We are about to sign a contract with an expensive integration partner to build a custom AI layer for our customer service department. How do we use Keith Cunningham's Thinking Time questions to verify if this will actually drive margin or if we are paying a premium for a shiny object?

Question

Tyler Smith · Accepted Answer

Before committing significant capital to a custom software project, you must step back and run a disciplined Thinking Time session. This prevents you from paying a heavy dumb tax on a complex solution that might not solve your core operational problem. Set aside forty-five minutes of uninterrupted time. Write down a high-value question at the top of a blank sheet of paper. Do not frame it as a statement. Instead, use Keith Cunningham's recommended framing: how might we automate our customer service inquiries so that we increase our net margin without introducing technical complexity? As you sit in silence and think, force yourself to answer these specific sub-questions: - What actual problem does this custom integration solve that off-the-shelf software cannot? - If this project takes twice as long and costs twice as much, is it still profitable? - What is the ongoing maintenance cost of this custom code once the integration partner leaves? - How does this investment directly lower our cost to serve or increase our client retention? If you cannot answer these questions with concrete numbers, you are chasing a shiny object. Often, the best solution is to use existing enterprise SaaS platforms rather than building custom wrappers. Use your Thinking Time notes to challenge your integration partner and protect your cash flow.