Featurebase

What it is

Featurebase started as a feedback board — the Canny shape of voting, a public roadmap, and a changelog — and spent 2026 turning into a full helpdesk. The same workspace now runs live chat, email, and Slack conversations in one inbox, a help center, proactive messaging, and Fibi, an AI agent that answers customers and bills per resolution. The pitch is consolidation: one per-seat tool in place of a feedback board plus an Intercom-style support stack. The meter to understand first is the opposite of Canny’s. Featurebase charges for your team’s seats and for Fibi’s resolutions, not for the end users who post and vote.

Why it shows up in Customer Success stacks

• Feedback and support share one record. Feedback 2.0 (24 September 2026) renamed posts to Requests and backs each one with Insights pulled from support conversations and call recorders — Fathom, tl;dv, Grain, Gong, Fireflies, and Granola. Agents link a chat to a Request from the inbox, and status changes flow back to the conversation where the customer first asked. That is the loop-closer a separate board and helpdesk can only approximate through an integration.
• The support surface is now real. Outbound (7 July) folded the changelog, banners, email, surveys, and targeted chat into one module. Slack became a native support channel on 10 August, with SLAs and conversion to Tickets. Salesforce two-way sync and HubSpot data sync shipped on 20 August, with requests sortable by Opportunity Value.
• AI is priced on outcome, not seats. Fibi bills only when it resolves a conversation; hand-offs to a human are free. Featurebase’s own fleet figure after Fibi 3.0 (18 June) is a 65% resolution rate, up from 50% — a vendor number, not an independent benchmark.
• Agent access is on the entry paid plan. API, webhooks, and MCP sit on Growth at $29/seat, and the July Reports release lets you query support data through the MCP server.

Pricing reality

Four plans, priced per seat; end users, posts, and comments are not metered.

• Free — $0, one seat, one feedback board, one roadmap, live chat and inbox. No Fibi.
• Growth — $29/seat/month billed yearly, $37 month-to-month. Fibi, Copilot, API/MCP, reporting.
• Professional — $59/seat/month yearly, $75 monthly. Adds workflows, SLAs, round-robin assignment, Salesforce and HubSpot, and 20 free Lite (read-only) seats.
• Enterprise — $99/seat/month yearly, $129 monthly. SSO, multi-brand, 50 Lite seats.

The seat price is not the bill. Five meters stack on top: Fibi at $0.49 per resolution (up from $0.29 on 1 July 2026, for new and existing customers alike); Copilot at 10 free conversations per seat per month, then $19/seat/month; outbound email at 5,000 free, then an unpublished rate; and $69/month to remove Featurebase branding. A worked example: three Growth seats plus 400 Fibi resolutions a month is about $283/month ($87 + $196). The same load on Intercom — Essential seats from $29 plus Fin at $0.99 per outcome — lands near $483. Early-stage companies (under two years old, fewer than six employees) qualify for 86% off.

Best for

Founders and first CS hires at B2B SaaS companies with 2-15 support seats who want feedback voting, a public roadmap, a help center, and an AI first responder in one tool for under $300/month. It fits best when your public board is large or consumer-heavy, because Featurebase’s cost does not move when 20,000 users start voting.

Do not buy Featurebase if you run 25+ support agents, need voice or workforce scheduling, or need a vendor whose AI agent has been benchmarked outside its own marketing. That is Zendesk territory, or Intercom when Fin’s deflection quality and ecosystem justify paying double per resolution.

Versus the alternatives

• Canny — pick it when you only need the feedback board and want the more mature one; Canny’s Autopilot intake is unmetered, but its bill climbs with tracked users, so a large public board flips the math to Featurebase.
• Productboard — pick it when the buyer is a product org that needs scored prioritization and feature hierarchies, not a support inbox.
• Intercom and Zendesk — the two market-share leaders in support; pick either once support volume, channel breadth, or compliance review outgrows a small vendor.
• Pylon — the fastest-growing B2B support entrant; pick it when enterprise accounts live in shared Slack Connect channels and you need account-level health views. Plain and Help Scout are the lighter API-first and shared-inbox options.

Watch-outs

• The AI rate moved 69% on 13 days’ notice. The $0.29-to-$0.49 change was announced 18 June and applied to existing customers on 1 July. Guard: on an annual contract, get the per-resolution rate written into the order form with a cap, and model your budget at $0.49 plus headroom rather than today’s number.
• “Resolution” has no published definition. Featurebase has not documented a quiet period, reopen rule, or how partial escalations bill, where Zendesk and HubSpot publish 72-hour rules. Guard: get the definition in writing before you switch Fibi on, then reconcile one month’s invoice against conversation logs.
• Revenue-sorted feedback needs Professional. The Salesforce and HubSpot sync that lets you sort requests by deal value sits on the $59 tier, not Growth. Guard: if ARR-weighted triage is why you are buying, price Professional from day one.
• Fibi is a black box to admins. Featurebase’s own feedback board carries user requests for visibility into which tools Fibi calls and why. Guard: review a sample of Fibi transcripts weekly for the first month and keep sensitive topics routed to humans by workflow.

For the CS platforms that consume this demand signal alongside health scores, see Gainsight and Vitally.