The Operational Experience Crisis

• Correlation between distributedness and OES: –0.22
  This negative relationship indicates that as applications are deployed across more environments, operational experience declines. Each environment introduces unique tooling, policies, and integration challenges, fragmenting control and slowing response.
• Correlation between application count and OES: +0.05 
  On its own, application volume has a negligible correlation with operational friction. But when combined with distribution, it paints a clearer picture: organizations with both many apps and many environments suffer the lowest OES scores.

Application scale adds vertical pressure: more apps mean more configs, policies, and updates to manage. Distribution adds horizontal complexity: more environments require more integrations, visibility tools, and specialized expertise.

Put together, they create a compounding drag on operational velocity.

Industries like Financial Services, Energy, and Manufacturing, which score lowest on OES, are also among those with the highest distribution and application scale. In contrast, Education and Healthcare, which typically have fewer applications and simpler infrastructure topologies, report significantly higher OES scores.

Despite years of investment in automation tooling, a significant portion of operational work remains manual. This isn’t just about legacy systems, it's also about the gaps between tools, the lack of standardization, and the friction introduced by scripts and APIs that were never meant to scale.

• 41% of respondents said that interacting with vendor APIs or writing scripts is the most time-consuming task in their workflow.
• Even in cloud-native organizations, automation often means "a collection of scripts," not streamlined orchestration.

This signals a critical need for more intelligent, adaptable automation, the kind that reduces dependency on hardcoded logic and tribal knowledge.

The data reveals that delays and inefficiencies are not isolated to a specific tool or process. Rather, they are systemic and structural, caused by fragmented approval chains, process silos, and non-integrated systems.

• Nearly 30% of respondents cited waiting for approvals or handoffs as a key contributor to delays.
• 18% reported that poor integration across tools was a major source of inefficiency.

This demonstrates that improving workflow efficiency isn't just about deploying more tools, it's about reimagining processes and creating feedback loops that reduce latency between intent and action.

While automation maturity varies, openness to AI tools is consistently strong across roles and industries. This interest isn’t hypothetical, it correlates strongly with areas of high operational friction.

• The correlation between AI interest and OES is –0.62, meaning those in the most pain are the most ready for AI intervention.
• Teams want AI not to replace humans, but to reduce the cognitive load of repetitive, error-prone work: log summarization, policy generation, status reporting, auto-scaling, etc.

This willingness suggests a powerful opportunity: AIOps adoption can be accelerated bottom-up by empowering the very teams who are feeling the most strain.

The combined impact of application scale and distributed infrastructure creates a unique operational challenge. While each factor alone has a limited effect on OES, together they produce measurable friction.

• Industries with both high application counts and multi-environment deployments (e.g., Financial Services, Energy, Manufacturing) scored lowest on OES.
• In contrast, sectors like Education and Healthcare, where app portfolios and environment diversity are smaller, scored significantly higher.

This reinforces that complexity is a compound function: it increases as scale (vertical burden) and distribution (horizontal sprawl) rise. And it’s here that automation, especially AIOps, becomes not just useful, but essential.

The data from the global State of Application Strategy Report 2025 is unequivocal. Across industries, the top operational challenges are not exotic, they are painfully familiar:

• Fragmented deployment topologies spread across public cloud, private data centers, colocation, edge, and SaaS.
• Rising application volume and velocity, especially in highly digitized sectors.
• A continued reliance on brittle scripts and tribal knowledge, which make even basic automation fragile and difficult to scale.

These realities aren’t new, but what is new is the growing recognition that human effort alone cannot solve them.

And yet, there is a silver lining: practitioners are ready. 

Survey responses show that the people closest to the pain are also the most willing to change:

• 56% want AI to assist with repetitive or complex tasks, such as summarizing logs, generating scripts, or enforcing policies.
• 38% want AI to automatically adjust configurations and make decisions, not just suggest them.
• Only 6% expressed disinterest in AI tools altogether, a minority that underscores how widespread the appetite for change has become.

This signals a cultural shift: teams aren’t afraid of AI, they’re asking for it. Not to replace human judgment, but to augment it, accelerate it, and free it from drudgery.

Every script written to fix a broken handoff. Every delayed deployment due to manual approvals. Every ticket created because systems can’t talk to each other. These are not minor inconveniences; they are compounding sources of operational debt.

And in a world where digital performance is business performance, that debt translates directly into:

• Slower time to market
• Increased security and compliance risk
• Talent attrition due to burnout and toil

The OES reveals this hidden cost with precision, and it shows which organizations are at greatest risk of falling behind.