AI FinOps meets cost accounting: why showback is not enough
In two years, the share of FinOps teams managing AI spend went from a third to nearly all of them, and the Linux Foundation has announced an intent to standardise the discipline as a Tokenomics Foundation. AI FinOps is real and it matters: it meters AI usage and tags it to teams and products. But tagging answers one question, where did the cost land, and leaves two unanswered: why did it occur, and how much capacity did we pay for without using. Those are cost-accounting questions, and they are where activity-based costing completes what FinOps starts.
AI FinOps, increasingly called tokenomics, brought visibility to a cost that used to arrive as one opaque bill. It meters tokens and inference, attributes them to teams, products and customers through tagging, and supports showback, showing each unit its cost, and chargeback, moving the cost into that unit budget. The FinOps Foundation now treats the token as the atomic unit of AI consumption. This is a genuine advance, and any company spending seriously on AI should have it. It is necessary. It is just not sufficient.
Tag-based showback tells you that a team or product consumed a certain amount of AI cost. It does not tell you why: which activities drove the consumption, which cost drivers, and whether the work was even successful. Nor does it surface the cost of unused capacity, the GPU time you paid for and did not use, which on single-digit utilisation is most of the bill. Showback shows the cost that was incurred; it is silent on the cost that was wasted. That silence is expensive.
Activity-based costing supplies the allocation logic underneath the tags. It identifies the activities that consume AI, the drivers that scale them, and the cost objects, products, customers, processes, that should carry the cost. Time-Driven Activity-Based Costing adds the practical-capacity rate, which makes the cost of unused capacity a visible line rather than an inflated blended rate. Put together, FinOps gives you the meter and the tags; activity-based costing gives you the why and the waste. The first tracks spend; the second turns it into profitability.
SHOWBACK VS ACTIVITY-BASED COSTING
Illustrative. FinOps and tokenomics meter and tag AI spend, answering where cost landed. Activity-based costing adds why it occurred and how much capacity was wasted, the layer showback leaves out.
Showback moves the number into a budget. Activity-based costing explains the number, and reveals the capacity you paid for but never used.
Common questions
- What is AI FinOps, or tokenomics?
- AI FinOps, increasingly called tokenomics, is the discipline of metering, attributing and managing the cost of AI usage, especially token and inference spend. It tags AI cost to teams, products and customers and supports showback and chargeback. The FinOps Foundation treats the token as the atomic unit of AI consumption, and the field is formalising fast, with a Linux Foundation Tokenomics Foundation announced in 2026.
- What is the difference between showback and chargeback?
- Showback shows a team or product the AI cost it generated, for visibility, while keeping the cost in a central budget. Chargeback moves the actual cost into that team or product budget. As a phrase used in FinOps puts it, showback moves information and chargeback moves money. Both rely on accurate allocation, which is where activity-based costing strengthens them.
- Why is AI FinOps not enough on its own?
- Because tag-based showback tells you where cost landed, not why it occurred or how much capacity was wasted. It does not identify the activities and drivers behind the consumption, and it does not surface the cost of unused capacity, which on typical single-digit GPU utilisation is most of the bill. Activity-based costing supplies that missing logic, turning spend visibility into profitability management.
- How do FinOps and activity-based costing work together?
- FinOps provides the meter and the tags: it captures AI usage and attributes it. Activity-based costing provides the allocation rigour: the activities, the cost drivers, the cost objects, and, through TDABC, the practical-capacity rate that exposes unused capacity. Used together, you get both accurate spend tracking and a defensible unit cost and profitability view of AI.
Turn AI spend tracking into profitability.
We add the costing rigour beneath your FinOps tags, so you see why cost occurred and what was wasted.
Take the Profit Check