Every Sunday, you'll get a new lesson about product, design & startups to your inbox. Researched, heavily user focused & without fluff.
Most features die quietly. No one uses them. In fact, over 80% of new features are rarely or never used, according to a feature adoption study by Pendo. Every feature is a bet. Here’s how I approach the full lifecycle of a product feature: 1. Discovery – Where the idea comes fromFeature ideas usually come from one of three places:
Ideally, you build features that solve real pain, not just ideas that sound cool. If you’re not getting real feedback, go find it. Ask them about their problems, not about your product, but about their work or life. They won’t give you the solution. But they’ll give you the pain. Once you identify a strong pain point, write it down as a user story. Then, turn that into a hypothesis: “We believe solving [pain point] with [this solution] will result in [measurable outcome] because [rationale].” Example: “We believe solving the lack of visibility for non-technical users with a weekly insights email will increase engagement by 30% because users are more likely to open emails than log into a dashboard.” This is key. It gives you something to measure later and a reason to remove the feature if it doesn’t deliver. 2. Design – Explore ugly and beautiful solutionsThis is where you think through how to solve the problem. I usually do two things in parallel:
Then we find a middle ground with the team. What’s shippable? What’s valuable? What’s fast? And don’t forget to involve developers early. Their input will save you time and pain later. Tools like Lovable make it super fast to turn rough ideas into prototypes, so you can test early without waiting for dev time. 3. Testing – Put the idea in front of humansBefore you build anything, you want signals. Take your prototype and test it:
Ask:
If reactions are lukewarm, rethink the idea. You’re better off pivoting now than launching something no one needs. 4. Delivery – Ship it fast, but prep for what’s nextYou’ve got alignment. It’s time to build. But don’t just write code and push it live. You also want to:
Shipping should be fast, focused, and reversible. 5. Activation – Make sure users know it existsSo many features die here. They’re built. They’re shipped. But nobody sees them. Your job isn’t done after deployment. You need to activate users. How? Depends on your audience:
One smart trick: Beta tester mode. Let users opt in to new features early. They’ll give better feedback and feel part of the process, even if the feature isn’t perfect yet. 6. Impact – Measure, then decideOnce the feature is live and people are using it, start measuring. Look at:
If your original hypothesis was “this will increase weekly active users,” check the data. Did it? If yes → keep it If somewhat → evolve it If no → kill it No hard feelings. No ego. Just product thinking. Zombie features do more harm than good. Final thoughtEvery feature is an experiment. You don’t just build it once. You learn from it. Treat your features like bets. And be brave enough to remove the stuff that doesn’t work. That’s how great products are built. |
Every Sunday, you'll get a new lesson about product, design & startups to your inbox. Researched, heavily user focused & without fluff.