Choosing your first AI use case
Pick a use case that scores high on all three points of the Golden Triangle: high pain (felt daily), low complexity (achievable in 90 days), and clear ROI (a metric you can measure and move). Skip anything that misses one.
The Golden Triangle
The Golden Triangle screens use cases on high pain, low complexity, and clear ROI. A use case must score well on all three; strong on two but weak on one is how pilots end up in the 95% that fail.
High pain means people feel the problem every day, so adoption is pulled rather than pushed. Low complexity means it is genuinely achievable in 90 days with the data you already have. Clear ROI means there is a specific metric, handle time, conversion, error rate, that you can baseline today and prove you moved. Use cases that miss any one corner are where roadmaps stall: high ROI but high complexity becomes a two-year project; low complexity but low pain gets ignored by users.
- High pain: users feel it daily and will pull the solution in.
- Low complexity: achievable in 90 days with existing data.
- Clear ROI: one baselined metric you can prove you moved.
Frequently asked questions.
Should the first use case be customer-facing?
Not necessarily. Internal, high-friction tasks (support triage, document processing, data entry) often score higher on the Golden Triangle because you control the environment, the data is available, and you can measure the metric cleanly before exposing anything to customers.