Exec view

The quarterly spend decision

Portfolio return on AI spend, its trajectory, and the risks to it — the three things the budget decision actually depends on.

Recommended move: grow the budget selectively — expand Data Science and Engineering, hold Support, review Marketing.

Blended return is 8.6x and has climbed 3.1x → 7.1x over 26 weeks while spend grew — the marginal dollar is still earning more than the average dollar. The constraint is not budget, it is allocation: +$612k unlockable sits behind a team cap while bottom-quartile spend continues unreviewed.

Annualized portfolio

$1.5M → $12.9M

Metered spend to attributed value across six processes

Return trajectory

3.1x → 7.1x

26-week blended ROI while weekly spend grew $41K → $74K

Primary risk

Concentration

85% of Search Platform's attributed value comes from 2 people

Organizational Token ROI

Marginal value per $1 of token spend · 26-week deployment

Spend vs. attributed value

By business process · trailing 30 days ($K)

Token spend Attributed value

Weekly resource signals

Automated reallocation intelligence from the OLS

Reallocation opportunity

$280K/yr recoverable

Engineering: 3 members in the bottom ROI quartile despite top-20% token spend.

Concentration risk

Succession risk

85% of AI-attributable value in Search Platform is generated by 2 individuals.

Budget expansion case

+$612K unlockable

4 Data Science users are token-constrained by team cap; marginal ROI averages 34x.

Platform phase

Each phase produces the asset the next one requires.

VisibilityLive

Token efficiency and the marginal return on every token spent.

Fleet autonomy

Autonomous management, allocation, and optimization of the agent fleet.

Outcome distillation

Positive outcomes become post-training data that makes models natively efficient at the task.

Full company arc