spirby.Spirby home
all posts
by Dima

flat pricing is a feature, not a limitation

tracked-user pricing punishes growth. flat pricing aligns incentives. why we copied plausible's model instead of canny's, and what we give up.

your feedback board has 100 active visitors today. tomorrow you ship a launch and it has 3,000. with canny, your bill quintuples. with spirby, it doesn't change.

this is the part of pricing nobody likes to write about explicitly, because it's uncomfortable for the vendor. it's also the part that matters most to a growing team.

the canny example

canny's published pricing as of may 2026:

  • 0–100 tracked users: $79/mo (starter)
  • 500 tracked users: $311/mo (growth)
  • 2,500 tracked users: $661/mo (business)
  • 10,000+: contact sales

a "tracked user" is a unique identified end-user who interacted with your boards in a 30-day window. as your audience grows, this number grows. the price grows with it.

if you read that table cold, the reaction is "$79 is fine, i'll take it." what you don't account for is that "tracked users" looks small at the start of the year and very large by the end of the year. teams that thought they were on a $79 plan find themselves on a $311 plan within a quarter and a $661 plan within a year. nothing about their usage of the product has changed except that more people are voting on their boards.

the bill goes up because growth happened. the product didn't get better. the team didn't get more value. there is no per-tracked-user infrastructure cost that justifies a 4x markup at the 500-user threshold. the pricing is structured this way because it's the most lucrative way to extract willingness-to-pay from a growing customer.

why this is bad for everyone

the case against tracked-user pricing isn't that it's expensive (it is, but that's a different argument). it's that it misaligns the customer and the vendor.

the vendor's incentive is to not improve the dashboards' performance with high traffic, because high traffic is what generates the bill. the vendor's incentive is to count "tracked users" as broadly as possible, because broader counts mean higher tiers. the vendor's incentive is to obscure the cost of growth at signup, because honest disclosure would cost them sign-ups.

the customer's incentive is to throttle their own product. don't share the board too widely. don't email customers about it. don't link it from the public roadmap. all of those would generate tracked users. all of them are exactly what you want feedback boards to do.

we've been on the customer side of this with three different metered tools (analytics, error tracking, customer support). every time, the team eventually starts performing the choreography of "how do we keep our usage in this tier." that's a tax on growth, paid in calendar time.

what plausible got right

plausible analytics' pricing is the obvious counter-example, and it's the model we copied. plausible meters on pageviews, but the price scales gradually and predictably. you don't 4x at the 500-pageview mark. their tiers also don't change what features you have. you get the same product on the entry tier as on the highest tier. you just get more headroom.

supabase does the same thing for postgres-as-a-service. small projects pay $25, large projects pay more, but the api is the same product end-to-end. linear flattened seat pricing for years before raising it. every tool that explicitly markets to indie teams has done some version of this.

we wanted spirby to feel the same. flat $19 entry. higher tiers (planned post-launch) raise headroom (board count, vote count, team size) but don't change which features are available. the api is on the entry tier. webhooks are on the entry tier. custom domains are on the entry tier. growth doesn't get punished.

what we give up

flat pricing isn't a free win. there are real trade-offs we've absorbed.

we leave money on the table. a team with 100,000 tracked users on canny is paying canny $1000+/mo. that team would also probably pay us more than $19/mo. we don't have a sales-ops layer to extract that value, and we're not going to build one for v1. that revenue goes to canny.

we're more sensitive to runaway accounts. if someone uses spirby for a 10-million-vote launch on the entry tier, we eat the cost. we plan for this with rate limits (1000 req/hour per api key) and queue workers, but the math has to work even at the long tail. that constrains how aggressively we can spend on infrastructure compared to a metered competitor.

we have to be honest about feature gaps. canny can absorb a feature gap because the customer is paying $300+ either way. we can't. on $19/mo, we have to ship the product that's actually worth $19/mo on day one, not in a roadmap promise.

we're fine with each of those trades. they pull the product back toward "obsessive about the small team's experience," which is where we want it.

the fairness argument

the strongest argument we've heard for tracked-user pricing is that it's "fair" because it scales cost with usage. teams using more should pay more.

we don't buy this. the cost of running a feedback board does not scale linearly with tracked users. the marginal cost of the 501st tracked user to canny is approximately zero. the bill jumps because the customer's willingness-to-pay went up, not because the cost went up.

"fair" pricing aligns price with cost. tracked-user pricing aligns price with extraction. the words "fair" and "extractive" are doing different jobs and shouldn't be confused.

a fair model: the customer pays a price related to the value the product provides. flat pricing approximates this badly (a heavy user gets a deal). tracked-user pricing approximates it backwards (growth-on-the-vendor's-side gets bigger bills). plausible's model (gradual, predictable scaling) approximates it well.

we picked the cleanest of the three because we'd rather a customer who joined at $19 still feels like they're at $19 a year later. that doesn't feel like a limitation. it feels like a feature.

the bet

every feedback tool has to decide whether it's a vendor that grows revenue per customer over time or a vendor that grows by adding customers. canny is on the first track. spirby is betting on the second.

it's a slower business. it's also less likely to get acquired by a private-equity firm and immediately pivot to enterprise sales. that's the whole point.

if a flat $19/mo with full api access on day one sounds like the kind of thing you'd build a workflow on top of, start a trial. the price won't change because you grow. that's the entire pitch.