The growth strategy on the other hand largely focuses on maximizing products existing value proposition. Effective growth strategies connect acquisition, retention, and monetization. It's important to move away from the funnel thinking towards growth loops where we have an input and action and an output which will feed the next loop and so on. The product market expansion strategy tries to add value in two ways. One is to adapt products to new complementary markets. The other on adding new complementary products with the goal to overcome saturations like market saturations, market capture reaches the natural ceiling or product saturations where the product becomes fully optimized for its use case. The last product strategy pillar is scaling where you invest in supporting processes, infrastructure, and strategies which support the three previous mentioned strategy layers. Key pillars here are tech scaling, platformization, technical depth management, monetization of UX, process scaling, process improvement evaluations, value stream mapping, and user scaling. Value added use cases, underserved user segments, or identifying bad behavior. So with our feature strategy work our goal is to improve the ability to create and capture value. Features can create value in three ways through acquiring new users, retaining assisting ones, and monetizing. It's crucial to evaluate and enhance existing features as this will inform your future build strategy and help you develop new features and continue the loop of evaluating new feature performances post launch with the goal to achieve feature product fit. There are strategic insights you need to consider like what is our company vision, what are our group or team objectives, and how will this feature help achieving them. Then we have user insights. These are our qualitative insights together from interviews, user research, or surveys to fully understand the problems our users have. In addition we have data insights, and here we track the actual consumption or behavior to understand patterns, identify, and validate hypothesis we have. Qualitative evaluations are very important as they help you with questions like how many users is this feature or product designed for, how does the feature add value to the users, and how does the feature add value to our business. It's also crucial to think about user problem severity. If the user has low disruptions usually they found easy workarounds. With moderate disruptions these workarounds become more complex and painful. High disruptions on the other hand prevent the user's experience in the value proposition, which is very critical. Another aspect is identifying user categories. Who are my core users, those for whom the feature is designed for? Who are my adjacent users who are getting some value from the feature but they haven't been considered why designing it? And non-adjacent users. These aren't getting any value from the feature or product. Based on all previous mentioned areas you're able to create this qualitative feature map. This map has three pillars. The first is the target population in which you describe the target group. The segment your features is designed for and the target size. The estimated percentage of users or segment represents. The second pillar is the user value. Here you describe the user problem. The feature or product tries to solve the problem frequency. So how often the user is experiencing it and the problem severity. How painful is it for our users? The last pillar is the business value. What impact will this feature bring and how strategically important is it? A great mythology that you can use to help you be better prepared for the upcoming release is the pre-mortem. It helps you think about what could happen, good or bad, so that you can plan before it starts. Some questions you can ask are how does the feature add value to a product? What target segment are the primary set of users? How much confidence do you have in the success of the feature? How will we launch this feature? What metrics will indicate the success? What outcomes are possible for the feature performance? What would each outcome teach us about the performance? And what would the next step be after each of those outcomes? These are only, of course, recommendations and you should identify outcomes and metrics tied to your company and product. This framework is great as you're able to validate and compare your assumptions with the actual results post implementation and it will also help you strengthen your product sense. And before you roll up your sleeves, you should make sure to pressure test your insights. Did you use the right data? Have you avoided biases? Make sure the right feature has been prioritized and sense check the launch plan in order to minimize the costs and maximize the value. Now we should be well prepared to release our features but how shall we do it? Which model is appropriate? We have the experimental release which focuses on running experiments to validate key assumptions we have for our feature in order to learn and shape it over the time. Then we have the minimum viable feature release, a fully functioning version of the feature but with a minimum functionality as our goal is to validate the core value proposition. And lastly the face release which delivers a more robust version of the feature but broken down into faces in order to continuously release a value when ready. These methods don't compete with each other it's actually the opposite. A feature needs to run through each of those release methods based on the ambiguity spectrum. If you have high ambiguity and not enough insights from our already mentioned sources like strategic user or data insights then the experimental release is your choice as you need to figure out what the right feature product is to solve the biggest problem for our users. If you have high confidence and you validated your hypotheses through the other releases then the face release is your choice as this focus more on building the feature or product right in order to deliver incremental value. Let me dig deeper into each of those methods. With the experimental release our goal is to learn and shape the feature continuously with user behavior. We're able to do it when we list all assumptions we believe in the success of the feature. Turn these into experiments that can be validated with users and refine our features with every experimental outcome. Hypothesis driven validations are key to success as clearly defined incremental experimentations lead to faster learnings and deeper insights. They're much more valuable than writing detailed specifications as I have done previously. Here you can see a step-to-step guide. First you identify your assumptions then you reframe your assumptions and hypotheses rank them in order of importance design appropriate experiments conduct these synthesize your learnings and act. The validated learning loop is really powerful you start to build your minimum viable product experiment based on your hypotheses you have then you measure the qualitative and quantitative user data learn from the results and create newly improved hypotheses and start the process again until you gain enough evidence and confidence in your solution. I receive a lot of questions around what I consider to be an experiment and the answer is whatever helps you validate your assumptions as fast as possible it could be a spreadsheet with which you validate the business logic UI sketch on paper a user flow a low no-code prototype so basically anything that makes you learn so once we have enough confidence that we have identified the right feature we can move to the minimum viable feature release method since here our goal is to validate the core user value proposition by using minimal design and functionality we can achieve that by minimizing the number of platforms the feature is built for so for example only releasing it on iOS and build desktop or Android versions later or by minimizing the number of integrations the feature needs or minimizing design and engineering resources needed now that we have validated the core user value proposition we are very confident that this is the right thing to build we can use the face release method now for smaller features this means that we can fully build the entire feature and release it for more complex features we will be breaking down the feature into faces which can be released independently by continuously delivering value we did it we have successfully released our feature but here our job doesn't end we need to evaluate the performance of the feature post launch a great tool that is doing that is the so-called retention score based on our pre-work creating the quality feature map we know what our target users are then we check how many of those users have tried the product once so adopted and how many use the feature regularly now so are retained we then need to divide the retained users with our target audience and have the retention score this exercise helps us evaluate our features as calculating the retention score is the first step which is followed by plotting the score against the strategic importance with this approach we are able to evaluate our features and the great tool for that is the so-called feature matches on the y-axis you have the retention score and on the x-axis the strategic importance we're able to categorize our features in four areas the core overperforming project and liability features we as product folks we need to have a special eye on the liability features as they have a very high strategic importance but are not satisfying our users yet for core and overperforming features it's important to maximize the value capture and try to expand the target audience for our project features you should ask yourself if it's worth investing or if we should sunset them as it's important to the overall health and success of our product if you let too many features creep into your product the core value proposition and vision can easily get diluted i hope you enjoyed and found this webinar useful please start experimenting with some of the frameworks and ideas i mentioned i look forward to hearing your feedback thank you and all the best