# Switching Cost Design The deliberate engineering of workflow embedding such that the cost of leaving the product exceeds the benefit of an alternative. One of the most durable [[7 Powers]] for software and AI companies. ## Dimensions of switching cost - **Data** — customer data inside the product (history, configuration, training data, artefacts). The deeper the history, the higher the cost to recreate. - **Workflow** — integrations into the customer's daily operating rhythm. Every process that runs through the product is a line that would need to be re-stitched elsewhere. - **People** — trained operators whose muscle memory is tied to this product's UI, its conventions, its shortcuts. - **Network** — other users inside or outside the organisation who already interact with the product. - **Audit / compliance** — historical records tied to this vendor for regulatory purposes. ## Test > "What happens if the customer stops using the product next Monday?" - **Weak answer:** "They'd switch to a competitor in a week." - **Strong answer:** "They'd lose X months of training data, break Y integrations, and retrain Z operators." ## Design principles - Own the **source of truth**, not just a view over someone else's. - Generate **artefacts** customers keep returning to (reports, dashboards, playbooks) rather than ephemeral outputs. - Embed in **multi-person workflows** — solo-user products switch cheaply; team products don't. - Accumulate **history** that no competitor can hand-wave into existence. Related: [[7 Powers]], [[AI era Defensibility]], [[Defensibility Principles MOC]], [[AI Agent Vertical SaaS DD MOC]] --- Tags: #defensibility #productstrategy #firstprinciple