I put together a roundup of seven feature request tools recently. Seven is a good number for seeing the landscape and a bad one for making a decision, each tool gets a paragraph, and a paragraph can't tell you what something costs in year three.
So here's the narrower cut. Canny, Featurebase and Loopful overlap with each other more than with anything else on that list: all three target a SaaS team that wants boards, voting, a roadmap and a changelog without a full product management suite. On a feature grid they look almost identical. What differs is the billing model, and that shapes your invoice in completely different directions as you grow.
| Canny | Featurebase | Loopful | |
|---|---|---|---|
| You're billed for | Tracked users | Seats | Nothing variable |
| Bill goes up when | Your customers participate | Your team grows | — |
| Entry paid tier | $79/mo Pro | $29/seat/mo | $12.50/mo |
| Plus metered | — | $0.49 per AI resolution | — |
| Free tier | 25 tracked users | 1 seat | 1 board, 1 team member |
All prices annual billing, checked August 2026.
Canny
Canny has positioned itself around revenue. The headline feature, Autopilot, reads Intercom, Zendesk, Gong and Slack conversations automatically, extracts feedback without anyone retyping it, deduplicates it and connects feedback to customer context. Deeper CRM and opportunity-value features, HubSpot, Salesforce, revenue attribution, are on the Business tier. Prioritization is framed in opportunity value, a feature tied to $500K in open deals can outrank one attached to $10K of existing revenue. If your product decisions need to survive a conversation with someone who wasn't in the room, that framing is useful.
It's also the deepest integration catalogue of the three, with Jira, Linear, GitHub, Asana, Salesforce, HubSpot, Segment, Zapier and an MCP server that lets you pull Canny data into Claude or ChatGPT.
The billing model: a tracked user is anyone with an attributed post, vote or comment. The count doesn't reset monthly, although users whose activity is only tied to completed or closed posts may stop counting. As more customers participate, the count can cross into the next increment. You can set a spending limit from the billing page to cap surprises.
Three plans, from Canny's pricing page:
- Free — 25 tracked users, 5 managers, unlimited boards and feedback.
- Pro — from $79/mo annually at 100 tracked users. Climbs on a published schedule.
- Business — custom, 5,000+ tracked users, SSO and CRM integrations.
The Pro ladder, from Canny's help center:
| Tracked users | Pro, annual | Pro, monthly |
|---|---|---|
| 100 | $79 | $99 |
| 200 | $129 | $161 |
| 500 | $279 | $349 |
| 1,000 | $529 | $661 |
| 1,500 | $629 | $786 |
| 2,000 | $729 | $911 |
| 5,000 | $1,079 | $1,349 |
Featurebase
Featurebase started as a feedback tool and now covers the whole support surface: omnichannel inbox, AI agent (Fibi), help center, feedback boards, roadmaps, changelog and surveys, all in one subscription. The consolidation argument is real: if you're currently paying separately for a helpdesk and a feedback tool, collapsing them has a straightforward ROI.
Featurebase moved from flat plans to per-seat on 4 December 2025. Existing customers were grandfathered; new customers are on the seat model. Half the comparison posts still describe the old plans, so treat any source that doesn't mention per-seat with caution.
Current tiers, from Featurebase's pricing page:
- Free — $0, one seat, one board, one roadmap
- Growth — $37/seat/mo. Widgets, integrations, API, MCP, webhooks
- Professional — $75/seat/mo. Workflows, SLAs, AI deduplication, moderation, end-user SSO, 20 Lite seats
- Enterprise — $129/seat/mo. Admin SSO (SAML/OIDC, five-seat minimum), multi-brand, 50 Lite seats
Fibi, the AI support agent, adds $0.49 per resolved conversation on top of every paid plan.
Loopful
Loopful's scope is narrower by design: feedback boards, voting, roadmap, changelog, widget, surveys and insights. The entire product is the feedback loop.
Three plans, from Loopful's pricing page:
- Free — one board, one team member, unlimited posts, votes and tracked users, roadmap, widget, moderation, duplicate merging, smart tags. No time limit.
- Pro — $12.50/mo. Unlimited boards, team members and feedback; changelog with email announcements; insights; private boards; custom statuses and fields; surveys; segment filtering; API, webhooks, Slack, Asana, Jira, SSO; custom domain and CSS.
- Max — $20.83/mo. Unlimited workspaces, white-label portal and widget.
A few tier placements are worth comparing directly. SSO is on Pro at $12.50. On Canny it requires the custom-priced Business tier. On Featurebase, end-user SSO is $75/seat (Professional), and admin SSO over SAML/OIDC is $129/seat (Enterprise, five-seat minimum). The embeddable widget is on Loopful's free plan; Featurebase's free tier doesn't include it. Canny's free tier caps tracked users at 25.
The flat model means nothing scales with participation and nothing scales with headcount.
The numbers side by side
Annual billing, feedback features only.
| Scenario | Canny Pro | Featurebase Growth | Loopful Pro |
|---|---|---|---|
| 4-person team, 200 engaged users | $129/mo | $116/mo | $12.50/mo |
| 8-person team, 1,500 engaged users | $629/mo | $232/mo | $12.50/mo |
Canny's second row jumps because 1,500 customers participated. Featurebase's jumps because four people were hired. The numbers are just doing what the billing model says they should.
If you're also replacing your helpdesk
This one isn't like-for-like: Loopful has no support agent, so the comparison doesn't extend here. For a 10-seat team resolving 2,000 conversations a month with Fibi, Featurebase Professional is 10 × $59 + 2,000 × $0.49 = $1,570/mo, with the AI meter larger than the subscription. Whether that's expensive depends on what you're replacing, for teams moving off Intercom or Zendesk it can be the cheaper side of the trade. The point is that the AI line needs modelling before you sign.
The monthly snapshot also understates the trajectory. Run the 8-person scenario forward three years, with the team growing from 8 to 14 and the board still active: Featurebase reaches $406/mo without a single price change, Canny steps up the ladder as the tracked-user count crosses each pricing threshold, and at today's pricing, Loopful's invoice is the same number it was on day one.
Five questions worth asking in any demo
- Does the billing metric reset? Canny's tracked-user count doesn't reset monthly, check how historical participation affects your current number as posts complete.
- What's your seat count in 18 months? Per-seat models are easy to budget at four people and harder at fifteen.
- Is loop closing automatic, and what triggers it? Worth asking precisely: at Canny, changelog emails go to subscribers; at Featurebase and Loopful, voters on linked requests are notified directly.
- Which SSO, and on which tier? "Yes we have SSO" can mean three different things across these three tools, the tier placement matters more than the yes.
- What does data export look like? All three import. Check what leaves.
What it comes down to
Each of these tools is a bet on which number in your business grows fastest, customer base, team size, or neither. If your core need is feedback collection, voting, a roadmap and a changelog, the feature overlap is substantial enough that the billing model can become the deciding factor.
