Should I prioritize feature requests by votes?

Not on their own. A vote count mixes how many people want something with how loudly a few of them want it, and it ranks requests (guessed solutions) rather than needs. Use votes as one signal after you have grouped requests by need and counted the distinct customers behind each need.

What is the difference between a feature request and a customer need?

A feature request is a proposed solution (“add an Excel export”). A need is the problem or goal behind it (“get month-end numbers to our accountant”). Needs are more stable, several requests often share one, and there is usually more than one way to meet a need.

How do I handle feature requests I won’t build?

Say so, with a reason, to everyone who asked. Keep the request as evidence: if the same need keeps coming up, that is information worth having. A clear no keeps trust better than silence.

Which prioritization framework should I use: RICE, ICE or Kano?

Scoring frameworks help once you know what the need is and how many customers share it. They cannot fix an input that counts clicks instead of customers, so do the grouping and counting first, then score the needs you are considering.

How often should I review feature requests?

Triage new requests continuously, ideally within days, so people hear back. Review the patterns monthly: which needs are growing, which have no work behind them, and how quickly your team replies.

Guide · Feature requests

How to prioritize feature requests

The most-voted request is rarely the most important one. Here's a method that turns a pile of requests into decisions you can explain, with any board or spreadsheet.

The method

Seven steps from requests to decisions

Each step fixes one way a request list misleads you.

1
Treat every request as a clue, not a spec
“Add an Excel export” is one customer’s guess at a solution. Ask what they are trying to get done, and write the need in their words: “get month-end numbers to our accountant”. Several different requests often turn out to be one need, and one request can hide two.
2
Count customers, not votes
A vote count adds up enthusiasm, so one power user can look like a groundswell. Count the distinct customers behind each need, across your board, interviews, support tickets and sales notes, and count each customer once however many places they said it.
3
Know who is asking
Ten customers on a free plan and ten of your largest accounts are different signals. Attach what you know about each customer (plan, size, segment) so “what do our enterprise accounts need?” is a question you can answer, not a guess.
4
Make votes scarce
When voting is free, people vote for everything. Give everyone a small budget, a handful of votes with a cap per request, so a vote says “this is in my top few”. Let people take votes back freely so the budget stays current.
5
Look for the gap
Put each need next to the work already in flight. The most valuable row is a need many customers share that nothing you are doing answers. The second most valuable is work in flight with no customer need behind it.
6
Close the loop, including the no
Tell everyone who asked what you decided: planned, shipped, or declined with a reason. A no with a reason keeps the customer, and the request still counts as evidence the next time the need comes up. Measure how long people wait for a first reply.
7
Review how it is going
Once a month, ask three questions: are people still asking, are they hearing back, and do we ship what they ask for? Watch new supporters per week rather than all-time vote totals, which only ever go up.

A worked example

38 votes, or eleven customers?

“Export the monthly report to Excel” has 38 votes. On its own, that number says the request is popular. Read the requests behind it, though, and most of them are about one thing: getting month-end numbers to an accountant. Eleven customers share that need, across the board, a few interviews and a handful of support tickets.

Now check the work in flight. If nothing you're doing answers that need, you have found the gap, and you can say why it matters in one sentence. Maybe an export is the answer. Maybe a scheduled email to the accountant is better. Either way, you're deciding about a need, not a vote count.

▲38votes

Export the monthly report to ExcelOpen

Every month I copy the numbers into a spreadsheet by hand so our accountant can close the books.

Dana Okafor · Northwind

What it turned out to mean

Get month-end numbers to our accountant
Heard from 11 customers
Share results with people who don't log in
Heard from 6 customers

Worked on in

No opportunity cites this yet. Eleven customers share the need, and nothing in your trees is working on it.

Common mistakes

Ranking the backlog by vote count.
Treating the requested solution as the requirement.
Letting the loudest customer set the roadmap.
Going quiet on requests you won’t build.
Counting the same customer once per channel.

Scoring frameworks like RICE, ICE or Kano help once you know what the need is and how many customers share it. They can't fix an input that counts clicks instead of customers.

Doing this in Outcomify

A feedback board built around this method

You can do every step with a spreadsheet and patience. Outcomify’s feedback board does the heavy lifting.

The need behind each request

Approve a request and Canopy reads it into the needs in your research corpus.

Customers, not votes

Each need counts the customers behind it across requests, interviews and tickets.

Who is asking

Customer and account data from your own login or the API, so you can narrow to one group.

Scarce votes

Seven votes each by default, up to three on any one request.

The gap

Every request shows the opportunities citing its needs, or says no opportunity cites them yet.

Close the loop

Status notes and replies reach the person who asked, by email, in your words.

Review

The board report: requests, new supporters, what shipped and time to first reply, week by week.

Ask your agent

The board is MCP-enabled: Claude or any MCP client can tell you what customers ask for that nobody has picked up.

See the feedback board Start free trial

Frequently asked questions

Prioritizing feature requests

Keep exploring

Stop counting votes

What a vote count hides, and what to count instead.

The Outcomify feedback board

Requests become evidence, votes become priorities, and customers hear back.

Customer interviews to opportunities

Turn what customers tell you into needs you can act on.

What is an Opportunity Solution Tree?

A map from outcomes to the needs and solutions behind them.