A diagnostic for revenue leaders
Seven questions to answer before your board asks them. Phase one bought tools. This is how you find out whether it bought efficiency.
How to use this. Answer each question with a number or a name. Not a story, not a sentiment, not "the team says it's helping." If the honest answer is "I'd have to check," mark it unanswered — that's the point of the exercise. Ten minutes, alone, before you socialize any of it.
What is your revenue per GTM head today, against the quarter before you started buying AI?
Why it's the first question. It is the only figure that shows whether AI changed your economics or only your experience. Faster work that produces the same revenue with the same headcount is a nicer job, not a better business.
Which steps did AI remove — not speed up?
Why it matters. Accelerating a step is a convenience. Deleting one is a structural change. Most phase-one deployments made every existing step faster and removed none of them, which is why the cost line moved and the output line didn't.
What does it cost you to check the AI's work?
The hidden line item. Every draft a human reviews carries supervision time that no one budgeted and no dashboard reports. When the review is heavier than the task, the tool is a cost centre wearing a productivity badge.
Where does the work still get hand-carried?
Where the hours hide. Deal desk. Quote and proposal assembly. CRM hygiene. Board-deck and QBR preparation. Ramping a rep into an unfamiliar segment. These rarely appear in a tooling review because no tool owns them.
When you enter a new segment, does the motion transfer or get rebuilt?
The expansion tax. A motion that works in one market and is rebuilt by hand for the next is not a system — it's a talented person doing it again. That cost is invisible until you're running two or three segments at once.
Who owns the AI operating model?
Tools have owners. Systems need one too. Phase two runs on decisions someone has to make and hold: what data the models reach, what an agent may do without a human, who reviews it, who is accountable when it acts. That is a role, not a committee.
What happens when it's wrong in front of a customer?
The question your board will reach eventually. Efficiency stops being interesting the moment automation embarrasses you in a live account. Knowing the answer in advance is what separates a programme from an experiment.
Count only the questions you answered with a number or a name.
You are measuring the right things and you can defend the spend. Put it in front of the board before they ask again — the advantage here is going first.
The most common result, and not a failure — it's what phase one produces. The gap is an operating model, not more software. Buying another platform now will make the reporting prettier and the economics identical.
Start with the baseline, not a vendor. You cannot evaluate a solution before you can state the problem in numbers — and any vendor who says otherwise is selling to the gap, not closing it.
I spend fifteen minutes with revenue leaders working through exactly this — no deck, no demo. If the baseline turns up something worth fixing, you'll know what to do next whether or not you work with me.
Book fifteen minutesLove Hudson-Maggio — fractional AI executive. She builds the operating model behind the tools: shipped agents, signed governance, and human-plus-agent team design for enterprises from a $9B health system to a national nonprofit — and runs her own firm on the same stack. calendly.com/lhmaggio