John O’Connell Featured in Kiplinger: How Businesses (and Advisers) Can Budget for AI Use When the Bill Keeps Changing

John O’Connell was recently featured in Kiplinger writing about how businesses and advisory firms should budget for AI. The problem isn’t whether to spend, that decision is made, it’s that firms treat AI like standard software when it’s actually a variable cost that tracks consumption rather than headcount, so two advisers on the same license can generate wildly different bills, and usage only climbs as people get better at the tool. Firms also tend to budget only the token bill the vendor invoices, ignoring the human-in-the-loop costs that scale with adoption: reviewing output before it reaches a client, training staff, and maintaining governance and compliance records. Meanwhile the proven return is thin, with less than 1% of executives reporting AI returns of 20% or greater and Gartner expecting 30% of generative AI projects to be abandoned after proof of concept. John’s advice is to start with a metered pilot to establish a true cost per task, build the line as a range with a floor and ceiling rather than a fixed number, budget the token spend and the review, training, and governance costs together, tie each item to a measurable result like hours saved in a named workflow, and revise quarterly since both prices and usage keep moving.