A feature request tool is a public place where customers post ideas, vote on the ones they want, and watch a status move from under review to planned to shipped. Underneath that description sit three kinds of product: a hosted subscription, a self-hosted open-source project, or a static site you own outright.
What are the parts of a feature request tool?
Every tool in this category is built from the same six pieces, whether it costs nothing or nine hundred pounds a year.
- A public board. Customers see what others have already asked for, so the same idea doesn't get filed forty times under forty titles.
- Voting. A number next to each request that tells you which twenty ideas matter more than the other two hundred.
- Statuses. Under review, planned, in progress, shipped, declined. The status is the whole point: it is the difference between a suggestion box and a board people trust.
- A roadmap. The planned and in-progress statuses, shown as their own page. Whether that page should be public is its own decision, covered in public roadmaps.
- A changelog. Where a shipped request becomes an announcement, ideally linked back to the request that caused it. The case for keeping one properly is in changelogs and release notes.
- Notifications. An email to everyone who voted, the moment the status changes. Skip it and voting becomes a one-way action nobody repeats.
None of the six is hard to build alone. What is hard, and what the pricing in this category is really charging for, is running all six as one product that does not break.
How do you manage what comes in?
A board fills up faster than anyone ships from it. The unglamorous work is deciding which twenty requests earn the next three months, not adding a seventh status. That process, and the traps in doing it by vote count alone, is covered in how to manage and prioritise feature requests.
What are the three ways to get one?
Every product in the category, ours included, is one of three things wearing different branding.
| Approach | What it is | Who runs it | Examples |
|---|---|---|---|
| Hosted vendor | A subscription you log into | They do: patching, backups, uptime | Canny.io, UserJot, Featurebase, Frill.co, Nolt.io |
| Self-hosted open source | Software you install on your own server | You do: the database, the mail server, the updates | Fider, LogChimp, Quackback, ClearFlask |
| A static site you own | Files plus a couple of small endpoints, no database to run | Nobody has to: there is no server to keep alive | Docket |
The differences between the three are not really about price. They are about who answers when something breaks, and how much of the product you actually need.
What separates the hosted vendors from each other?
The five hosted vendors are not interchangeable, and the concessions below are worth reading before assuming the cheapest one wins.
Canny.io runs Autopilot, which reads Intercom, Zendesk, Freshdesk, Gong and Zoom conversations and captures, dedupes and triages feedback without anyone filing anything, syncs with HubSpot and Salesforce, carries SOC 2, CCPA and GDPR, and gives unlimited boards, including private ones, even on its free tier.
UserJot catches duplicates with AI as the customer types, drafts the changelog entry itself once a request ships, emails everyone who voted, and ships a two-minute importer that carries boards, posts, votes, comments, statuses and users over from Canny.io. It also runs a remote MCP server.
Featurebase is not really a feature request board, it is a suite: a full helpdesk with live chat, a help centre with AI search and multilingual translation, NPS and CSAT surveys, and outbound email, built on SOC 2 Type II and EU hosting in the Netherlands and Germany. Startups under two years old with fewer than six staff get 86 per cent off.
Nolt.io speaks 22 or more interface languages, integrates two ways with Jira, Linear, Asana, monday.com, Intercom, Slack, Discord and Teams, and backs Enterprise with a 99.9 per cent uptime SLA and a separate air-gapped EU product.
Frill.co embeds NPS, CSAT and idea polls directly inside your product, schedules and segments announcements, and does similar-idea detection, smart merge and a priority matrix the others do not.
What all five share: somebody else runs it, patches it, backs it up and answers for the uptime. That is a genuine cost to skip if nobody on the team wants to own a server.
How do the pricing models differ?
The pricing model matters more than the number on the page, because two of the four models grow while you use the product and two do not.
| Model | Who does it | What grows the bill | Three-year effect |
|---|---|---|---|
| Per tracked user | Canny.io | Anyone who posts, votes, comments; the count never falls | Entry $79/mo billed yearly rises to $1,079/mo at 5,000 tracked users, same billing |
| Per seat | Featurebase | Every teammate with a login | $348/year at one seat, $1,044/year at three, before the badge add-on |
| Per workspace | UserJot, Frill.co, Nolt.io | Nothing, until board or feature limits are outgrown | Flat: $1,044/year (UserJot, Nolt.io), $792/year (Frill.co) |
| One payment | Docket | Nothing; there is no second charge | $99 once, $0 after |
Canny.io defines a tracked user as anyone who posts, votes or comments, and the count only moves upward. Exceed the limit and Canny.io "automatically upgrade[s] you to the next tier" with a prorated charge, unless a spend limit is set, at which point it "will stop tracking feedback from new users". Full three-year figures for every tool in the category are in what feedback tools actually cost.
What does self-hosting actually cost?
Self-hosting means running the software yourself, and the open-source field is smaller and rougher than the hosted one.
Fider is the largest project: AGPL-3.0, 4,476 stars, 842 forks, still pushed to as recently as 9 August 2026. It needs Docker, PostgreSQL 12 or newer and your own SMTP or SES, gives you a roadmap, voting and private boards, but no changelog, and patching is your job: three security releases landed inside three months in 2026. Their own hosted option, fider.io, run by Northern App Labs Ltd, exists because most people who try to self-host it end up not doing it.
LogChimp is open core in the way that matters least conveniently: Boards, Roadmaps and Comments, three of the six pieces this guide opened with, sit behind a $24-a-month Pro tier even on a self-managed install.
Quackback is smaller (237 stars) and the only open-source project doing real content marketing about its own running cost: a $10-to-20-a-month VPS for a small team, $50 to $100 for a larger one, against a SaaS range it puts at $29 to $999-plus a month.
ClearFlask exists, Apache-2.0, 448 stars, but is very small.
GitHub Issues and Discussions cost nothing on a public repository, with genuinely strong threading, reply-by-email and full-text search. Against that: posting needs a GitHub account, which most buyers of a consumer product will not have; reactions are not votes; there is no changelog, testimonials or email capture; and every complaint is public and indexed against your product's name.
The honest running cost, from UserJot's own costed breakdown of self-hosting Fider: Postgres at around $20 a month, a host at $20 to $40, transactional email at $10 to $35, monitoring and TLS at $20 to $30, totalling $70 to $125 a month before any engineering time, plus five to ten hours a month once traffic is real. Canny.io's own build-versus-buy piece, published 19 August 2026, prices building a feedback portal from scratch at around 420 engineering hours, about $31,500, plus $45 to $550 a month to run: "the code was never the hard part."
Both figures describe database-backed self-hosting. A static site avoids the database entirely, the case made in self-hosting a feature request board; the open-source field is compared tool by tool in the open source options compared.
One payment, no subscription, unlimited products.
What should you check before committing?
Three questions the pricing pages never raise, and all three matter more after you have committed than before.
What does your export actually contain? Canny.io's CSV export is post data only, with no comments; those need a separate call to the JSON API. Canny.io's own public board carries open, unshipped requests asking for full data export and user export. Since 12 September 2025 the EU Data Act has capped the switch away from a data processing service at 30 calendar days, though that clock only starts after a notice period the contract may set at up to two months, and no vendor in this category mentions any of it on a pricing page. We are not lawyers and this is not legal advice, and whether a feature request board counts as a data processing service under the Act is a judgement we are not qualified to make.
What does removing the badge cost? On Featurebase it is $69 a month, purchasable only alongside Professional or Enterprise, so badge removal alone runs to $828 a year, more than most of the tools here cost in total. Canny.io puts it on Business tier only. Frill.co removes the tagline free on every tier but keeps full white-labelling as a paid add-on. Check it before signing up, not after.
What happens if the vendor is acquired or shuts down? This category has already lost several products. Olvy announced closure in January 2025 and closed the following month. Astuto's repository was archived read-only on 8 February 2026, after its hosted service ran under a year. Pendo's classic Feedback product, formerly Receptive.io, was sunset on 1 August 2026. Delighted closed with renewals blocked from July 2025. A board carrying years of requests, votes and comments is not a small thing to lose.
Which approach is actually right for you?
None is correct by default.
A hosted vendor is right the moment nobody on the team wants to own infrastructure, or when the product needs something specific to one vendor: Canny.io's Autopilot triage, Featurebase's full helpdesk, Nolt.io's language coverage, Frill.co's embedded surveys. Paying a subscription for those features is a reasonable trade, and pretending otherwise is the advert this guide is trying not to be.
Self-hosting an open-source project is right for a team that already runs Postgres, already has someone who owns a server, and wants the roadmap and voting pieces without a subscription, accepting UserJot's own figure of $70 to $125 a month and five hours a month of upkeep. It is the wrong choice for anyone without that person already in place.
A static site you own, which is what Docket is, fits a narrower case: wanting the six pieces above, owning the data outright, and not wanting a database to run at all. It costs $99 once, or $249 for the automated and notification pieces on top. It is the wrong choice for anyone who needs AI triage, a full helpdesk, or multilingual coverage, because a static site cannot sync a CRM or run live chat.
Where to go next
This page is the overview. Each part of it has a page of its own.
- Choosing a tool. Best feature voting tools in 2026 ranks the hosted options on a stated criterion, and free and open source feature request tools separates the four different things "free" means in this category, which is where most people get caught out.
- The changelog half. Best changelog tools in 2026 covers publishing the other end of the loop.
- Collecting what arrives. How to collect feedback from customers ranks the methods honestly, and how to write a good bug report is the craft version for the people filing them.
- The forms themselves. A bug report template you can copy and a feature request template you can copy.
- Neighbouring categories, and which one you actually need. Best product feedback software separates the four shapes people conflate, best idea management software is for ideas from your own staff rather than customers, suggestion box software is the cheap simple end where the answer is often not software at all, voice of the customer tools is the enterprise end where almost nobody publishes a price, and website feedback widgets capture a screenshot from a live page, which is a different job again.
- Showing the good feedback to other buyers. How to collect testimonials, and where to put them.
- If you already live in GitHub. A GitHub feature request issue template turns issues into something closer to a board without adding a tool.
- Leaving a specific tool. One page per vendor, priced, with what each one still does better: Upvoty, Sleekplan, FeatureOS and Feature Upvote.
- Looking for an open source replacement. Open source Productboard alternatives and open source UserVoice alternatives both start from the same honest place: no open source project replaces the whole of either. Fider alternatives is for the opposite case, where the software is fine and the operating burden is not.
- Chasing a name that no longer exists. Roadmunk alternatives, which is now Strategic Roadmaps under Tempo.
Frequently asked questions
What's the difference between a feature request tool and a helpdesk?
A feature request tool tracks ideas customers have not got yet: a public board, votes, a status, a roadmap. A helpdesk tracks problems customers already have: tickets, chat, replies. Featurebase merges both into one product; most of the category keeps them separate, which is worth checking before assuming any one tool covers everything you need.
Do I need a public roadmap?
Not necessarily. Showing planned work in public stops the same request landing forty times and gives customers something to check before they file, but it also commits you to dates you might miss and shows competitors what is coming. Most tools let a roadmap stay visible to logged-in customers only, rather than the open web.
Is voting the same as prioritisation?
No. A vote count tells you which twenty requests are loudest, not which twenty matter most: a single enterprise account can be worth more than two hundred votes from free users, and vote counts say nothing about how hard a request is to build. Treat votes as one input into prioritisation, never the whole decision.
What's the cheapest feature request tool over three years?
It depends on the pricing model, not the sticker price. A one-time purchase like Docket's $99 stays $99. Workspace-priced tools like Frill.co and UserJot run $792 to $1,044 over three years. Per-seat and per-tracked-user tools can run far higher: Featurebase at three seats reaches $3,132, and Canny.io's entry tier reaches $2,844 before any growth in tracked users.
Can I move to a different tool later if this one doesn't work out?
Sometimes, and it depends on the vendor's export. UserJot ships a two-minute importer that pulls boards, posts, votes, comments, statuses and users straight out of Canny.io. Canny.io's own CSV export is post data only, with comments needing a separate API call. Check what an export actually contains before you need it, not after.