How do I decide which feature to build next?
19 September 2026 · by Nordva, who make Nordva Launch
The short answer
Weigh requests by who asks, not how many times: a request from paying customers who use the product every week counts more than ten from people who tried it once. Look past the feature they name to the job they're trying to do, group requests by that job, and build the one that helps the most paying users do what they came for.
Count people, not requests
A long list of requests hides the thing that matters: who asked. Put each request next to the person, with three facts about them:
- Do they pay?
- Do they use it regularly?
- Are they the kind of user you built it for?
A request from a paying customer who uses the product every week is worth more than ten from people who signed up once and left. Some requests come only from people who will never pay; building those makes the free version better and the paid one harder to sell.
Find the job behind the feature
People ask for a solution: “add a CSV export”, “integrate with Notion”. Ask what they’d do with it. The answer is the job: “I need to send a weekly report to my boss”. Three different requests often turn out to be the same job, and one feature (sometimes not the one anyone asked for) solves all three.
Group requests by job, not by feature name.
Choose with three questions
For each job on the list:
- How many paying or regular users need it?
- Does it help people reach the first useful moment, or keep them coming back? Those beat nice-to-haves.
- How long will it take? A small thing that many paying users need goes first.
Say no to the rest for now, and say so to the people who asked. “Not now, and here’s why” keeps trust better than silence.
Close the loop
When you ship something people asked for, tell each of them. It turns a request into a reason to come back, and people who see their feedback built tend to send more.
A warning about the loudest users
The users who write most are not always the users you want more of. Check where your most engaged paying customers came from and what they asked for; build for more people like them.
With Nordva Launch
Feedback arrives in one inbox, sorted into feature requests, bugs and more, with the sender attached (on Free by keywords, 50 messages a month per project; on Indie by AI). On Indie, a person’s contact record puts their feedback next to their waitlist signup and Stripe payments, so “do they pay?” is on the same screen. Link a request to a changelog entry, and when you publish it, Launch emails the person once to say it shipped, if they left an email.
When Launch is the wrong tool
If you have a large team and a roadmap tool with voting, you may already have this process. This page is for founders deciding alone, with a handful of users and a long list of wishes.
- How do I prioritise feature requests?
- What should I build next for my SaaS?
- How do I handle feature requests from users?