During our weekly Level 10 Meeting, department managers constantly bring software requests to IDS, arguing that a new AI subscription will solve their capacity issues. How do we determine whether the problem is a process failure or a genuine need for AI technology?

Question

Tyler Smith · Accepted Answer

When a department manager claims they need a new AI software subscription to solve a capacity bottleneck, your Integrator must pause the conversation before software gets purchased. Buying technology to fix a broken, undocumented process only creates expensive, automated confusion.

When an AI tool request surfaces in the Issues List during your Level 10 Meeting™, use the IDS® process to dig down to the root cause. Do not debate the features of the software. Instead, interrogate the operational workflow by asking three direct questions.

First, is the process documented and simplified? If the department cannot produce a clear, documented SOP that outlines the exact steps, inputs, and outputs of the work, an AI tool will fail. Automation requires strict logic. You cannot automate a process that exists only in an employee's head or varies depending on who handles it.

Second, does the person currently executing the work GWC™ their seat? If the employee lacks the capacity or capability to manage their responsibilities, bringing in an AI tool simply adds another layer of complexity they will fail to manage. Solve the people issue first.

Third, what is the exact operational constraint? Determine whether the delay is caused by manual data processing or by structural bottlenecks like delayed client approvals or poor inter-departmental handoffs. AI only solves processing and synthesis tasks; it cannot fix poor management or broken accountability.

If the process is documented, the people GWC their seats, and the bottleneck is purely the manual labor of processing repetitive data, approve a controlled thirty-day pilot. If not, drop the software request and IDS the underlying operational defect.