Key takeaway
- Why FinOps adoption stalls and how to fix it: awareness, silos, cost allocation, culture and executive sponsorship, using FinOps Foundation guidance.
The challenges of adopting FinOps are mostly organisational, not technical. The State of FinOps 2026 survey of 1,192 practitioners, representing more than $83bn in annual cloud spend, shows the work now reaching well past public cloud. Getting engineers to act on cost data, agreeing who owns shared costs and keeping executives engaged are the hurdles that decide whether a practice sticks. The FinOps solutions page sets out how we support this work.
This guide is for CTOs, CIOs, CFOs, heads of platform engineering and FinOps leads. It lists the six adoption challenges we see most often, the fix for each, and an order of work using the Crawl, Walk, Run approach. Phases and maturity are covered in FinOps phases and maturity levels; this post stays on why adoption stalls and how to unblock it.
What does the FinOps Foundation say FinOps is?
The FinOps Foundation defines FinOps as "an operational framework and cultural practice which maximizes the business value of technology" (source: finops.org). The word "cultural" matters for adoption: the tooling is the easy part, and most failures trace back to ownership and habits.
The FinOps Principles give the cultural test. They state that teams need to collaborate, that business value drives technology decisions, that everyone takes ownership for their technology usage, that FinOps data should be accessible, timely and accurate, that FinOps should be enabled centrally, and that teams should take advantage of the variable cost model of the cloud. A practice that cannot meet the first three is usually the one that is stuck.
Which FinOps adoption challenges come up most, and what fixes them?
| Challenge | What it looks like | First fix | FinOps Framework capability |
|---|---|---|---|
| Low awareness | Engineering, finance and product use different terms for the same cost | Role-based training and a shared glossary | FinOps Education & Enablement |
| Siloed ownership | Teams optimise their own bill; shared costs have no owner | Cross-functional working group and named FinOps practitioner | Allocation; FinOps Practice Operations |
| Cost data that cannot be trusted | Untagged spend, disputed allocation, late reports | Allocation, tagging and shared-cost strategy agreed with finance | Allocation; Data Ingestion |
| Cultural resistance | Engineers see cost work as a tax on delivery | Show cost per unit of value; reward savings; keep the ask small | Executive Strategy Alignment |
| Weak executive sponsorship | Savings targets with no decision rights behind them | Name a sponsor and set guardrails | Executive Strategy Alignment; Governance, Policy & Risk |
| Scope creep into AI and SaaS | AI and SaaS spend sits outside the cloud practice | Add scopes one at a time with an owner for each | FinOps Scopes; Governance, Policy & Risk |
The capability names come from the FinOps Framework, which the Foundation updated in March 2026 and which lists these as capabilities inside its four domains.

Why do teams not understand FinOps?
Many people in finance, engineering and product have never worked under a variable cost model. The Foundation's FinOps Education & Enablement capability describes the fix as training by persona and team, internal events, applied learning and a shared knowledge repository (source: finops.org).
Train engineers on unit cost and finance on usage drivers, not the same course for both.
Publish a one-page glossary so that "commitment", "showback" and "unit cost" mean the same thing everywhere.
Keep a single place for documents and decisions that any persona can reach.
How do siloed teams block FinOps?
When each team optimises its own bill, shared platforms, support and networking costs fall between owners, and savings in one team can raise costs in another. The usual fix is a small cross-functional group with finance, engineering and product in it, plus one named practitioner who owns the practice. Decision rights should be written down, not assumed.
Create a working group that meets on a fixed rhythm and keeps a decision log.
Name FinOps champions in each engineering group to carry the practice into daily work.
Publish cost views per team so that transparency does not depend on a monthly meeting.
Why is cloud cost data so hard to allocate?
The Foundation's Allocation capability says allocation needs an allocation strategy, a tagging or metadata strategy and a shared cost strategy. It notes that tagging compliance is hard to maintain, that some costs cannot be tagged in certain environments, and that many organisations "informed ignore" some shared costs by budgeting them centrally. Its stated goal is not perfect allocation but enough to support good decisions at your current maturity (source: finops.org).
Agree the shared-cost rules with finance before you publish team reports.
Enforce tags through infrastructure as code and policy, not reminders.
Accept a known unallocated percentage and track it down over time.
For the data-platform side of this problem see FinOps for data mesh and data fabric and DataFinOps. For provider-specific work see FinOps for AWS, FinOps for Google Cloud cost optimisation and FinOps for Snowflake.
How do you deal with cultural resistance?
Resistance usually comes from three places: engineers who fear cost work will slow delivery, product owners who fear budget cuts, and leaders who want proof of benefit before they commit time. The State of FinOps 2026 page reports that incentive structures have not caught up with the need for engineers to act, and quotes one practitioner asking how to give developers credit for shift-left work. Treat that as a design problem for your own practice.
Start with one team and one visible saving, and share the result.
Report cost per unit of value, such as per transaction, alongside the total bill.
Credit engineering teams for avoided cost, not only for cost removed.
Make requests small and specific: one action, one owner, one date.
What does executive sponsorship need to cover?
The Foundation's Executive Strategy Alignment capability connects technology spend, usage and adoption to business strategy, and covers decision rights, guardrails (including for AI) and the FinOps operating model and its reporting line (source: finops.org). The State of FinOps 2026 page reports that 78% of practices report to the CTO or CIO and 8% to the CFO, and that practitioners engaged at VP or C-suite level report more influence on technology decisions. The page quotes different figures for the size of that gap in two places, so we do not repeat them here.
Why does scope keep growing?
FinOps no longer stops at public cloud. The State of FinOps 2026 page reports the share of respondents managing each scope: AI 98% (up from 63% in 2025), SaaS 90%, licensing 64%, private cloud 57% and data centre 48%. Each new scope brings a different data source and a different owner, so adoption plans that were built for one cloud bill need revisiting. Our post on the cost side of AI is in FinOps for managing AI adoption costs.
What has changed in FinOps adoption in 2026?
| Change | What the source says | Source |
|---|---|---|
| Wider scope | AI, SaaS, licensing, private cloud and data centre are now managed by large shares of practices | data.finops.org |
| Top forward priority | FinOps for AI is the top forward-looking priority; waste reduction remains the top current one | data.finops.org |
| Framework update | The Framework carries 2026 updates (20 March 2026) on executive strategy alignment and scopes | finops.org |
| Diminishing returns | Practitioners report many smaller savings that take more effort to capture | data.finops.org |
In what order should you tackle adoption?
Assess the current state: Find where spend sits, who owns it and how reliable the data is. The Foundation offers a FinOps Assessment capability for this.
Name a sponsor and a practitioner: Settle decision rights and the reporting line before you buy tools.
Fix allocation first: Agree tagging and shared-cost rules, because every later report depends on them.
Train by role: Give engineering, finance and product training that fits what each must do.
Pick one team and one scope: Prove a result, then widen. This is the Crawl, Walk, Run approach: start small and grow in scale, scope and complexity.
Set guardrails and automate: Add budgets, anomaly alerts and policy once the data is trusted.
Review each quarter: Revisit scopes, owners and targets; add AI and SaaS when you can name an owner.
How should governance work as FinOps spreads to AI agents?
As teams run AI agents that can start workloads and spend money on their own, cost decisions move from monthly reviews to runtime. That needs policy, an owner and a record of why each decision was made. ElixirData Context OS keeps governed context, versioned policy and decision traces in one decision layer above AI agents, and AgentGRC keeps autonomous agent actions inside policy, with the audit trail risk teams need (both per their own sites).
When is a full FinOps programme the wrong first move?
Cloud spend is small and one team owns it: a monthly review and good tagging may be enough.
No one can own shared costs yet: fix ownership before tooling.
The data is untrusted: a savings target will start disputes, not savings.
The organisation cannot act on findings: reporting alone does not change spend.
To map one cost decision your agents should make, and the context and policy it depends on, book a Context Workshop. See what happens in the workshop.
FAQ on the challenges of adopting FinOps
Why do FinOps adoptions fail?
The Foundation's principles centre on collaboration, ownership and accessible data. Adoption usually stalls where these are missing: unclear owners, untrusted cost data and no executive sponsor.
Who should own FinOps in an enterprise?
The State of FinOps 2026 page reports that 78% of practices report to the CTO or CIO and 8% to the CFO. Finance, engineering and product are core personas in the Framework regardless of the reporting line.
Do we need perfect tagging before starting?
No. The Foundation's Allocation capability says the goal is not perfect allocation but enough to support good decisions at your maturity level.
How do we start small?
Use the Crawl, Walk, Run approach: begin with one team and one scope, show a result, then widen scale, scope and complexity.
Does FinOps cover AI and SaaS costs?
Yes. The Framework lists AI, SaaS, licences, data platforms, private cloud and data centre as scopes, and the State of FinOps 2026 page reports 98% of respondents managing AI spend.
Follow XenonStack on LinkedIn: XenonStack on LinkedIn for new guides on agentic AI, data and platform engineering.
Ready to put AI agents to work? Explore ElixirClaw, the agentic OS platform to build, orchestrate and govern AI agents.
Explore the XenonStack platform family: ElixirData · ElixirClaw · ElixirHub · NexaStack · MetaSecure · Neural AI · AgentGRC · AgentSRE · AgentAnalyst · AgentInstruct · AgentRAI · AgentScrum · AgentSearch · AgentForce · AI Aviator · SAIF Aviator · CohortIQ · XenonStack AI · XenonStack UAE