Markets and business models
Developer land, enterprise expand
Self-serve API key first, then usage growth, then procurement, commits and expansion. The product must serve both ends.
The canonical usage-business motion (Twilio, Stripe, Snowflake) is developer land, enterprise expand. A credit card and an API key beat a six-month sales cycle for the first wedge. Later, procurement wakes up at a spend threshold. Your product and packaging must support both ends.
Concept
Land is the first paid (or seriously used) foothold — often a developer on a credit card. Expand is more usage, more products, more regions, or a commit. See usage-based pricing.
The four steps
- Developer discovers docs, signs up, ships a feature on a credit card.
- Usage grows with the customer's product traffic.
- Procurement wakes up at some spend threshold — security review, MSA, volume discount, commit.
- Sales and success attach to expand: new products, new regions, higher support tiers.
Pricing pages that only speak "contact sales" kill the land. Pricing that only speaks "pay as you go" leaves enterprise money on the table. You need delightful self-serve and enterprise controls: SSO, audit logs, commit pricing, invoices.
Snowflake-style credits and cloud commit contracts turn volatile usage into something CFOs can budget — often trading discount for commitment. See commits and credits.
Walking through a threshold
A team spends $200/month on your API from a founder's card. Traffic grows. At $4k/month, finance notices. At $12k/month, security asks for SSO and a DPA. At $40k/month, they want a 12-month commit at a lower unit price. If you have no commit SKU, no audit logs and no invoice flow, that expansion stalls — or they leave for a vendor who looks "enterprise" at the same usage.
The opposite failure: you hid the API behind "talk to sales" and never got the 1,000 developer lands that would have become those $40k accounts.
A healthy book often shows a power law: a minority of accounts drive most usage. Enterprise success on the whales, self-serve tooling for the long tail.
Common mistakes
- Sales selling seats while COGS are usage — misaligned quota.
- Contact-sales-only pricing. You never get the developer land.
- Pay-as-you-go only. CFOs cannot budget, so enterprise spend never commits.
Check yourself
Why do usage companies need both a credit-card checkout and a commit SKU?
Checkout lands developers. Commits let finance budget and unlock enterprise expansion.
A customer's spend jumps from $200/month to $12k/month. What usually happens next?
Procurement and security wake up. They want SSO, a contract and a volume discount — not another credit-card receipt.
Key takeaways
- Path is self-serve land → usage growth → enterprise commits and expansion.
- Product must serve both ends: DX for land, controls and commits for expand.
- A minority of accounts usually drive most usage. Staff success accordingly.
- Forced commits without usage history stall deals. Earn the threshold first.
Further learning
Next, the AI-specific trap: wrapper economics.