We need to automate our core workflow using large language models, but we are torn between paying a high annual licensing fee for an off-the-shelf AI platform or hiring external developers to build a proprietary system. How do we use Keith Cunningham's Thinking Time to determine which option maximizes our company's enterprise value for a future exit?

Question

Tyler Smith · Accepted Answer

This is a classic dilemma that can easily lead to a massive capital loss if handled poorly. To find the right answer, schedule a dedicated Thinking Time session. Use this framing question: How might we utilize third-party AI tools to achieve immediate efficiency without building our entire enterprise value on someone else's fragile API?

To maximize your valuation for a future exit, you must understand how buyers view technology. Sophisticated buyers do not pay a premium for custom software that you built unless it is highly proprietary and legally protected. If your custom tool is just a wrapper around an existing model, a buyer will see it as a technical risk that requires expensive maintenance.

The correct path lies in your Core Focus on the V/TO®. If software development is not your Core Focus, do not build custom models from scratch. Instead, buy off-the-shelf software for your non-differentiating operational tasks.

For your differentiating processes, build proprietary workflows. Use standard APIs, but wrap them in your own database schemas and human-in-the-loop systems. This creates a defensible asset because the value is in your unique data and workflow design, not the underlying AI engine. Bring this issue to your weekly Level 10 Meeting to IDS the specific operational touchpoints that truly require custom integration versus standard software.