AI adoption is becoming a weak way to distinguish mature customer experience organizations. Five9, a contact-center technology vendor with a commercial stake in CX technology adoption, says in its “2026 Business Leaders CX Report” that 90% of CX organizations are piloting or deploying AI. Five9 also reports that respondents are almost evenly split among end-to-end platforms, hybrid environments, and best-of-breed solutions.

Companies can reach similar levels of AI adoption while making different choices about the systems, models, and infrastructure underneath it. Adoption alone says little at that level. Architecture is also an incomplete maturity test because AI can operate across several approaches. A stronger test is whether leaders can produce measurable value while controlling security, integration, reliability, compliance, and customer risk as technology changes.

AI adoption is converging. CX architecture remains diverse.

A standard architecture has yet to emerge alongside widespread AI adoption. Five9 describes deployment approaches as nearly evenly divided among end-to-end, hybrid, and best-of-breed models. An end-to-end approach consolidates capabilities within a broader platform, while a best-of-breed environment selects individual components for specific needs. A hybrid environment retains a mix of systems or deployment models.

Architecture shapes the cost and difficulty of future changes. A company selecting one broad platform may reduce some integration and governance work while becoming more dependent on that platform’s technology choices. A company retaining multiple components may preserve more choices while taking on more integration and operational coordination. These are management trade-offs.

The practical question is whether the current design lets the company put AI into useful workflows safely and economically. Leaders also need to understand what replacing a model, provider, or component would require across data work, integration, testing, security, and governance. Architecture then becomes a way to manage technology decisions over time.

Traditional maturity signals are becoming less useful

Cloud migration has provided an intuitive measure of modernization: move workloads away from on-premises systems and reduce the migration backlog. Five9’s findings complicate that measure for customer care. It says 84% of organizations are somewhere in the transition from on-premises systems to the cloud, while current customer-care environments remain heavily hybrid.

Customer-care environment Share of respondents
Hybrid 74%
Fully cloud-based 16%
Entirely on-premises 10%

These figures weaken attempts to classify every hybrid organization as a technology laggard. AI is already being piloted or deployed across most CX organizations while 74% of customer-care environments remain hybrid. The two measures are moving at different rates within Five9’s survey population. That coexistence is clear in the reported data, though the survey cannot establish why each respondent remains hybrid.

A hybrid estate can result from different management decisions. Some systems may remain because leaders value the available choices or consider migration risk too high today. Others may stay because a difficult migration is unfinished. The same visible architecture can reflect different constraints and decisions, so executives need to examine the reasons behind it.

That changes the maturity question. Leaders should be able to identify which workloads remain in place because of an active design decision, which remain because moving them would create unacceptable risk, and which are waiting on unresolved technical work. Those answers show whether the architecture is being actively managed and where intervention is required.

Okoone experts
LET'S TALK!

A project in mind?
Schedule a 30-minute meeting with us.

Senior experts helping you move faster across product, engineering, cloud & AI.

Please enter a valid business email address.

Migration friction makes the finish line a weak measure

Five9 reports migration obstacles across security, integration, data, reliability, compliance, cost, skills, and customer experience. Just 4% of organizations report no migration challenges. The breadth of these barriers makes it difficult to infer that widespread hybrid customer care reflects a shared preference for a long-term hybrid design.

Migration issue Share citing it
Data security and privacy 36%
Integration with existing IT infrastructure 35%
Data migration 34%
Reliability 34%
Customer experience disruptions 33%
Regulatory compliance 33%
Software adaptation 32%
Implementation costs 31%
Technical support 29%
Scalability 29%
Staff training 29%

For CIOs, the breadth of that friction matters. Integration with existing IT infrastructure sits almost level with security and privacy, while data migration and reliability follow closely. Migration brings technology, controls, and customer-facing operations into the same change program. The reported barriers show why migration progress provides an incomplete view of execution quality.

Customer experience disruption creates a direct operating constraint. A contact operation has to preserve service while infrastructure changes underneath it. Reliability, data security, and regulatory compliance can affect both the value and risk of the transition. Migration quality therefore deserves separate scrutiny from migration speed.

Five9 characterizes its findings as suggesting that flexibility is becoming part of the strategy. Five9 benefits commercially from continued investment in contact-center technology, so executives should treat that interpretation as vendor framing rather than a motivation established for every respondent. The migration data support a narrower conclusion: hybrid environments are common while concrete migration barriers are also common.

Execution amid uncertainty is a stronger maturity signal

Five9 reports AI deployment across several operational use cases. The leading applications map to identifiable business processes such as answering requests, reviewing interactions, analyzing conversations, monitoring compliance, and assisting agents. That pattern gives executives concrete workflows in which to assess outcomes, controls, and the cost of failure.

AI use case Deployment
Self-service automation 42%
Quality management and automated quality assurance 41%
Speech and text analytics 40%
Real-time compliance monitoring 39%
Agent assistance 38%
Knowledge authoring 30%
Journey analytics 28%
Personalization 26%

Quality assurance illustrates the management logic. An organization already has interactions to inspect, review processes to improve, and human practices against which AI-assisted work can be evaluated. Agent assistance similarly inserts AI into an existing employee workflow. In both cases, management can ask whether the intervention improves operations and whether its outputs remain dependable.

Knowledge authoring, journey analytics, and personalization show lower deployment in Five9’s figures. Five9 attributes the greater difficulty of these applications to richer customer-data requirements, stronger governance, and greater confidence in AI-generated decisions. The deployment figures establish the reported adoption difference; Five9’s explanation remains a vendor interpretation.

This supports a maturity model based on selective deployment. Use cases have different data requirements, potential failure costs, and control needs. Executives can assess where automation is reliable enough to use, where human review is required, and where additional controls are needed before expansion. Each deployment should connect to an operating outcome and an acceptable risk level.

Five9 also reports changes in agents’ work. Overall, 45% of respondents say AI provides better data-driven decision support, while 44% say it helps agents manage exceptions and judgment calls. Five9 says AI increasingly handles routine activities including documenting interactions, retrieving knowledge, translating conversations, summarizing interactions, and handling self-service requests. This shifts the management question toward how human attention is allocated as software performs more repeatable work.

Among U.S. respondents, Five9 says 45% cite increased efficiency and productivity as AI’s biggest benefit, while 42% report improved decision-making and 42% say AI has improved customers’ perceptions of their organizations. These measures span operating performance, judgment, and customer response. They give executives several outcomes against which to evaluate deployment.

Architectural diversity can coexist with this deployment pattern. Organizations can introduce AI into individual workflows while cloud migrations continue and platform choices remain unsettled. Five9 also reports interest in environments that let organizations use different models for different tasks. Such a design may make technology easier to change as model performance and business requirements evolve, although the reported data do not establish why organizations adopted diverse architectures.

The useful capability is controlled change. A company can test whether a new model improves a specific workload, determine what data it needs, assess its security and compliance requirements, and calculate the cost of introducing or replacing it. Leaders can then inspect optionality in practical terms: which component can change, what the change costs, what risks it creates, and which controls must be repeated. Architecture becomes a set of decisions management can inspect.

Positive ROI is one maturity signal

Financial returns sharpen the issue. Across every AI use case measured, roughly nine in 10 respondents report positive ROI, according to Five9. At the same time, only 4% say they have encountered no significant AI implementation challenges. Five9’s commercial stake in AI-enabled contact-center technology is relevant when evaluating its characterization of those returns.

The challenge profile shows why ROI is one measure among several. Five9 says data security is the leading AI implementation concern at 31%. Reliability, scalability, and customer consent each concern 27% of respondents, while ethics, regulatory compliance, infrastructure, AI expertise, budget constraints, and customer discomfort all register above 20%. Reported returns and implementation concerns coexist within the survey findings.

For a CEO or CIO, positive returns establish a case for examining whether a use case is worth expanding. Scaling still requires separate evidence on reliability, customer consent, security, infrastructure, and compliance. Executives also need to know how easily critical models and components can change as requirements evolve. A mature program makes those costs, controls, and dependencies visible before expansion creates harder-to-reverse commitments.

Waiting for every architectural choice to settle can carry an opportunity cost. Five9 reports positive returns from current AI uses while respondents continue to report migration and implementation difficulties. The executive problem is ongoing allocation: decide which gains justify deployment now, which risks require more control, and which technology commitments would make future change too expensive. Those decisions can be revisited as evidence, models, and business requirements change.

Main highlights

  • Manage architecture for change: CX organizations are converging on AI adoption while remaining split across end-to-end, hybrid, and best-of-breed environments. CIOs can assess architecture by the cost, risk, and effort required to change models, providers, or components.
  • Diagnose why hybrid systems remain: With 74% of customer-care environments still hybrid, cloud migration alone provides an incomplete measure of maturity. Technology leaders can separate deliberate architecture choices from risk constraints and unresolved migration work.
  • Measure migration quality: Security, integration, data migration, reliability, compliance, and customer disruption all create significant migration friction. CIOs can evaluate whether infrastructure changes preserve service and controls alongside tracking migration progress.
  • Deploy AI selectively: AI maturity depends on putting suitable use cases into production with measurable outcomes and appropriate controls. CX leaders can match each workflow to its data needs, failure costs, human oversight, and acceptable risk.
  • Test ROI alongside operational risk: Roughly nine in 10 respondents report positive ROI across measured AI use cases, while implementation challenges remain widespread. CEOs and CIOs can use returns to identify expansion candidates, then test security, reliability, compliance, customer consent, and technology dependencies before scaling.

Alexander Procter

September 18, 2026

8 Min

Okoone experts
LET'S TALK!

A project in mind?
Schedule a 30-minute meeting with us.

Senior experts helping you move faster across product, engineering, cloud & AI.

Please enter a valid business email address.