
Everyone is talking about getting started with AI
“Are we ready for AI?” is the wrong question. The right one: are the foundations in place for a first agent? Three testable facets of readiness, from our own practice.
Three weeks ago, this space was about where data sits. Today it is about what data means. Because in all three fields we work in at AIC Group, the debate is currently shifting to the same point: it is not the model that decides whether results are good, but what it finds. Models change every few months. Definitions stay.
What is an ‘active customer’? Someone who bought in the last year? Someone who opens the newsletter? Someone who logs in to the shop? As long as campaigns are built by hand, someone catches the vagueness. Once a campaign runs automatically, an unclear definition does not take effect once, but with every send. That is why our Marketing Automation Studio starts with a central data base and not with the first mailing.
An agent works only as well as what it can look up. We have been experiencing this with our own agent for a good year now. If the same thing is stated differently in two places, it cannot know which one applies. So we check its knowledge base for contradictions again and again, the way you reconcile metrics.
Querying data in natural language is becoming normal. The question is asked quickly, the answer comes in seconds. But what is ‘revenue’? With or without returns, by invoice date or by payment received? If there are two definitions, the AI picks one and still answers convincingly. In the past, the contradiction surfaced in the meeting. Today it becomes invisible.
Where is a definition sparking more discussion than the tool at your company right now? Feel free to write to me at meetme@aicgroup.de.
More articles…

“Are we ready for AI?” is the wrong question. The right one: are the foundations in place for a first agent? Three testable facets of readiness, from our own practice.

What happens when the AI agent becomes the most convenient colleague? On the quiet crowding-out of human collaboration, and how we deliberately push back.

An AI agent is only as fast as the team that can carry it. Why good AI does not solve the speed problem but shifts it, and what that means for working together.
*Articles are partly written with the help of AI.