06/04/2026
98% of agencies are planning AI investment in 2026.
Most will still be running the same pilot in 2027.
The gap between the two isn't budget. It isn't the model. It isn't vendor selection.
It's whether the work underneath is structured well enough for AI to operate on.
You can't automate a vague queue. AI needs to know which tasks exist, who owns them, what their inputs and outputs are, and which ones require a licensed person. Most agency service operations don't have that structure. Work lives in email threads, AMS activity logs, and people's heads.
So the AI gets handed work that's partially defined, with no feedback loop and no measurable outcome. It does what it can. What it can do isn't enough to compound.
The agencies that will have a real AI story in 18 months are not the ones picking the best model. They're the ones instrumenting the operating layer first.
S&G Mitchell built that foundation. 17.9% to 60%+ EBITDA in 12 months. Same book, same carriers. The operating model changed. Everything else followed.
Structure the work. Then let automation run on top of it.
Read the full breakdown at https://hubs.la/Q04k8CqY0