The Dealer Relationship Trap
You start by saying yes to every dealer. Makes sense—you need volume. You integrate with their CRM, you don't enforce data standards, you let them skip verification steps. It feels collaborative.
By year two, you have 800 dealers. But they've built habits. They expect you to bend. You ask for standardized customer data—suddenly you're the villain. They start using competitors who ask less.
You're now trapped between two bad options: enforce standards and lose dealers, or stay loose and your data becomes unusable. A dealer that's been with you 18 months is much harder to fix than a dealer you onboard clean.
The compounding cost? You can't model demand properly. You can't forecast inventory. You can't build reliable underwriting. Three years in, you're still fighting data quality while competitors who were stricter in year one are moving to B2B2C models.
Data Fragmentation You Didn't Know You Caused
In year one, you integrate with dealer systems loosely. You get vehicle registration data, you get some service records, you get some financing details. It's messy but it works for your immediate use case—maybe loan approvals or insurance quotes.
Now scale to 200,000 vehicles. The same vehicle has three different registration dates across your systems. Customer names are spelled four ways. Service records exist for 41% of your vehicles, 67% for another dealer network.
This isn't a data engineering problem anymore. It's a network problem. Dealers now have no incentive to give you clean data because you've never enforced it. You've trained them that messy is acceptable.
Think of it like credit bureau data in 1995—early compilers accepted whatever lenders gave them, and spent a decade unwinding bad data standards. Auto-tech in India is there now, except the fragmentation happens faster.
By the time you realize it, your unit economics on B2B products (financing, insurance, fleet management) are broken because you can't reliably match a vehicle across transactions.
The CAC Trap Into Unprofitable Segments
You optimize CAC in year one. You find that B2C used car buyers are your cheapest customer acquisition—maybe Rs. 800-1,200 per customer. So you lean into that channel.
By year two, 60% of your revenue is from that segment. It feels great—you've found product-market fit.
Then you hit a wall. Used car buyers are price-sensitive and low-frequency. LTV is Rs. 3,000-4,500 per customer. Your actual payback period is 14-18 months. You need constant new volume just to stay flat.
Meanwhile, the segment you deprioritized—fleet operators buying new vehicles—has worse CAC (Rs. 2,500-3,500) but LTV of Rs. 25,000-40,000 and 3-4 repeat purchases. Payback is 4-6 months.
Year three: you're profitable only on paper. You're reinvesting 100% of revenue into CAC because you're addicted to the cheap B2C channel. Switching now means admitting you optimized for the wrong metric.
This isn't random—it's baked into how you built your product, your team's compensation, and your dealer relationships. All because in month 3 of year one, B2C looked easiest.
The India Stack Timing Problem
DiGi-Locker went live in 2015. PAN-Aadhaar linkage accelerated post-2019. Vehicle registration APIs opened up in fragments across states from 2020 onwards.
If you started an auto-tech company in 2018, you built workarounds for data you can now access directly. Your workarounds—dealer relationships based on gatekeeping, manual verification processes—became your moat by accident.
Now that the data is available, your moat is a liability. You're overly dependent on dealers who no longer have exclusive information. Newer competitors can integrate with RTO data and Setu APIs without needing dealer goodwill.
You spent year one solving a problem that stopped existing in year three. And you can't unfund the relationships you built to solve it.
What This Means for Year One
Start with standards, not convenience. Make dealer onboarding harder, not easier. Eat the short-term volume hit—your year-one sales will be 20% lower, but your year-three margins will be 300% better.
Build data quality into product, not post-launch. If you're not rejecting messy data in month 2, you'll be cleaning it in year 4.
Be honest about LTV in month 6. Don't optimize CAC in a segment just because it's cheap. Calculate payback period, repeat purchase rate, and margin by segment. Bias toward complexity if the unit economics work.
These aren't sexy moves. They won't make year-one board meetings exciting. But by year three, when your competitor is paralyzed by dealer revolt and data chaos, you'll be the only one who can actually scale.