What I Kept Seeing Before the AI Ask
A peek inside a cross-functional meeting to add AI layer. Find out what happens behind the scenes versus the market noise on AI.
What I kept seeing in the field.
Teams wants AI.
Nobody could name the digital layer underneath it.
What I kept seeing
The conversation would start with a new ask: AI copilots, modernisation, data cleanup, governance. But the moment someone asked for the list of recurring dashboards, workflows, scripts, forms, bots, data flows, and embedded SOPs already running work, the room went quiet.
Why it kept happening
The market rewards what sounds new. Operators live inside what already works or works half the time. So the shiny ask gets airtime, while the baseline stays invisible. The result is the same every time: fuzzy ownership, duplicates, hidden risk, and no clean entry or exit criteria.
What broke downstream
Priorities got argued emotionally or qualitatively instead of compared side by side.
Risk stayed hidden because nobody knew which digital assets were genuinely critical.
Duplicate work survived because it didn’t look like duplication yet.
AI and upgrade conversations kept restarting because the base layer still wasn’t named.
The operator move
Pick one boundary. Build a 10 to 15 row Digital Work Portfolio. Use owner + reuse / risk / value. Force one keep / fix / stop criteria and one next action.
Do it before the next AI workshop, not after it.
What changed once it was named
Once the base layer was visible, the conversation changed tone. Ownership gaps start to show up earlier. Duplicate work became visible and easier to retire/decommission. The AI talk got less airy and more useful because people could finally see what they were trying to improve.
Related read
