Dasseti Insights

The Context Moat: Why 95% of AI Projects Fail and What the Winning 5% Know

Written by Lachlan Fogarty | Jun 11, 2026 4:38:15 PM

A few years ago I bought a top-of-the-range bike computer. GPS, route tracking, elevation, heart rate, power output - everything a data-obsessed cyclist could want. There was just one problem. My power meter wasn't calibrated correctly.

The hardware was flawless. The software was state of the art. But because the underlying data feeding the system was wrong, every output it produced was wrong too. My training zones were off. My sessions were miscalibrated. I undertrained for months without realising it, and when the events I had been building towards finally arrived, I missed the goals I had set.

The technology didn't fail me. The data did.

I think about that power meter a lot, because I see the same story playing out across the investment management industry right now - just with considerably more money at stake.