Lower prices can coincide with greater consumption
Resource prices may fall while calls multiply. Retries, duplicated data and idle environments change the cost structure. Procurement discounts alone may not explain spending.
FinOps and technology value management.
Technology costs often turn when patterns of use change.
As platforms spread, more technology is consumed as shared services. We believe cost management must evolve accordingly: contracts no longer determine the budget alone. Calls, configurations and service commitments shape ongoing spend. Understanding the bill must be followed by understanding who can change it.
A SIGNAL WORTH EXAMINING89%
of respondents report using an internal developer platform3
2024 · Cross-industry software-delivery surveyTHE ZHICE INTERPRETATION
Once platforms are common, how they are used makes the difference.
Buying or building a platform is not sufficient evidence of digital maturity. We see its value in reducing repeated work on common problems while preserving meaningful business differences. When standards become another approval queue and shared capabilities have no service owner, centralisation can slow change. The next question is how the platform changes delivery behaviour.
THE ARGUMENT BEHIND THE VIEW
A judgement needs a complete argument.
Technology costs often turn when patterns of use change.
THE BUSINESS LOGIC OF TECHNOLOGY
THE REASONINGThe business conditions behind the view
Sharing complicates accountability
One resource can support several businesses. Without understandable allocation and cost drivers, teams cannot distinguish useful consumption from inherited practice.
Reliability is part of the cost choice
Redundancy, recovery and field support for critical energy activities are not arbitrary extras. Compare equivalent service conditions rather than creating savings by transferring operational risk.
IMPLICATIONS FOR LEADERS
What should leaders reconsider?
Explain cost in business units
Relate spending to useful device connections, completed tasks or transaction services. Move the discussion beyond totals while retaining quality conditions.
Put decisions with the right owners
Platform teams manage shared efficiency, business teams specify service needs and finance maintains comparable definitions. Consumers should see the cost consequences of choices.
Make value a rolling decision
Adjust capacity and services as demand, performance and costs change. A one-time acceptance decision cannot substitute for ongoing operating judgement.
WHEN TO RECONSIDER
A falling unit cost does not establish rising value. Changes in volume, quality and risk exposure require the comparison to be interpreted again.
Which costs grow with useful business activity, and which persist because nobody has changed old consumption patterns?
FROM PERSPECTIVE TO PRACTICEPut this perspective to work.
ExpandCollapse
Put this perspective to work.
Cloud, software subscriptions, data platforms and AI inference appear on different bills, while budgets may not move with business demand. Price negotiation alone cannot distinguish necessary consumption from waste and can shift cost into service risk or manual work.
Create a shared cost view
Reconcile bills, contracts and usage. Establish ownership, shared-cost rules and data gaps, separating one-off implementation from recurring operation.
Bring cost into design
Compare capacity, service levels and model-call patterns during architecture and use-case reviews. Define budget thresholds, anomaly alerts and business approval routes.
Optimise against value
Review idle resources, unit service cost and demand with business, finance and technology teams; record effects on performance, reliability and workload.
Work products
Technology cost-allocation model
Budget, consumption and unit-economics dashboard
Optimisation backlog and operating-review mechanism
What to measure
- Explainable-cost coverage
- Measure spending attributable to an owner or explicit sharing rule, with unallocated cost reported separately.
- Cost per useful service
- Measure per useful business request, device service or accepted AI task, accounting for quality and demand mix.
Savings depend on usage, contracts and service requirements; the engagement does not assume a universal savings percentage.
Data notes and research foundationsExpandCollapse
- 1IEA · Key Questions on Energy and AI, 2026
Central projection for all global data centres, not AI alone or a capacity plan for an individual enterprise.
IEA 2026 · Global data centres · 2030 projection - 2IBM Institute for Business Value · 2025 CDO Study
Self-reported confidence across 27 geographies and 19 industries; not an energy-sector adoption rate or observed project success rate.
IBM 2025 · 1,700 senior data and analytics leaders globally - 3DORA · 2024 Report Infographic
The study uses a broad platform definition; associations with productivity, throughput and stability are mixed.
DORA 2024 · Cross-industry software-delivery survey
THE NEXT CONVERSATIONThe question worth discussing:
Contact ZHICE