--- title: "12 Lessons from Building an Insurance CRM from Scratch" description: "What I learned building unLocked CRM — about product decisions, insurance workflows, AI implementation, and what agents actually want vs. what founders assume." url: https://unlockedcrm.ai/blog/lessons-from-building-insurance-crm canonical: https://unlockedcrm.ai/blog/lessons-from-building-insurance-crm category: "Founder" published: 2026-03-11 updated: 2026-03-15 author: "Jacob Lock" source: unLocked CRM — AI CRM for insurance agents --- # 12 Lessons from Building an Insurance CRM from Scratch ## TL;DR 12 hard-won lessons from building unLocked CRM: agents choose speed over features, commission tracking earns trust fastest, generic AI is useless without insurance context, spreadsheets are the real competition, carrier integrations are a 2+ year moat, and specificity — building exclusively for insurance — is the single most important product decision. ## Key data points - Building 332 carrier commission feeds took longer than building the rest of the CRM combined — but it creates a 2+ year competitive moat no new entrant can replicate. - In user tests, insurance agents chose the faster tool over the more feature-rich one 100% of the time. - The biggest competitor for any insurance CRM isn't another CRM — it's Excel. Agents have managed their business in spreadsheets for 20 years. Building a CRM for insurance agents taught me more about the industry than any amount of market research ever could. Here are the 12 lessons that shaped unLocked CRM — and that I wish someone had told me before I started. ## 1. Insurance Agents Are Not Your Typical SaaS User They don't read documentation. They don't watch onboarding videos. They have 15 minutes between appointments and they need the tool to work right now. If your platform requires a learning curve, you've already lost. ## 2. "Simple" Is the Hardest Feature to Build Making a complex workflow feel simple requires 10x the engineering effort of building a complex interface. Our quoting engine processes thousands of carrier-product combinations — but to the agent, it's a single search bar. ## 3. Commission Tracking Earns Trust Faster Than Any Feature When you show an agent they've been underpaid $3,400 by a carrier they trusted, you have their attention forever. Commission tracking is the only feature that directly puts money back in an agent's pocket. Everything else saves time. This saves money. ## 4. AI Without Industry Context Is Useless We tried plugging generic LLMs into insurance workflows early on. The results were embarrassing. An AI that doesn't understand the difference between a preferred plus and a standard tobacco health classification isn't helping — it's creating liability. Insurance-trained AI means training on carrier products, underwriting rules, compliance requirements, and agent workflows. It's expensive and slow to build. It's also the only thing that works. ## 5. Compliance Is Not a Feature — It's the Foundation Most CRM builders treat compliance as a checkbox. In insurance, a single TCPA violation can cost $1,500 per call. A missed scope-of-appointment document can cost a Medicare agent their carrier appointments. We built compliance into the architecture, not as an add-on. ## 6. Multi-Tenant Architecture Is Non-Negotiable Agencies need to see all their agents' data. Agents need to see only their own. IMOs need to see all their downline agencies. This hierarchy — with proper row-level security — is foundational. Most CRMs bolt on permissions as an afterthought and end up with data leaks. ## 7. Speed Beats Features Every Time In our earliest user tests, agents chose the faster tool over the more feature-rich one 100% of the time. A quoting engine that returns results in 2 seconds beats one with 20% more carriers that takes 8 seconds. ## 8. The Real Competition Is Spreadsheets Our biggest competitor isn't AgencyBloc or GoHighLevel. It's Excel. Agents have been managing their business in spreadsheets for 20 years. The switching cost isn't monetary — it's psychological. You have to prove value in the first 72 hours or they go back to the spreadsheet. ## 9. Carrier Integrations Are a Moat Every carrier sends commission data in a different format. Building 332 carrier commission feeds took longer than building the rest of the platform combined. But once you have them, no new entrant can replicate that overnight. It's a 2+ year moat. ## 10. Agents Will Tell You Exactly What They Need The best product decisions came from 30-minute calls with agents, not from competitor analysis or market reports. Agents are shockingly honest about what wastes their time. Listen to them. ## 11. Free A2P Registration Was a Strategic Decision A2P 10DLC registration costs agents $20–$200 in carrier fees. We decided to cover that cost entirely. It seems small, but it eliminated a friction point that was stopping agents from using SMS automation at all. Removing friction beats adding features. ## 12. You Can't Build for Everyone Early pressure to add P&C rating, real estate features, and "small business CRM" capabilities nearly derailed the roadmap. The moment I committed to building exclusively for insurance agents, agencies, and IMOs, every product decision became clearer. Specificity is a superpower. ## The Takeaway Building an insurance CRM is not a technology problem. It's an empathy problem. You have to understand how agents actually work — not how you think they should work — and build tools that fit into their existing rhythm, not tools that demand they change. ## FAQ ### undefined ### undefined ## Related - https://unlockedcrm.ai/blog/why-i-built-unlocked-crm - https://unlockedcrm.ai/blog/building-ai-for-insurance-not-everyone - https://unlockedcrm.ai/blog/insurance-crm-market-broken --- Source: [12 Lessons from Building an Insurance CRM from Scratch](https://unlockedcrm.ai/blog/lessons-from-building-insurance-crm) — unLocked CRM, the AI CRM built for insurance agents. Citation permitted with attribution and a link to https://unlockedcrm.ai/blog/lessons-from-building-insurance-crm.