By Sachin Salian, SVP & Global Business Head – Cloud and Data Services, Writer Information
A cloud provider can meet its contractual obligations while the enterprise continues paying for resources it does not need. Servers remain available, incidents are resolved and service reports look reassuring. Yet an idle environment survives another billing cycle because the provider cannot retire it, the application team has other priorities, and finance cannot identify who must decide.
That is the cloud accountability deficit: a gap between responsibility for keeping technology running and authority to make it serve the business efficiently. Indian enterprises should treat closing that gap as a condition of cloud maturity.
Flexera’s 2026 State of the Cloud Report 1estimates respondents wasted cloud spend at 29%, reversing a five-year downward trend. The same report finds that 63% of organisations have established FinOps teams. These are global findings, rather than a measure of waste under managed providers. Their significance is that governance structures and persistent inefficiency can exist together.
The lesson is uncomfortable for enterprises and providers alike. Outsourcing operational work does not settle who approves spending, challenges architecture or retires obsolete workloads. Those decisions cross finance, engineering, procurement and business teams. Unless contracts connect these responsibilities, each participant can complete its assigned task while waste continues.
The enterprise cloud conversation must therefore move towards business accountability. A managed cloud agreement should establish who must act, what outcomes they must deliver and how those outcomes will be verified. Cost, Control, Compliance and Continuity provide a practical Cloud Accountability Framework for testing that agreement.
Cost is the first test because it reveals how easily activity gets mistaken for achievement. An optimisation report listing ₹10 lakh in potential savings demonstrates an opportunity. It becomes a business result only when changes are implemented and savings are verified against an agreed baseline.
That baseline must account for changes in demand, service levels, prices and implementation costs. Otherwise, a quieter business month can be presented as an optimisation success, while genuine efficiency gains disappear beneath rising transaction volumes. Finance should validate realised savings, and business owners should confirm that service quality remains acceptable.
For an Indian retailer, cost per completed order may reveal more than the monthly infrastructure bill. For a manufacturer, the relevant measure might be cost per production workload. Such measures connect cloud expenditure to commercial performance and make it harder to claim success through discounts on capacity that remains unnecessary.
Provider incentives belong in this discussion. Where fees increase with consumption, reducing waste can reduce service revenue. Enterprises should understand that relationship and negotiate rewards for verified efficiency and reliable delivery. Providers, in turn, need agreed authority to implement approved changes. Accountability becomes credible when incentives and decision rights support the same outcome.
Control turns those contractual commitments into daily practice. Every material workload needs an owner, spending limits and a review date. Test environments need expiry rules. Anomalies need escalation deadlines. Long-term capacity commitments need demand forecasts that someone is prepared to defend.
These controls matter when legacy systems, multiple clouds and AI experiments sit across different budgets. A consolidated invoice offers limited insight if leaders cannot connect expenditure to utilisation, ownership and purpose. Unified visibility should enable decisions across the technology estate, supported by governance that operates continuously.
The FinOps Foundation’s 2026 report 2supports this broader direction, highlighting governance, forecasting and organisational alignment alongside optimisation. Financial scrutiny should begin during architecture and procurement, when choices remain flexible. Waiting for the invoice makes it an avoidable expenditure for somebody else’s urgent problem.
Compliance tests whether that visibility extends to obligations and evidence. CERT-In’s directions require specified cyber incidents to be reported within six hours of being noticed or brought to notice, and ICT logs to be retained for a rolling 180 days within Indian jurisdiction. Applicable obligations must be reflected in operating procedures.
An enterprise should know who retrieves those logs, establishes what happened and coordinates reporting. Responsibilities across providers and subcontractors must be explicit. Annual assurances cannot establish whether a newly deployed workload follows approved access, retention and incident-response requirements. Evidence must come from the environment running.
Continuity brings the business consequence into view. A platform’s availability percentage cannot establish whether a customer can complete an order, or a factory can access essential production information. Recovery objectives should reflect acceptable disruption and data loss for each critical service.
Festive demand makes this tangible for Indian retailers. Capacity, application dependencies and recovery procedures need testing together. A successful backup is useful evidence only when restoration also works. Leaders should examine provider dependencies and the practical ability to retrieve data or transition services.
The four pillars must be judged together. Removing spare capacity can lower costs while weakening recovery. Retaining every dataset indefinitely can increase expense and exposure. Mature governance records why such trade-offs were accepted, who authorised them and when they will be reconsidered.
Quarterly reviews should document verified savings, unresolved control exceptions, compliance evidence and recovery results. Each gap needs an owner and closure date. The C-suite should require those answers before renewal. A managed cloud relationship earns its value when the enterprise can demonstrate what it gains. That evidence should shape procurement, investment decisions and the next phase of cloud expansion.