One Premium tier, all features: why we don't do upsell ladders

April 13, 2026·4 min read·The Kovra team

Look at the pricing page of any major Discord bot. You'll see four tiers. Free, Premium, Pro, and "Custom / Contact Us". Each tier adds a handful of features and raises limits. The best features live in Pro or Custom. You're expected to start at Premium and upsell yourself up the ladder.

This is a ladder because ladders convert. Every rung adds one or two must-have features and most customers can't resist climbing. It works well for the bot company. It works badly for the customer, who ends up paying $30/mo for features they use once a month and still hits a cap on the ones they use daily.

We rejected the ladder. Kovra has one paid tier.

What one tier looks like

Premium is $5.99/mo, $59/yr, or $99 lifetime. All three get the same 15 gates. There is no Pro tier above Premium. There is no "Business" or "Enterprise." Agency is an add-on, not a higher tier — it strips branding for client servers and sits on top of any Premium sub.

The 15 gates are:

  • Kovra Custom — rename the bot, change its avatar, pick an accent colour.
  • Social alerts — YouTube, Twitch, Bluesky, Kick, RSS, Podcast.
  • Invite tracker — leaderboard with fake/left attribution.
  • Retention curve — 1d/7d/30d/90d cohort retention analytics.
  • Config import/export — clone your setup across servers.
  • Caps lifted — custom commands (20 → ∞), auto-responders (5 → ∞), scheduled messages (3 → ∞), voice hubs (1 → 5), role rewards (15 → ∞), active giveaways (1 → ∞).
  • Deeper retention — cases 7d → 30d, full retention analytics.
  • Better tickets — HTML transcripts replace plaintext.

Why this works for us

A simple pricing page is a marketing advantage. A customer who lands on our pricing page spends 30 seconds deciding vs. five minutes comparing tier tables. Every pricing study we ran showed simpler = higher conversion, because choice fatigue is real.

A simple gate matrix is an engineering advantage. Every feature has the same boolean check: guildHasPremium(guildId). No "is this a Pro feature or a Business feature" tier-checking logic scattered across the codebase. Adding a new feature takes one line of gate check, not a spreadsheet of "which tier gets this."

When we'd add a tier

We won't add a higher tier for these reasons:

  • "We ran out of features to add to Premium." This is never true.
  • "Some customers will pay more." Probably, but that's their problem, not ours.
  • "Our competitors have four tiers." Not a reason.

We'd add a tier for these reasons:

  • Volume workloads that cost real money. If a server has 500 000+ members and is hammering our infra, we'll eventually need a custom tier that covers the hosting cost. But we'll negotiate it per-customer, not via a self-serve Pro tier.
  • White-label reselling. Agencies running Kovra for dozens of clients may want a reseller arrangement with bulk discounts + support SLAs. We'll build that as a partnership program, not as a ladder rung.

Feedback we get

Two kinds of customer push back on this pricing. The first says "$5.99 is too cheap — you'll never make money." We appreciate the concern. The math works out because we keep engineering lean and don't carry sales staff.

The second says "I only want one feature — can I get a cheaper plan?" The answer is no. We'd rather lose the $3/mo customer than build a second cheaper tier and complicate the gate matrix. The free tier is genuinely useful; if it's not enough, $5.99 is the price of the whole stack.

One tier, all features, cancel any time. That's the deal.