YOUR BUSINESS IS INVESTING IN AI. WHAT WOULD JUSTIFY INVESTING MORE?
AI activity is easy to count. Business returns are harder to prove.
Seven questions for examining the evidence behind an AI investment decision.
Not a verdict. A way to test the thinking.
A private thinking experience prepared for Rowan De Klerk. No analytics or external services. Your answers remain in this browser session.
WHAT EVIDENCE WOULD JUSTIFY APPROVING THE INVESTMENT?
An AI proposal can sound compelling without establishing a financial case. What evidence should exist before funding is committed?
If only one of these were missing, which absence would make approval most difficult to defend?
ARE WE MEASURING PRODUCTIVITY OR FINANCIAL RETURN?
Saving hours does not automatically create financial value. A productivity gain only becomes a business result when something changes.
What must happen after time is saved before you would count it as a return?
WHAT COSTS ARE MISSING FROM THE BUSINESS CASE?
The licence price is often the easiest number to see. The full economic cost may be less visible.
Which overlooked cost could change the investment decision?
WHAT WOULD HAVE HAPPENED WITHOUT AI?
If performance improves after an AI rollout, how much of that improvement should be attributed to AI rather than other changes?
What would persuade a sceptical CFO that AI caused the improvement rather than merely coinciding with it?
WHO OWNS THE PROMISED RETURN?
Technology teams may deliver a working system while the financial benefits depend on decisions elsewhere in the business.
Who has both the authority to change the process and accountability for the commercial result?
WHEN SHOULD LEADERSHIP STOP FUNDING THE EXPERIMENT?
Continuing to invest can be justified, but only if there is a credible path to learning or value.
What evidence would trigger a stop, redesign or scale decision?
WHAT MAKES AN AI INVESTMENT DEFENSIBLE TO A SCEPTICAL CFO?
Imagine reviewing the investment six months after approval. Which evidence would make the original decision credible even if the experiment did not succeed?
Could the business explain not only what it spent, but why it continued, changed course or stopped?
ONE FINAL QUESTION.
What if the biggest risk isn't investing in the wrong AI technology?
WHAT IF IT'S APPROVING INVESTMENTS WITHOUT KNOWING WHAT EVIDENCE WOULD JUSTIFY CONTINUING?
The question isn't whether AI can do more. It's whether the business can show what changed, why it matters and when to change course.
Carlos Batista
AI Growth Advisor
No pitch. No analytics. No data is sent or saved.
