# Switching costs *Embed deeper than any demo can reach.* --- You've built a better product. Cloud-native architecture, modern interface, open APIs. The incumbent's system looks like it was designed in 1995, because it was. You schedule a demo with one of their largest customers, a community bank that's been running the same core platform for twelve years. The demo goes well. The CTO is engaged, the operations team is impressed, the procurement team requests a proposal. You start scoping the migration. --- The first call reveals the core banking system: every deposit, withdrawal, loan, and payment flows through it. You've built migration tooling for that. But core banking is the first layer. Over twelve years, this bank added payments processing, then digital banking, then lending, then risk management, then compliance reporting. Each module was configured by people who understood this bank's specific needs: which branch managers need approval oversight, how the board expects its reports formatted, what the regulators require in quarterly submissions. More than 950 fintech partners integrate through the incumbent's APIs, and every integration is a dependency your migration plan needs to account for. Your estimate started at three months. After the scoping calls it's closer to twelve, with a parallel-run period where both systems operate simultaneously while the bank prays nothing breaks. A failed core conversion can destroy a bank, and everyone in the industry knows the horror stories. The CTO who loved your demo stops returning calls. The switching cost was never in the contract. It was in twelve years of configuration encoding how this specific bank operates, built up gradually, impossible to recreate from a feature comparison. --- About 200 core processing contracts get bid annually across the US. Half stay with their current provider without looking. Of the half that do look, most move between three incumbents. [[Jack Henry]] retains 99% of its customers each year. Contracts run five to seven years, but the contracts matter less than what customers have built on top of the platform. Cross-sell is as much a retention strategy as a growth strategy. The more services a bank runs on the system, the more the system mirrors the bank's operating knowledge, and the less anyone else's feature list matters. --- Jack Henry's depth accumulated over fifty years. Other companies engineer the expansion from the start. In 2007, a founder left Salesforce to build CRM for life sciences. Pharmaceutical sales reps operate under constraints generic CRM can't handle: regulated sample tracking, physician-level territory management, compliance-driven call reporting. By 2012, the product had 80% of the pharmaceutical CRM market. CRM was a beachhead. In 2011, the next product launched: content management for regulated industries, built on its own technology stack. Clinical operations followed, then regulatory affairs, then quality management. Each product entered a new department within the same customer. A pharmaceutical company using the CRM for sales discovered the vendor also served their clinical team, then regulatory, then quality. [[Veeva]] now offers eight application families and serves over 1,400 customers. Net revenue retention exceeds 120%. A company running one application might evaluate alternatives at renewal. A company running six has built its operational backbone on the platform. Clinical trial data, regulatory submission histories, validated quality processes, years of work stored in Veeva-specific formats and referenced by auditors. Replacing one application is a project. Replacing six is an organisational reconstruction. --- This is the dynamic [[Constellation]] recognised at industrial scale. [[Verticals|Vertical markets]] are too small for large competitors to enter but deep enough to support one dominant player. A transit scheduling system doesn't compete with Salesforce. A marina management platform doesn't compete with Oracle. They compete with two or three alternatives, and the deeper customers embed over time, the wider the gap between technically possible to switch and practically worth switching. Constellation has bought over a thousand of these businesses. Niche vertical, embedded workflows, high switching costs, recurring revenue. What Jack Henry does in community banking and Veeva does in life sciences, Constellation's portfolio does across club management, transit operations, dealership systems, and dozens of sectors most people have never heard of. Mark Leonard found a type of business with these characteristics and bought hundreds of them. --- Customers shape the software around how they work. Configuration accumulates over years: approval routing, compliance templates, report formats, integration logic, trained habits. Each adjustment is small. The aggregate is a system that mirrors the organisation's operating knowledge, built up one decision at a time, impossible to recreate from a spec sheet. A competitor can demo a better interface. They can't demo the twelve years underneath it. ---