Growth and monetization
Good-better-best packaging
Three tiers so buyers self-segment. Gate admin and scale, not the aha.
Packaging is how you bundle capabilities into plans. The classic good / better / best ladder lets buyers self-segment:
- Good: core job for individuals or small teams; cheap enough to start.
- Better: the plan you want most customers on; collaboration + the features that drive retention.
- Best: enterprise administration, security, support, and scale limits.
Feature gating should push upgrades along a real pain ("need SSO for IT," "hit usage limit," "need shared workspaces"), not arbitrary locks that break the core job on the free/low tier. Gate admin and scale; avoid gating the aha moment if you run a product-led motion.
Add-ons (extra modules, premium support, dedicated capacity) let large customers expand without forcing every mid-market buyer through a custom SKU explosion. Prefer three core tiers + a few add-ons. Too many SKUs and neither sales nor customers can explain the difference; engineering drowns in entitlement matrix bugs.
Walking through Pipekit
You build Pipekit, an API for background job workflows. Value metric: workflow steps per month, plus seats for the dashboard.
| Free | Pro ($49/mo) | Business ($299/mo) | Enterprise | |
|---|---|---|---|---|
| Steps / mo | 5,000 | 100,000 | 1,000,000 | Custom |
| Seats | 1 | 5 | 25 | Unlimited |
| Retention / concurrency | Basic | Higher limits | Priority | Custom SLAs |
| SSO / audit logs | — | — | SSO | SSO + audit + VPC |
| Support | Community | Slack connect | Dedicated |
Self-serve checkout for Free/Pro/Business; Enterprise via sales. Soft warnings at 80% and 100% of step quota; overage on Pro/Business at a published per-1,000-steps rate so power users expand without a forced plan jump.
A startup averaging 70,000 steps/mo and 4 seats fits Pro. They grow to 400,000 steps → Business, or stay on Pro with overage if spike-y. An enterprise security review needs SSO + audit → Business or Enterprise. Indie user at 2,000 steps stays Free until they feel the cap — the free tier still lets them hit the aha (first successful job run) without a card.
Store entitlements as data (max_steps, sso_enabled, seat_limit) rather than if plan == 'pro' in every service. Packaging then becomes a config change with a migration, not a rewrite.
Common mistakes
- Gating the aha. If the free/trial plan cannot deliver the job, PLG dies.
- Too many SKUs. Nobody can explain the difference.
- Hard-coding plan names in every service. Centralize
can(workspace, 'sso').
Check yourself
Free tier cannot run the first successful job. What's wrong? You gated the aha. The low tier must complete the core job.
Why not a dozen add-on SKUs for every enterprise checkbox? Sales cannot explain them and engineering drowns in entitlement bugs. Three tiers plus a few add-ons.
Key takeaways
- Good-better-best lets buyers self-segment. Put most customers on Better.
- Gate admin and scale, not the aha. Add-ons beat SKU explosion.
- Entitlements as data, not scattered
if plan ===checks.
Further learning
Validate with interviews across segments, then ship and revisit quarterly. Next: freemium vs trials.