The Subscription Trap
Most open banking startups price like Stripe. That's wrong. Stripe's unit economics assume infinite API call scaling. Open banking APIs have regulatory audit costs. Data governance scales linearly, not logarithmically.
Subscription model math: ₹50K/month customer churns at 15% monthly. You need 7 new customers monthly just to stay flat. Customer acquisition cost in India for fintech B2B is ₹2-4L. You're underwater for 5-8 months.
Worse, subscription creates pressure to add features. More features mean more compliance review cycles. More compliance cycles delay product launches. Cycle time stretches to 90+ days. Competitors using usage models iterate faster.
Why Usage-Based Actually Works
Usage-based pricing mirrors what banks actually charge for data access. NPCI charges per transaction. RBI compliance costs scale with data volume accessed. Your pricing should too.
A B2B lender using open banking APIs accesses customer data 100-500 times per loan decision. One subscription tier covers 50 decisions. Next tier covers 500. Customer pays for what they consume, not what they might use.
Churn dynamics flip. Customer stops paying when they stop accessing data. You stop serving when they stop paying. No prolonged dead weight. Clean unit economics.
Netflix tried this in 2011. Netflix abandoned usage-based and went subscription. But Netflix had content scarcity. Open banking has data scarcity—the opposite. Your analogy should invert Netflix's lesson, not copy it.
The NRR Problem
Subscription-model fintechs in India report NRR of 105-115%. That looks respectable. But it's misleading.
NRR is calculated as (ARPU_new - ARPU_lost) / ARPU_start. In fintech, when a customer grows, they often build internal capabilities or negotiate direct bank partnerships. They don't pay you more; they pay you less. Negative dollar churn is common by month 18.
Usage-based models hit 125-140% NRR naturally. A lender grows loan volume 40%. They access data 40% more. Billing scales automatically. No negotiation. No churn negotiation, either—hard to argue with usage-based bills.
Subscription Tier Design Error
You'll be tempted to create "Starter, Pro, Enterprise" tiers. This is operational debt in India.
Tier misalignment happens at month 4-6. Customer is on Starter (₹25K/month), hits API limits, needs Pro (₹75K/month). They resent the jump. They build their own integration instead. You lose them entirely.
Usage-based sidesteps this. Customer hits 10K calls in April, 15K calls in May. Bill scales. No cliff. No negotiation call. Done.
Real Numbers: Churn Benchmarks
B2B SaaS churn in India sits at 5-7% monthly for software businesses. Fintech B2B churn is 12-15% monthly. Why? Regulatory changes, direct bank partnerships, and internal builds.
Open banking adds another layer: open banking APIs themselves are new. Customers pilot, then pivot to building in-house. Expect 18-22% monthly churn on subscription models at scale.
Usage-based models show 3-5% monthly churn for cohorts above ₹10L ARR. Why the drop? Usage naturally correlates to business health. If a customer's business sinks, API calls drop. Churning happens gracefully, not suddenly.
Timing: The India Stack Window
Account Aggregator mandate is 2025-2026. Every bank must expose APIs. Supply is unlimited. Pricing power evaporates.
Subscription models can't survive in unlimited supply. Only usage-based pricing survives because cost is variable, not fixed.
Your window to lock subscription customers closed in 2023. Now you're competing on margins. Usage-based pricing is the only way to compete and build 120%+ NRR simultaneously.
The Founder Implication
If you're building open banking infrastructure today, price by usage. Not API calls—value delivered. Link pricing to loan amount approved, customer acquired, fraud prevented.
Subscription models will consolidate into 2-3 players. Usage-based winners will emerge from the long tail. The long tail will have better margins by year 3.