Why AI is not yet showing up in the growth plan

You have asked a fair question. The growth plan in front of you shows almost no AI revenue, and the activity around the company tells a different story. The gap is real, but it is not what it looks like. AI is in the economics of this business today. It is not yet in the line of the plan you are reading, and that is a problem with the gauge, not with the work.

The real question is not how much AI revenue is in the plan. It is whether the plan is measuring what AI is actually doing in this business. The honest answer is no, not yet. This memo sets out where AI is showing up, why the growth numbers cannot register it, and what would change that.

Where AI’s value is actually landing

AI is contributing to this year’s economics in three places. None of them is a new ARR line.

• Gross margin and cost-to-serve. Inside delivery, support, and implementation, AI is letting us serve revenue with less labour. Costs that used to grow with revenue are not growing at the same rate. The benefit shows up in margin, not in growth.
• Retention quality. Features that hold renewals in accounts where a competitor’s demo would otherwise have created a switch. Saved revenue does not appear as growth in the plan. It appears as the absence of a loss, which the plan does not credit.
• Productivity captured by the buyer. A meaningful share of the value our AI creates is being absorbed by the customer as their own efficiency. We have not yet re-priced to recapture it. The contract was written for the SaaS era and has no mechanism to share the upside back.

The first two are real and measurable. The third is a leak.

Why the growth numbers cannot see it

The plan uses three instruments: new ARR by logo and motion, net revenue retention through expansion, and forecast pipeline by seat count and ACV. Each was the right instrument when value scaled with seats. None of them registers the work AI is now doing.

• New ARR misses cost-to-serve. Margin gains land below the growth line. From the gauge’s perspective they are invisible.
• NDR misses retention saves. NDR rewards expansion in accounts, not loss avoided. A renewal AI helped us hold counts the same as one we would have held anyway.
• Seat-and-ACV pricing misses outcome value. The buyer captures the surplus because the contract has no mechanism to share it back. Until pricing changes, AI value leaves the building.

This is not a forecasting failure. It is a measurement failure. The dials on the plan were built for a different business model.

What would make AI legible in the growth numbers

Three changes, in sequence. None of them is fast, and none should be skipped.

1. A value metric we can charge for and the buyer accepts. Per outcome, per task, per agent action. Until the contract has a unit of AI value attached to it, the upside leaks to the buyer and the growth line stays blank.
2. Retention-saved reported separately. Track and credit the renewals we would have lost without AI. Not as new growth, which would flatter the plan, but as defended base. Today this work is unrewarded in the numbers, which biases capital allocation against it.
3. A single AI growth line, underwritten against three tests. A value metric in market, a unit margin proven after inference and support on live accounts, and a repeatable motion that does not depend on heroics. Put a number on the line only when a use case clears all three. One marquee deal is an anecdote. A motion is a forecast.

What this means for this plan cycle

This is the operating agenda I am asking the board to support.

• Stop. Counting defensive AI as new growth. Reporting usage as if it were revenue. Funding features whose only case is that competitors have them, unless they defend a specific and quantified renewal risk.
• Test. Concentrate investment on one or two use cases run end-to-end as the underwriting experiment: priced on a value metric, measured at unit margin, sold by a motion that does not depend on me personally. The aim of the spend is not revenue this year. It is a proven, repeatable motion we can underwrite next year.
• Watch. Three signals will tell us AI is ready for the growth line before the revenue itself does: the value metric our category is converging on, our AI gross margin on live usage, and renewal quality where AI is written into the contract.

The questions this board should be asking

The familiar question, how much AI revenue is in the plan, rewards optimism and hides the work AI is already doing. I would ask the board to replace it with three.

1. What share of this year’s gross margin and renewal book is AI already defending, and is the plan crediting that work anywhere?
2. Where is AI value leaking to the buyer because the contract structure has not changed, and what is the unit we could charge for instead?
3. Of what we are funding, how much is defending the existing business and how much is building the next growth curve, and are we honest with ourselves about which is which?

In short

AI is in this business already. It is in the margin, in the retention book, and in the surplus we are giving to customers. It is not yet in the growth line, and it should not be until the contracts and the instruments can carry it. The cost of putting AI in the plan too early is a forecast we miss and a board that trusts the next one less. The cost of leaving it out for another quarter is small beside that. We are choosing to earn the line before we write it.