Our cheapest quarter of growth, by far, was one where we signed almost nobody new. Expansion from entrenched accounts carried the entire number. That quarter changed my job description in my head: retention stopped being a support metric and became switching-cost engineering. The five barriers below are the whole system.
Acquiring a customer costs five times more than keeping one, so the highest-leverage retention work is making departure expensive. Not through contracts and threats, but through value structures that make leaving genuinely painful: data gravity, embedded workflows, and team-wide dependence. Platform dominance is the endgame; switching costs are how you pave the road there.
There is an ethical line here, and crossing it backfires: hostage tactics create resentful customers who leave loudly. The goal is earned lock-in, barriers that exist because the product became load-bearing, not because legal trapped anyone.
The 5 Earned Exit Barriers
- Data gravity. Every report, dashboard, and historical record stored in your system raises the cost of leaving. Accelerate it: one-click imports, unlimited history retention, analytics that get smarter with age. Contrast with artificial lock-in that blocks exports, earned gravity lets data leave in theory while making it pointless in practice.
- Workflow embedding. Count your integrations per account. Each connected tool, Slack alerts, Salesforce sync, Zapier flows, is a thread tying you into operations. Products with 5+ integrations churn at a fraction of single-integration accounts. Your onboarding goal is not activation; it is entanglement.
- Team penetration. Single-user tools get cancelled in budget reviews; tools eleven people open daily survive. Drive multi-seat adoption deliberately: shared workspaces, @mentions, collaborative artifacts. Every additional weekly-active user cuts churn probability measurably.
- Skill investment. Certifications, power-user shortcuts, custom templates, when users invest learning, leaving means admitting the investment was wasted. This is the sunk-cost trap working for you, and it is why education programs are retention programs in disguise.
- Compounding outputs. AI models trained on customer data, benchmarks built from their history, automations refined over months. If the product gets better the longer it is used, tenure itself is the moat, a new vendor starts at zero while you start at three years of tuning.
Measuring Your Moat
Build a switching-cost score per account: integrations count, WAU seats, data volume and age, custom objects, API calls per week. Watch it predict churn better than NPS. Accounts below threshold get entanglement campaigns (integration prompts, team-invite nudges); accounts above it get expansion offers, because entrenched customers buy more.
The Line
Earned barriers survive procurement reviews and angry CFOs; artificial ones (throttling exits, hiding export buttons, degrading the product to force upgrades) generate G2 reviews titled “hostage situation.” Rule of thumb: if you would be embarrassed explaining the mechanism on a sales call, it is not a moat, it is a liability with a timer.
Earned lock-in survives procurement reviews. Hostage tactics do not.
Frequently Asked Questions
How do you measure switching costs per account?
Score integrations count, weekly-active seats, data volume and age, custom objects, and API usage. Composite scores predict churn better than NPS and tell you exactly which accounts need entanglement campaigns.
What is the fastest way to raise exit barriers ethically?
Drive multi-seat adoption and integrations in the first ninety days: shared workspaces, @mentions, connected tools. Early entanglement compounds; late upsells to entrenched accounts feel like rewards.
When does lock-in become a liability?
When barriers punish rather than serve: blocked exports, degraded downgrades, hostage data. Earned gravity survives procurement; artificial cages generate the reviews that kill you.
Continue Reading: Platform Power & Lock-in
This article is part of our series on Platform Dominance: Moving From Tool to Indispensable Ecosystem. Related reading: