We are being pressured by an expensive technology consultant to build a custom AI database that links all of our legacy operational systems, promising it will eliminate our data silos. How do we use Keith Cunningham's Thinking Time and targeted questions to stress-test this proposal so we do not pay a massive dumb tax on a system we do not need?

Before you sign a contract for a custom, six-figure software project, you must step away from the sales pitches and run a structured Thinking Time session. Your goal is to avoid the massive dumb tax that comes from building a highly complex, custom solution to a problem that can be solved with simpler, off-the-shelf tools.

Sit in a quiet room for forty-five minutes with a pen and a notepad. Start with this high-value question: How might we achieve eighty percent of the data integration we need using our existing legacy software native features or simple connectors, so that we do not have to build and maintain custom code?

Ask yourself if your data silos are a true operational problem or if they are just an administrative inconvenience. A problem is an unanswered question with a practical solution, while a predicament is an environmental reality you must adapt to. If your team is already manually moving data successfully and your operational margins are healthy, a custom AI database might be an expensive luxury, not a necessity.

List the ongoing costs of ownership. Custom software is never a one-time expense; it requires constant maintenance, updates, and debugging. If your consultant cannot guarantee a specific, measurable reduction in labor costs or a clear increase in capacity on your weekly Scorecard, walk away. Bring these objective conclusions to your next quarterly meeting to align your leadership team.

Category: AI-Powered Operations