The pattern is common enough at UAE enterprises to be predictable: a cloud cost optimisation engagement identifies 25-35% waste, produces a comprehensive optimisation report, achieves partial implementation over the following 6-8 weeks, then progressively decays. By month six, measured spend has drifted back to within 5-10% of pre-engagement baseline. By month nine, the optimisation is functionally reversed.
The decay is not a failure of the optimisation identification. The waste was real; the recommendations were accurate. The decay is a failure of the operating model — the assumption that governance can be delivered as a one-off report rather than as continuous operations.
Understanding why the decay happens (and doesn't happen through any specific operational fault) matters because the pattern repeats. Enterprises that commission a second optimisation engagement 12-18 months after the first typically observe the same decay pattern.
The three-quarter decay pattern
Quarter 1 post-optimisation: implementation. Highest-priority recommendations get actioned, typically 40-60% of identified waste captured. Team energy is high, executive attention is present, vendor coordination is active.
Quarter 2 post-optimisation: partial recovery. Additional 15-25% of identified waste captured through follow-up work. Operational cadence begins to slip — the FinOps working group meets quarterly instead of monthly, tagging discipline shows gaps as new workloads onboard, rightsizing decisions get deferred as other operational priorities emerge.
Quarter 3 post-optimisation: drift acceleration. New workload onboarding without tagging enforcement, expiring reserved instance commitments not renewed with adjusted commitments, rightsizing decisions on the remaining 40-60% of spend that didn't get addressed in Quarter 1 never get made. Measured spend begins to drift up.
Quarter 4 post-optimisation: functional reversal. Spend has drifted back to within 5-10% of pre-optimisation baseline. Executive attention on cloud cost has moved on to other priorities. The optimisation report is filed and referenced as historical context rather than active operational guidance.
Why the decay is structural, not operational
Three structural reasons the decay happens even when the operational team is capable:
Cloud spend is generated by many decisions across many teams, but governance requires one accountable point. Optimisation reports produce recommendations that teams should implement. Absent accountable point-of-decision, recommendations compete with team-specific priorities and lose. Continuous FinOps governance requires named accountability that persists beyond the optimisation engagement's timeline.
New workload onboarding produces new waste faster than optimisation reduces existing waste. Cloud adoption at growing enterprises means new workloads onboard continuously. Without governance operational discipline (tagging enforcement, rightsizing at onboarding, reserved capacity commitment at scale), new workloads accumulate waste at similar rates to the pre-optimisation baseline. Optimisation reports address the historical waste; ongoing governance addresses the new waste as it emerges.
Reserved capacity commitments and negotiated pricing decay without active management. Reserved instance commitments have defined terms (typically 1-year or 3-year). Without active management to reassess and renew commitments as workload patterns evolve, commitments either expire without replacement (lose the reserved pricing) or auto-renew at outdated commitment shape (over-commit or under-commit for current workload). Enterprise agreements with hyperscalers have negotiated pricing that requires periodic renegotiation. Both decay without operational discipline.
The alternative: continuous governance operations
The operating model that holds cloud cost trajectory is not a better optimisation report — it is continuous governance operations delivered as retainer or subscription:
Monthly governance cadence. FinOps working group meeting monthly with defined agenda: cost trajectory review, tagging coverage report, rightsizing candidates surfaced from the past month's workload changes, reserved capacity commitment review, negotiated pricing effectiveness review.
Continuous tagging enforcement. Workload onboarding with mandatory tagging (enforced at deployment pipeline rather than requested via policy documentation). Tagging drift detection and remediation running continuously.
Quarterly release cycle for governance model adjustments. Governance model itself gets adjusted quarterly based on operational learning — tagging standard refinements, rightsizing criteria updates, reserved capacity commitment shape adjustments.
Named accountable point for cloud spend trajectory. FinOps role with authority to enforce governance discipline across cloud-consuming teams. Typically inside the finance function or the cloud platform function with cross-functional authority.
The commercial economics comparison
For a UAE enterprise with cloud spend in the AED 4-8M annual range:
One-off optimisation engagement: AED 300-500K one-time cost. Captures 25-35% waste in Quarter 1. Decays to 5-10% below baseline by Quarter 4. Net year-one cost savings approximately AED 400-800K.
Continuous FinOps retainer: AED 800K-1.2M annualised cost. Captures 25-35% waste in Quarter 1 (equivalent to optimisation engagement). Holds trajectory through Quarters 2-4 (unlike optimisation-only). Net year-one cost savings approximately AED 1.2-2.4M.
The retainer costs 2-3x more than the one-off engagement. It delivers 3-4x the sustained cost savings. And unlike the one-off engagement, the savings hold in year 2 and year 3 without additional intervention.
The commercial preference for one-off engagements over retainer is not driven by economics — it is driven by the accounting treatment (one-time capex vs recurring opex) and by the internal preference for defined-scope engagements over ongoing commitments. Neither driver reflects the operational reality of cloud spend governance.
The retainer only makes sense if you're committed to acting
One caveat matters. The continuous FinOps retainer only delivers sustained savings if the enterprise commits to acting on what the retainer surfaces. Monthly governance cadence produces recommendations; the recommendations require implementation authority and organisational willingness to enforce discipline across cloud-consuming teams.
Enterprises that commission a retainer without committing to act typically experience the same decay pattern as one-off engagements, delivered at higher cost. The commitment-to-act prerequisite is what NexITC evaluates before recommending a retainer engagement.
Cloud spend governance is not an optimisation problem. It is an operational discipline problem with technical elements. Operating models that treat it accordingly hold trajectory; operating models that treat it as an optimisation problem observe the three-quarter decay pattern reliably.