A working assumption we used to hold: "AI automates tasks." A more honest version is that AI takes on roles — and the moment you accept that, your operating model has to absorb it the same way it absorbs a new hire.
The map that mislabels everything
Most AI rollouts we audit are still organized as a sprawl of disconnected tools. A copy assistant here, a meeting summarizer there, a chatbot grafted onto the website. Each one was bought to address a problem, and each one technically works. None of them have a manager.
The deeper issue is taxonomic. The team thinks of these systems as features. The business consumes them as roles. When marketing asks "who owns the brand voice across our AI surfaces?" — that is a role question.
If you cannot name the manager, you do not have an agent. You have an experiment running unattended in production.
Four properties of a real role
Over the last twelve months we have boiled the difference between an experiment and a role down to four properties. We will not ship without all four.
- Scope — a single sentence describing the decisions the agent is authorized to make.
- KPI — one business metric owned by a human and visible on a weekly review.
- Escalation path — the named human the agent hands off to when confidence drops.
- Manager — the human who tunes the role weekly, like a new hire.
What this looks like in production
A mid-market B2B sales team had a "lead enrichment GPT" running for nine months. It worked. It also lived in a Slack channel nobody owned.
We rewrote the same workflow as an AI Sales Qualifier with scope, KPI, escalation path, and manager. Six weeks in, qualified-to-booked rate went from 18% to 31%. The model was not smarter. It had a manager.
The performance lift on rollout almost always traces back to adding the role frame, not to model quality.
Pitfalls we keep seeing
A short list of things we have watched well-funded teams get wrong, in order of cost.
- The orphan agent — no named manager. Performance drift goes undetected for months.
- The five-headed scope — one agent doing qualification, follow-ups, CRM scrubbing, and support.
- The vanity KPI — satisfaction scores instead of P&L metrics.
- The black-hole escalation — handoffs to a generic queue with no SLA.
The org chart we actually ship
On a recent engagement we drew the company org chart on the left, and a second org chart on the right with the same boxes — Revenue, Marketing, Ops, Finance — but each box contained two columns: people, and agents.
That document is now the single most-used artifact in our handoff package. It is not a roadmap. It is an org chart with a second column.