Portfolio ROI

We are not going to concentrate AI and digital investments in one or two large bets or expect every project to be successful. The 10-5-4-1 model reframes this entirely: build a portfolio of initiatives, accept that some will fail, and design for the few that will generate outsized returns.

The math works at the portfolio level, not the project level. Five failures are not problems — they are expected. Four solid hits deliver meaningful returns. And one home run can transform the business.

A note on the numbers: The 5-4-1 split and the 2x or 10x thresholds are directional, not an exact formula. Every portfolio is different — the point is the mindset, not hitting precise ratios.

The Learning bucket is broad by design. It includes any initiative that didn't yield a good-enough ROI or was discontinued for any reason — lack of resources, technology gaps, funding cuts, shifting priorities, outright failure, or anything else. These are not failures to hide; they are part of a healthy portfolio.

Everything is evaluated on a 3-year horizon. Unless otherwise noted, all ROI calculations and portfolio assessments assume a 3-year timeframe. This is long enough to capture meaningful returns from AI and digital investments, and short enough to remain relevant for planning purposes.