Notes · 1 October 2026

Why new software fails in a business, and it's rarely the software

Most systems that get abandoned weren't bad tools. They asked people to work differently from how they actually work, and nobody sat with them until the new way became habit. That's as true for AI tools as it was for a real estate CRM.

I've been thinking about why new software fails in a business, and it's rarely the software.

In more than ten years of real estate I saw plenty of systems bought and then quietly abandoned. New CRM, new listing tool, new app for open homes. Six months later everyone was back to the notepad and the group chat. The tool wasn't bad. It just asked people to do something different from what they actually did on a Saturday morning, and nobody had sat down to work out what that was.

When I was at RE/MAX and LJ Hooker the systems that stuck were the ones set up around how people already worked, then someone sat with the team until it was habit. Not a training session. Sitting with them.

That's the bit I keep coming back to with AI. The model is the easy part. Getting a real person to change what they do on a Saturday morning is the work.

Happy to compare ideas if you're in the middle of rolling something out.

Before you roll anything out

None of that changes because the tool has AI in it. If anything it matters more, because the tool can do more things nobody asked for.

First shared on LinkedIn.

Dylan Hopp
Dylan Hopp
General Manager, Brisbane. Builds with AI. More about me

← All Notes