For more than a decade, contact centers have invested heavily in chatbots and conversational automation. Their promise has always been the same: deflect volume, reduce costs, and resolve customer issues faster.

Yet despite years of tuning, tooling, and AI upgrades, a familiar pattern persists. Customers still escalate to live agents far more often than leaders expect. Resolution rates plateau. Frustration rises. Automation teams work harder, but results remain stubbornly uneven.

The uncomfortable truth is this: most chatbot failures are not caused by weak AI models or poor intent recognition. They are architectural failures.

Many contact centers are running modern language models on top of systems designed for a much earlier era of automation. But as customer interactions grow more complex, emotional, and unpredictable, those foundations begin to crack.

Across the industry, self-service programs still see a large share of customer conversations escalate to agents, especially for billing, eligibility, and exception-handling scenarios.

This article examines why traditional chatbot architectures collapse and outlines a new software design approach, micro-GPTs, that offers a more resilient, governed, and leader-friendly path forward.

Why Legacy Bots Break at Scale

Most production chatbots today are built on some combination of three familiar patterns:

• Intent trees that route customers through predefined paths.
• Keyword matching layered onto structured flows.
• Rigid dialog orchestration optimized for predictable requests.

These approaches worked reasonably well when customer needs were narrow and transactional: checking order status, resetting a password, or updating an address.

But modern contact centers deal with something very different.

Customers arrive with partial information, emotional context, and multi-step problems. They expect systems to remember what they said five turns ago, adapt when plans change mid-conversation, and recognize when self-service is no longer appropriate.

Legacy bots struggle – and break – because they were designed around classification, not reasoning.

At scale, contact center leaders see the symptoms clearly:

• Intent libraries explode as teams try to model every variation.
• Conversation flows become brittle and hard to maintain.
• Escalation rates rise despite constant tuning.
• Automation teams spend more time maintaining bots than improving outcomes.

These are not tuning problems. They are structural limitations.

Why Better AI Doesn’t Fix A Broken Design

To address these issues, many organizations have embedded more advanced large language models (LLMs) into existing bot platforms. But while surface-level understanding improves, sustained resolution does not.

Why?

Because the surrounding architecture still assumes:

• One centralized decision engine.
• Static intent definitions.
• One conversational flow that is responsible for everything.

LLMs excel at flexible reasoning. But when constrained by brittle orchestration layers, their intelligence is throttled.

The legacy orchestration layer—specifically the centralized dialog manager and the intent-routing engine—forces the model to behave like a smarter classifier rather than a reasoning assistant. Leaders often interpret this as an AI maturity problem. In reality, it’s a design mismatch.

The technology has evolved. The architecture has not.

The Micro-GPT Model: Smaller Scope, Bigger Results

Micro-GPTs represent a fundamentally different approach to conversational automation.

Instead of deploying one general-purpose chatbot responsible for everything, this model breaks automation into purpose-specific agents.

In this article, “micro-GPTs” refers to a software architectural pattern for building governed, retrieval-grounded conversational assistants, not a specific product or vendor solution.

A micro-GPT is not a smaller model. It is a bounded system, defined by these four core principles.

1. Domain-bounded. Each micro-GPT operates within a clearly defined problem space: billing disputes, shipping issues, service eligibility, plan changes. Narrow scope reduces ambiguity and improves accuracy.
2. Retrieval-grounded. Responses are generated using approved knowledge sources—policies, procedures, FAQs, and structured data—not free-form guessing.
3. Policy-guarded. Business rules, compliance constraints, and escalation thresholds are enforced explicitly, not inferred probabilistically.
4. Composable. Multiple micro-GPTs can collaborate or hand off context, allowing conversations to evolve without collapsing into a single, monolithic flow.

This architecture mirrors how contact centers already operate: specialized teams, governed processes, and clear accountability.

How Micro-GPTs Improve Resolution While Keeping Control

For contact center leaders, the appeal of micro-GPTs is not novelty. It is control.

Loss of control is one of the most common executive concerns surrounding generative AI.

Traditional bots force organizations to choose between flexibility and governance. Micro-GPT architectures eliminate that tradeoff by embedding control at the system level rather than the dialog level.

This FIGURE provides a high-level comparison of legacy chatbot architectures and micro-GPT–based systems.

With micro-GPTs leaders gain:

• Higher first contact resolution (FCR) through constrained reasoning.
• Lower maintenance overhead by updating knowledge instead of retraining intents.
• Predictable behavior under stress, with safe failure and clean escalation.
• Clear ownership tied to operational domains.

For leaders accountable for both customer experience (CX) and operational risk, these attributes matter far more than raw model sophistication.

If your bot needs constant tuning, you don’t have an AI problem: you have a design problem.

Governance is the Difference

Loss of control is one of the most common executive concerns surrounding generative AI. Ironically, micro-GPT architectures improve governance rather than weaken it.

Because each micro-GPT is policy-guarded and retrieval-grounded, organizations gain:

• Transparent decision boundaries.
• Auditable response logic.
• Explicit escalation triggers.
• Consistent compliance enforcement.

Instead of asking, “Why did the model say this?” leaders can ask, “Which policy, source, or boundary was applied?”

That shift becomes critical as regulatory scrutiny increases and boards demand clearer accountability for automated decisions.

A Practical Migration Plan

For organizations with deep investment in legacy bot platforms, replacing everything at once is neither realistic nor necessary. Micro-GPTs can be introduced incrementally.

A pragmatic migration approach looks like this:

1. Identify high-friction interactions. Focus on use cases with high escalation rates or persistent customer dissatisfaction.
2. Define clear domain boundaries. Be explicit about what each micro-GPT application can and cannot handle.
3. Ground responses in authoritative knowledge. Connect agents to approved policies, procedures, and data sources.
4. Wrap automation with governance. Define escalation rules, confidence thresholds, and handoff logic up front.
5. Integrate alongside existing systems. Measure impact before expanding scope.

This approach reduces risk, preserves prior investment, and delivers visible wins that build executive confidence.

The Path Forward

The next phase of contact center automation will not be defined by larger models or more aggressive deflection targets. It will be defined by architectural maturity: systems designed to reason within constraints, collaborate across domains, and operate transparently at scale.

Traditional bots force organizations to choose between flexibility and governance. Micro-GPT architectures eliminate that tradeoff…

Micro-GPTs are not a silver bullet. But they represent a decisive shift away from brittle, monolithic bot designs toward automation that aligns with how contact centers actually work.

For leaders planning their inbound and automation strategies, that shift may be the difference between incremental improvement and meaningful transformation.

I recently spoke with the head of a contact center at a very large company that’s been around since before mobile phones.

When I asked what his key performance indicators (KPIs) for the year were, the answer was a roundabout “none.” This executive didn’t track any metric consistently on his contact center spend. A quick scan of online reviews of his company made me believe him.

Being unorganized breaks when scale and complexity rise or when the operations go through a platform shift. And we’re in the middle of just that.

Voice AI: The Platform Shift

Voice AI is triggering an enormous operating model change inside both existing and new contact centers alike:

• Established teams are redesigning workflows.
• In some organizations, new stakeholders are now accountable for customer conversations (Product, Engineering, Ops, Finance).
• Day-to-day decisions are happening faster and at higher volume.

The bar for measurement is rising because small failures can now scale instantly. In this environment, intuition isn’t enough. You need a measurable operating system.

Pick one KPI you’d bet your job on. Don’t worry, you can change it as your priorities shift. But you need focus now.

There’s a saying in Silicon Valley: startups = growth. In other words, startups are containers around growth.

Similarly, I think, contact centers are containers around customer success. We should borrow operating principles from Silicon Valley and apply them to contact centers, especially during this voice AI platform shift.

Pick one metric that defines winning, add a few guardrails that keep you honest, and run weekly goals like the best AI teams to drive compounding improvement. I will now explore these in detail.

Set a Primary KPI

A. Redefine KPIs

KPIs are a small set of quantitative metrics that tell you how “healthy” your operation is. The “health” part matters because it’s easy to confuse activity with progress.

KPIs force you to confront whether performance is improving or if you’re just keeping busy.

Setting KPIs is important for two reasons:

1. Obtaining objective truth

Numbers don’t lie: assuming you’re measuring the right thing(s). They keep you realistic about where you are.

2. Creating a feedback loop and prioritization

These tell you whether changes you make (policy tweaks, routing, training, tooling, voice AI iterations) are working and what to do next week. If you set them wrong, you can steer into circles.

B. Know what makes a primary KPI “good”

A good primary KPI should do these two things:

1. Quantify value delivered (to customers and to the business).

2. Be a usable feedback mechanism (moves frequently enough that you can act on it).

If a metric is easy to move but doesn’t represent value, it’s a vanity metric.

C. Know what are bad KPIs (and why they fail)

Before you pick your KPI “North Star,” it helps to see common pitfalls, especially during voice AI rollouts, such as these.

1. “% automated” or “# of AI calls handled”

Why it’s tempting: it’s easy to measure and looks like progress.

Why it’s bad: it measures activity, not outcomes. You can increase automation while customer success declines.

How it breaks: teams over-optimize containment; escalations become more frustrated; repeat contacts rise.

2. CSAT/NPS as the steering wheel

Why it’s tempting: executives recognize it.

Why it’s bad: it’s lagging, noisy, and biased. It’s great as a check engine light, not as a steering wheel.

How it breaks: teams argue about survey methodology instead of fixing the underlying operation.

Rule of thumb: if the metric can go up while customers get worse outcomes, it cannot be your primary KPI.

D. Select better KPIs

Pick one KPI you’d bet your job on. Don’t worry, you can change it as your priorities shift. But you need focus now. Too many KPIs cause fatigue, debates, and eventually non-use of the system.

Your primary KPI should:

• Move often (weekly is ideal).
• Tie directly to your business outcome.
• Capture what success looks like to you for the near future.

If you pick right, it shouldn’t feel reductive; it should feel focused.

Install Guardrails

Once you’ve chosen a primary KPI, pick two-three secondary health metrics. These are your operational guardrails: metrics that prevent you from “gaming” the primary number. Think of them as your contact center version of product “retention” and “quality” metrics.

Here is an example of a strong default set:

1. Repeat Contact Rate (7–14 days)

If this rises, your “resolution” is probably brittle.

2. Handoff Success Rate (a.k.a. escalation friction)

Measure whether escalations land cleanly: the human receives full context, the customer doesn’t have to repeat themselves, and time-to-first-human stays within your SLA.

3. Critical Defect Rate (compliance errors/over-refunding/etc.)

For refunds, cancellations, identity verification, regulated disclosures, or anything that can create liability.

Create an Operating Loop

A. Set weekly goals

There’s another Silicon Valley saying: do things that don’t scale.

In contact centers, that might look like:

• Listening to 20 calls yourself this week.
• Spending an hour rewriting the escalation policy yourself for the top two intents.

These aren’t long, glamorous projects with names, but they move the needle.

B. Aim for 1% – 2% weekly improvement

Set weekly goals for your primary KPI and target 1%-2% improvements. Those improvements compound (see CHART).

For example, a sustained 2% weekly improvement compounds dramatically over a year. Compare this to a large capstone project: an academic term that in business means a long, multi-team, big corporate effort that absorbs a lot of time: which doesn’t show any meaningful KPI movement until late, if at all.

That means you end the year at about 280% of baseline, without relying on a single giant capstone project. A small team (or even one operator) can often find a 2% win this week.

C. Draw the loop

1. Make the target visible

Draw a forward-looking chart for the next ~12 weeks. Put it somewhere the team sees weekly. Update it every week. Airbnb founders are famous for drawing it on the bathroom mirror of their offices, but you don’t have to go that far.

2. Set weekly prioritization

Every week, list the changes that could move the primary metric. Pick the one or two bets that are most likely to hit the weekly target.

3. Ship, measure, learn

Voice AI makes iteration fast if you have instrumentation in place to analyze calls and outcomes (intent tags, outcomes, recontacts, escalation reasons, and failure clusters).

Voice AI is bringing about not just a staffing change but a new operating model.

Early course correction and six months of focused engagement can pay dividends for years. You’ll build operational muscle for all members in your organization. You should also review your primary and secondary KPIs monthly to make sure they reflect the stage of business you’re in now.

When You Miss…

A. Constantly address the limiting factor

Missing a week (or two) is fine if you know why. Use “5 Whys” until you reach something actionable. 5 Whys is a simple root-cause method where you ask: “Why did this happen?” repeatedly (usually 5 times) until you get from the symptom to a specific, fixable cause. For example:

Problem: First Contact Success dropped from 62% to 54% this week.

1. Why did First Contact Success drop?

Because more calls were escalated to humans mid-flow.

2. Why were more calls escalated?

Because the voice AI couldn’t complete identity verification reliably.

3. Why couldn’t it complete identity verification?

Because the verification timed out so the agent started failing.

4. Why did the API time out?

Because traffic spikes caused rate-limiting and higher latency.

5. Why did rate limiting hurt us so much?

Because we had no retry mechanism and no fallback path (e.g., SMS verification).

Early course correction and six months of focused engagement can pay dividends for years.

So, the actionable fix is to ask the engineering team to add retries and a fallback verification path. Having proper instrumentation and call analytics in place helps you diagnose issues at startup speeds.

B. Understand if it is a metric problem versus an execution problem

When a KPI doesn’t move, it’s usually one of two things:

1. Metric problem: the definition is wrong or doesn’t reflect value.

2. Execution problem: you’re measuring correctly but not addressing the constraint.

It’s important to know what to fix.

C. Realize that instrumentation matters more in the voice AI era

Call analytics are no longer optional. Modern large language model (LLM)-based systems make it feasible to:

• Auto-tag intents and outcomes.
• Cluster failure patterns.
• Quantify policy violations.
• Surface the 20 calls that best explain why KPIs moved.

Finally…

Voice AI is bringing about not just a staffing change but a new operating model. The teams that win will treat their contact center like a modern growth organization:

• One metric that defines winning.
• A few guardrails that keep optimization honest.
• Weekly targets and a repeatable operating loop.
• Fast diagnosis powered by call analytics.

We can take lessons from the Silicon Valley playbook to improve our organizations and get compounding gains in customer success.

For years, many organizations treated customer service as a break-fix function. A customer had a problem, the contact center solved it, and efficiency meant speed, volume, and cost control.

However, with the advancement and incorporation of AI, that definition is much too small. World-class organizations are instead empowering teams to act as customer success advocates, focusing their efforts on proactive contact to develop more robust relationships between customers and brands.

In a generative model, every interaction contributes to future value…improving the entire customer lifecycle.

This work extends past fixing immediate concerns and leans into strengthening relationships, protecting account health, encouraging repeat business, and identifying opportunities for long-term growth.

Technology will accelerate this change, but the great differentiator is the human element, such as insight, empathy, and judgment.

The Four Stages of Service Evolution

This shift begins with culture and a firm understanding of CRM management. Service, and the CRM systems that support it, evolves through four distinct stages that map to when and how customer engagement occurs, which are before, during, and after contact.

1. Reactive Service: The Starting Point

Reactive service fixes what breaks. Customers reach out when something goes wrong, and agents work through queues that are focused on responsiveness and throughput. The problem is that these environments often reward closing the interactions rather than closing the loops.

Reactive service, by definition, starts after the customer has already experienced friction. Even when the agent resolves the issue correctly, the customer has still paid a tax in time, uncertainty, and emotional energy.

2. Predictive Service: Anticipating Needs

Predictive service operates before and during customer contact; it anticipates what is likely to break. Leaders treat service demand as something that can be understood and forecasted, not merely endured.

Before contact, organizations use data, AI models, and historical patterns to forecast likely issues, identify at-risk customers, and prepare the right responses.

During inbound or outbound interactions, predictive systems surface context in real time, helping agents understand intent, next-best actions, and potential outcomes.

AI is changing what CRM intelligence looks like before a customer interaction ever happens. Modern systems can synthesize account data, usage patterns, and prior interactions to give teams a deeper understanding of customer context before the customers ever experience friction.

3. Proactive Service: Preventing Disruption

Proactive service prevents issues from becoming disruptions. Instead of waiting for customers to report problems, proactive organizations intervene earlier. They correct errors before they trigger callbacks and deploy fixes before failures become outages.

The metrics that matter are the ones that measure avoided friction, such as repeat contacts, reopens, transfers, and customer effort.

4. Generative Service: Creating Future Value and Deepening Relationships

Generative service operates after and across interactions, using insights from every touchpoint to create future value and opportunities to strengthen the bonds with the customers.

It transforms the service from a moment of resolution to a continuous feedback engine that informs product design and policy decisions and improves digital journeys. This requires leaders to treat every interaction as data and an emotional touchpoint, rather than just a ticket.

In a generative model, every interaction contributes to future value, ensuring the organization focuses on high- context, relationship-building work and improving the entire customer lifecycle.

These stages are not isolated; they build on one another. Organizations typically operate across multiple stages at once. But maturity is defined by how far upstream they can move: from reacting to issues to predicting, preventing, and ultimately learning from them.

Overcoming the Barriers to Growth

Several common barriers can stall progress:

• Service Delivery versus Service Learning. Many contact centers produce insights but have no clear paths to operationalizing them or have owners accountable for translating what service learns into changes in product or policy.
• The Cost Center Mindset. As long as service is framed primarily as a cost center, investments in prevention and insight generation will always compete against near-term efficiency mandates.
• Metric Mismatch. If metrics prioritize speed over ease, and volume over value, the organization will optimize the wrong things.
• The Human Barrier. Service evolution asks leaders to protect employee capacity and emotional energy, not just productivity. You cannot scale proactive or generative service on a burned-out workforce.

The Path Forward

A practical way to lead this journey is to establish a roadmap and treat each stage as a set of explicit requirements.

• In the reactive stage, the cultural need is clarity. Customers want competence when something goes wrong.
• In the predictive stage, the cultural need is foresight and preparedness. Leaders must trust data enough to act on it, equipping teams with the tools and context to anticipate needs rather than simply responding to them.
• As you move to the proactive stage, the cultural need shifts to curiosity and cross-functional collaboration. There is also the need to build customers’ trust, like ensuring that the contacts are not seen by them as fraudulent (see BOX below).
• In the generative stage, success requires trust in service insights and trust in frontline judgment.

Proactive and predictive customer service, enabled by disciplined CRM management, is how the promise of experience-led growth becomes operational.

CRM is the mechanism that connects proactive service to business outcomes by linking customer context to predictive insight and orchestrating timely intervention.

Experience-led growth does not begin with selling more. It begins with listening, anticipating, and acting before customers are forced to ask, accomplished by integrating humans with technology. For this is how you build successful and valuable customer relationships.

Over the past decade, texting has become an essential support channel for modern contact centers. For good reason: it’s fast, convenient, asynchronous, and allows you to connect with customers via a device that’s rarely out of their reach.

But today, audiences expect a richer, more immersive, and more trustworthy texting experience. Fortunately, there is a newer protocol called rich communication services (RCS) that promises to deliver just that.

Here’s why contact center leaders should familiarize themselves with RCS as soon as possible.

What Is RCS?

RCS is a communications protocol that powers more interactive and modern texting experiences directly within a contact’s native messaging app. It supports high-resolution images and videos, customizable buttons, typing indicators, read receipts, and more.

RCS rivals the look and feel of conversations in over-the-top (OTT) messaging applications like WhatsApp without requiring users to download anything new. And since you can communicate more information (and even share files), it can help increase agent efficiency while slashing resolution times.

Beyond those engagement and productivity-boosting features, RCS also helps tackle one of text messaging’s biggest weaknesses: earning a recipient’s trust.

Thanks to verified sender profiles, where senders have been verified and approved by the individual telecom carriers and Google, messages arrive via a branded RCS agent, which displays a brand’s logo, colors, and a verified checkmark (rather than a random, faceless phone number).

In an era of rising impersonation scams and text-based fraud, having this visual assurance can give contact center agents instant credibility.

RCS…supports high-resolution images and videos, customizable buttons, typing indicators, read receipts, and more.

Plus, RCS lets you track more performance metrics than traditional texting. In addition to delivery stats, you’ll have insight into behavioral analytics like opens, clicks, and conversion rates. This can be incredibly useful when you want to determine which types of message content drive the best results.

Why Is RCS Suddenly So Important?

We say RCS is newer because, although it’s been around for a while (mobile users in the U.K. and EU have been using RCS for almost a decade); the North American rollout (U.S. and Canada) has been much slower.

Fortunately, now that major carriers have aligned on standards and Apple has given the green light to RCS with the release of iOS 18, we’ve seen a massive wave of adoption.

Additionally, as businesses clamor to adopt the new protocol and beat out their competition, it has recently become a popular topic of conversation in marketing circles.

In other words, if it seems like everyone is suddenly talking about RCS, you’re not imagining it. And we expect interest to grow even more in the second half of 2026 and into next year.

How RCS Compares to Other Channels

As a contact center pro, you’re probably wondering why you need RCS when you already have SMS texting, email, voice calls, and in-app messaging at your disposal. Do you really need to train your agents on yet another communication tool?

The truth is that RCS isn’t a new channel to master, nor is it a replacement for any of the channels you’re already using. Instead, it’s a complementary method and an essential part of any modern omnichannel strategy.

Here’s how it compares to other channels.

SMS

SMS is the universal texting protocol, meaning it’s supported by nearly every carrier and mobile device worldwide. But while SMS supports only text-based messages, as I noted earlier, RCS also supports high-quality visual content, more interactive features, and verified brand profiles (also see FIGURE).

Email

Email can be useful for communicating a large amount of information at once, but texting drives significantly higher open rates. And because RCS also allows you to include attachments, it’s an excellent tool for real-time conversations and more actionable follow-ups.

Voice

There’s no denying voice calls are still essential for those more emotion-driven, high-stakes moments. But in situations where customers are looking for a more asynchronous experience (without hold times), RCS offers contact centers a low-cost, high-engagement way to meet customer needs.

In-app messaging

Apps like WhatsApp and Facebook Messenger can be great for supporting customers who use those tools regularly. But RCS delivers equally immersive experiences for contacts who prefer their native texting apps.

SMS + RCS: The Future Is Hybrid

As RCS gains traction in the business world, more of our customers have been asking whether it will replace SMS. And our constant refrain is loud and clear: no. Not anytime soon.

Although RCS coverage has grown substantially over the past couple of years, it’s still not supported by every carrier and device.

So, if you were to rely entirely on RCS for texting, you’d inadvertently limit your reach. (We imagine the customers who still depend on SMS wouldn’t be too happy to discover they could no longer engage with support teams via text.)

Since SMS is universal, it’s important that you don’t count it out. Instead, we always recommend a hybrid approach where you set up SMS as your fallback.

Here’s what that looks like in action:

Suppose an agent supports a customer via a voice call and promises to follow up with some additional documents sent via text. They message the customer via RCS with an image carousel, and each image links to a separate PDF.

If the customer can’t receive RCS messages, the system will fall back to SMS and send the message in text-only format.

Depending on the texting platform your contact center uses, you may be able to set up a fallback version of the message. For example, since an SMS recipient wouldn’t be able to receive the carousel linking to PDFs, you might include a link to a page that contains all of the files.

SMS fallback then serves as a safety net to ensure everyone can receive your message, regardless of their device or carrier.

RCS Challenges

RCS offers some incredible features and benefits, but it isn’t perfect. Just like all communication methods, it has a couple of limitations and drawbacks.

Availability and reach

RCS adoption is growing fast. But, because it’s not yet universally adopted, you will likely always have at least a small portion of your audience that can’t receive these messages. (Which is why it’s critical you use SMS as a fallback.)

Operational complexity

The visual richness you enjoy with RCS also takes additional effort and tech. Creating these assets often requires careful coordination with other departments, and you’ll also need to make sure you have the right texting platform in place.

Although RCS coverage has grown substantially over the past couple of years, it’s still not supported by every carrier and device.

Ideally, you’ll want to work with a provider that has expertise in both SMS and RCS and offers a user-friendly interface with a shorter learning curve (so you can get agents up to speed more quickly).

RCS Success Best Practices

Since RCS is still relatively new, contact centers are often unsure how to get started. Should you dive in headfirst or take a slower, more methodical approach?

The answer depends on your resources, immediate goals, and the clients you serve. But, in general, here are three things I always recommend.

1. Get your verified sender status ASAP

Generally, it takes eight to 10 weeks to get approval from the major telecom carriers and Google for your RCS agent; some carriers may take longer than others.

We always recommend that our customers begin the registration process as soon as possible, even if they aren’t quite prepared to launch an RCS campaign. This way, once they’re ready to begin using RCS in earnest, they can hit the ground running with a verified, branded presence.

2. Plan for a progressive rollout

As with any new communication method or channel, it’s a good idea to test your effort with one or two use cases first, such as order tracking, customer onboarding, or a short-term campaign.

This allows you to familiarize yourself with RCS and work out any kinks in your process before you commit to it across the board.

3. Invest in the right platform and infrastructure

A lot of a company’s success with RCS comes down to having the right tools and technology in place.

Choosing a reliable platform will make a tremendous difference in how quickly and easily you can spin up your RCS program. And a great RCS partner will help you handle the registration process and support you in addressing security and compliance, too.

For contact center leaders, the message is clear: RCS adoption is accelerating, and customers are hungry for richer, more engaging, and more trustworthy experiences. Organizations that invest in the right expertise and technology today will be best positioned to take the lead in the months ahead.

For decades, organizations have managed contact centers through a narrow operational lens. They have viewed them primarily as necessary cost centers, measured by their ability to resolve issues as efficiently as possible while keeping expenses under control.

Metrics such as average handle time (AHT), abandonment rates, and cost per interaction have long defined success. These measures remain important for maintaining efficiency. Yet they capture only part of the contact center’s true strategic value.

As customer expectations evolve and interaction data becomes central to business performance, forward-thinking organizations now recognize the contact center as something far more consequential. That is, a real-time engine for relationship intelligence, revenue influence, and enterprise-wide decision-making.

This shift – from cost center to growth center – is not simply about improving service. It reflects a broader rethinking of how organizations engage customers and how everyday interactions can create measurable business impact.

The Evolution of the Contact Center

Customer expectations have changed dramatically. Consumers now insist:

• On seamless and personalized experiences across every channel they use, whether that’s voice, chat, messaging, email, or social media.
• That companies remember their histories, understand their preferences, and resolve issues quickly without forcing them to repeat information or navigate disconnected systems.

Yet many traditional contact center environments were not designed to meet these expectations. Customer engagement data remains distributed across multiple platforms, making it difficult for agents to access a comprehensive view of the relationships.

As a result, interactions can feel transactional rather than continuous, focused on resolving the immediate issues rather than strengthening long-term loyalty.

But when organizations begin to unify interaction data and provide agents with a holistic understanding of the customer journey, the nature of the conversation changes.

From an organizational perspective, leaders must expand how they define success, balancing operational efficiency with measures that reflect relationship strength and business impact.

Service interactions become opportunities to build trust, reinforce brand value, and identify needs that might otherwise go unnoticed. This evolution is redefining the role of the contact center from a reactive, expensive support function to a proactive growth-enabling engagement hub.

Why Context Matters

At the center of this transformation is the concept of persistent context, which is the ability to carry forward relevant customer information across interactions, channels, and time.

Persistent context provides continuity. It allows organizations to move beyond isolated service events and engage customers based on a deeper understanding of their experiences and intent.

When data from multiple touchpoints, such as purchase histories, behavioral signals, channel preferences, and prior engagement history, is connected in real time, organizations gain a far more complete view of customer intent and journey progression.

This continuity allows every interaction to begin with understanding rather than rediscovery. Instead of rebuilding context at each touchpoint, teams and systems can move forward with shared intelligence that reflects the full relationship over time.

The result is a more connected customer experience (CX), one that reduces friction, strengthens trust, and creates the foundation for more intelligent decision-making across the enterprise.

AI as an Enabler of Better Decisions

AI is accelerating the shift from cost management to value creation, particularly when it is grounded in live interaction context rather than in static historical data.

AI can analyze large volumes of engagement signals in real time, helping organizations move from reactive service to proactive decision-making while the customer relationships are unfolding.

For agents, AI-driven tools can surface insights during conversations, recommend next actions, and automate routine tasks such as summarization or data entry.

By reducing administrative burden, these capabilities free them to focus on the human elements of engagement, empathy, judgment, and relationship building.

This dynamic has important implications not only for revenue but also for employee satisfaction.

Contact center roles have historically been associated with high levels of burnout, driven by repetitive interactions and pressure to meet narrow efficiency targets.

When AI augments human capability and enables more meaningful conversations, agent engagement and retention often improve. Organizations that embrace this approach are finding that empowered employees are better positioned to deliver positive customer outcomes.

The Rise of the Connection Center

As engagement strategies mature, many organizations are redesigning their contact centers around continuity rather than channel. In these emerging connection center models, customer context flows fluidly across touchpoints, allowing interactions to feel coordinated rather than fragmented.

In this environment, agents are no longer viewed solely as problem-solvers. They become relationship managers who guide customers through moments that matter, from resolving service issues to navigating complex decisions.

But achieving this vision requires both technological and cultural change.

From a technology perspective, organizations need platforms capable of integrating interaction data, orchestrating workflows, and delivering AI-driven insights across channels, providing unified context.

These capabilities enable a more connected experience in which engagement signals can inform immediate action with intelligent guidance.

From an organizational perspective, leaders must expand how they define success, balancing operational efficiency with measures that reflect relationship strength and business impact.

Turning Service Moments into Growth Opportunities

Once this foundation of unified context is in place, contact centers can begin turning high-intent service moments into measurable opportunities for retention, expansion, and growth.

One of the most immediate benefits is improved customer retention. Research consistently shows that customers who feel understood and supported are far more likely to remain loyal to a brand, even when issues arise.

Contact centers are also uniquely positioned to influence revenue through contextual recommendations. Because they interact with customers at critical moments, often when intent and attention are already high, they can introduce relevant solutions in ways that feel helpful rather than intrusive.

For example, a customer calling about a subscription renewal may benefit from a bundled offering that better aligns with their usage pattern. With the right insights, the agent can recognize this opportunity and respond accordingly.

The goal is not to transform service professionals into sales representatives. Instead, it is to empower them to recognize when additional value can be created. Over time, these incremental moments of alignment can contribute meaningfully to revenue growth.

Breaking Down Organizational Silos

Transforming the contact center also requires stronger integration with the broader enterprise.

Historically, service teams have often operated separately from marketing, sales, and product teams. Yet the insights generated through customer interactions can provide a powerful signal of emerging needs, sentiment shifts, and operational challenges.

When these insights are shared across departments, organizations gain a deeper understanding of how customers experience their brand in real time.

These insights can inform product innovation, demand forecasting, and CX design. In this way, the contact center becomes more than a support function. It enables a strategic feedback loop that helps guide business decisions.

Measuring What Matters

One of the most significant barriers to reimagining the contact center lies in how performance is measured.

Traditional metrics like AHT and call volume will always play a role in operational management. However, organizations seeking to position the contact center as a growth driver must also adopt measures that reflect long-term value creation.

These measures may include traditional indicators such as customer lifetime value, retention, and satisfaction trends.

They may also include more advanced outcome-based metrics such as goal completion rates, revenue influenced by service interactions, and sentiment improvement during conversations that reflect how interactions contribute to business performance.

Organizations that embrace the strategic potential of customer interactions are redefining the role of the contact center within the enterprise.

Increasingly, organizations are evaluating revenue influenced by service engagements, the effectiveness of AI-supported resolutions, and the degree to which human and automated interactions work together to achieve successful outcomes.

Emerging performance frameworks also examine indicators such as goal completion rates, the cost required to achieve a successful resolution, real-time sentiment movement during an interaction, and the level of autonomy achieved before human intervention is needed.

In parallel, many leaders are tracking agent engagement and development metrics to better understand how technology augmentation affects workforce performance and experience.

By expanding measurement models to focus on value creation rather than activity alone, organizations can align incentives around strengthening customer relationships, improving operational decision-making, and enabling more sustainable revenue growth.

The Future of Customer Engagement

The contact center is undergoing a profound transformation. What was once viewed primarily as a cost center is becoming a strategic engine for connection and growth.

Advances in data integration, real-time analytics, and AI-enabled collaboration are accelerating this evolution, alongside rising expectations for more personalized and responsive experiences.

Organizations that embrace the strategic potential of customer interactions are redefining the role of the contact center within the enterprise.

Rather than treating engagement as a series of isolated transactions, they are building environments designed to turn everyday conversations into moments of understanding, trust, and value creation.

In a world where CX increasingly shapes loyalty and differentiation, the ability to turn everyday interactions into moments of connection may be one of the most powerful competitive advantages an organization can achieve.

As AI becomes more deeply embedded in customer service and contact centers, many organizations frame the conversation around efficiency: faster resolution, lower costs, fewer agents.

But that framing misses both the real risk and the real opportunity facing brands that serve large, price-sensitive customer bases.

The question isn’t whether AI will transform customer service. It already has. The question is whether companies will use it simply to check a box or to build trust and advocacy in an environment where good service is becoming increasingly rare.

The Less Service Myth for Mass-Market Customers

Much has been written about the emergence of a K-shaped economy where the middle class is hollowing out, wealth is concentrating at the top, and most consumers fall into increasingly price-sensitive tiers.

That shift has influenced how many organizations think about customer service. High-value customers are seen as deserving high-touch experiences, while mass-market customers are routed toward speed and automation.

But customers don’t experience this as efficiency. Instead, they experience it as a difference in care.

There’s a common assumption that customers outside the top income tiers expect less when it comes to service. In reality, they want less effort. They’re perfectly comfortable solving simple issues on their own: as long as the experience is fast, clear, and reliable.

This is where AI excels. Straightforward requests, such as checking an order status, updating account information, or resolving basic issues, can and should be handled immediately through self-service.

In fact, Zendesk reported that 51% of consumers say they prefer interacting with bots over humans when they want immediate service.

When done well, automation respects the customer’s time and reduces friction across the system. But we see problems emerge when automation is treated as an endpoint rather than part of a broader service strategy.

When “Checking the Box” Breaks Trust

Everyone has a story about bad customer service, and those stories tend to stick. Importantly, 95% of consumers say that customer service has an impact on their brand loyalty.

And as AI becomes more widespread, truly good customer service may become even more scarce. It’s not because the technology fails, but because it’s implemented without judgment.

Too often, AI systems are designed to follow rules rather than produce desirable outcomes. They confirm requirements, adhere to policy, and complete workflows. Yet they can sometimes miss the broader context of what the customer actually needs.

Humans Must Own the Outcomes

One of AI’s greatest strengths is consistency. Well-designed systems can reduce routine errors, surface patterns in customer issues, and ensure that simple requests are handled accurately every time. But accountability for the customer experience (CX) doesn’t disappear just because a system is automated.

Ultimately, humans are still responsible for outcomes. That’s why AI must be designed with humility as well as intelligence.

An effective AI experience recognizes when it has reached its limits. There needs to be a point where the chatbot or other AI tool can say “I don’t have the answer, but I can get you to someone who does,” before turning the interaction over to a human.

That handoff is critical. Customers don’t just want resolution; they want reassurance that a real person is available when the situation becomes complex, emotional, or high-stakes.

When escalation is difficult or opaque, frustration rises. But when it’s seamless and intentional, automation becomes an asset instead of a barrier.

The Right Work for Automation

The most effective customer service models don’t ask whether AI or humans are better. They ask where each belongs.

Simple, repeatable interactions should be automated. More complex moments, which are those involving judgment, exceptions, or trust, should escalate to people quickly and cleanly. The mistake many organizations make is treating escalation as failure rather than as a core feature of good service design.

This is especially important at scale. When customers feel trapped in automated loops with no clear path to a human, dissatisfaction grows rapidly. But when escalation is designed thoughtfully, AI helps manage volume while people focus on the moments that matter most.

…problems emerge when automation is treated as an endpoint rather than part of a broader service strategy.

Leaders need to look closely at their data. Examining call resolution rates, repeat contacts, and escalation drivers can help make more deliberate decisions about what to automate and what to protect as human-led work.

Why People Still Win the Retention Game

As AI becomes more accepted across industries, it will play an increasingly visible role in customer care and contact centers.

Over time, certain interactions may be handled end-to-end by intelligent systems. However, in the foreseeable future, people still win the customer retention game. Especially when they’re placed intentionally and supported properly.

The next evolution of customer service won’t be defined by how much is automated, but by how well automation and human judgment are orchestrated.

High-volume, low-complexity interactions will continue to move toward self-service. Moments that involve trust, exceptions, or emotion will demand faster escalation to empowered people.

At the same time, AI will increasingly operate behind the scenes. Quality assurance (QA), coaching, and performance feedback will become continuous and data-driven. Instead of replacing agents, technology will help develop them, raising the overall standard of service across contact centers.

Looking ahead, the brands that stand out won’t be the ones that deploy AI the fastest. They’ll be the ones that design customer care intentionally, using automation to remove friction, humans to build trust, and data to ensure the system learns over time.

In a marketplace where customers quickly forget average service and never forget bad service, the ability to combine efficiency with accountability may become the defining advantage.

Every few years, someone predicts the end of offshore BPOs:

• Sometimes it’s when a new technology gains momentum.
• Sometimes it’s when political pressure ramps up to “bring jobs home.”

Today, both forces are happening at once. And at first glance, you might assume the combination makes offshore support less attractive.

The reality is more complicated. AI/more task automation makes onshore service delivery more efficient. But it is also making offshore delivery stronger and more capable. Their value doesn’t disappear; instead, it shifts.

Here’s why. Offshore BPOs benefit enormously from AI that is fast to implement, easy for agents to adopt, and customizable to the nuances of complex service interactions.

The future of BPOs isn’t determined by where agents sit. It’s determined by how quickly and effectively BPOs can adapt to a world where more interactions are automated, complex interactions continue to grow, and political pressure creates noise that does not always align with economic reality.

So, the key question isn’t, “Do offshore BPOs have a future?” It’s, “Which BPOs will evolve fast enough to remain essential in a blended human + AI service model?”

From Labor Arbitrage to Augmented Expertise

Before AI, contact centers had two basic levers for improving service:

• Hire more people.
• Improve training and processes.

AI introduces a third, far more scalable lever: augmenting the people you already have.

Recent figures from McKinsey suggest that modern technologies could automate roughly 57% of U.S. work hours.

They also emphasize a critical point: automation replaces tasks, not the human skills that define service interactions. Emotional intelligence, complex task management, active listening, and contextual reasoning remain essential, and these skills are what contact center agents perform every day.

What AI really changes is which tasks humans spend their time on. Instead of toggling across systems or decoding policies mid-call, agents can now focus on the conversations and decisions that actually shape the customer experience (CX).

This shift increases the value of agents, both onshore and offshore, in handling:

• Multi-step reasoning.
• Complex scenarios.
• High-emotion interactions.
• Compliance-heavy requests.
• Fraud-sensitive workflows.

These scenarios require agents to track large amounts of information, follow nuanced workflows, and quickly interpret context.

Driving this shift is that customer journeys have become more fragmented, with interactions now spanning five or more channels, up from two or three just a few years ago.

…“Which BPOs will evolve fast enough to remain essential in a blended human + AI service model?”

That means agents must manage more complex scenarios, more exceptions, and more tools: thereby pushing cognitive loads beyond what humans can reliably maintain at scale.

This rising complexity is exactly what sets up the case for operationalizing AI in the call center, which I will discuss later in this article, both onshore and offshore.

AI Removes Offshore Friction, Not Jobs

At first glance, it’s natural to assume that if AI makes agents faster and self-service more capable, offshore BPOs might lose ground.

Offshore centers often carry a reputation (fairly or not) for slower navigation, communication friction, and inconsistent outcomes tied to system or knowledge limitations.

Historically, their agents faced disadvantages unrelated to talent:

• Unfamiliar or fragmented systems.
• Heavy multitasking under pressure.
• Unclear workflows or exception paths.
• Inconsistent policy documentation.
• Linguistic phrasing or formatting differences that slowed clarity.

Offshore programs often scale rapidly, and training cannot keep pace with constant updates, policy nuances, or new exceptions. Fragmented systems introduce extra seconds (or minutes) into every step. Also, cultural or phrasing differences can slow comprehension even when English proficiency is strong.

All of this lands in a service environment where customer frustration is breaking records. According to The Wall Street Journal, 77% of customers now report experiencing a service problem. And their biggest complaints mirror these very friction points:

• Long waits.
• Unclear processes.
• Interactions that feel harder than they should.

Over thousands of interactions, those frictions add up to longer handle times, higher variability, and greater cognitive strain.

This is exactly where AI shifts the equation. It reduces or eliminates many of these friction points. Modern agent-assist tools can:

• Auto-retrieve information across systems.
• Guide next steps in real time.
• Summarize context instantly.
• Reduce cognitive load.
• Ensure consistent adherence to policy.
• Flag risk before it escalates.

In effect, AI levels the playing field by removing the operational barriers that once created gaps between onshore and offshore teams.

As interactions grow more complex and journeys become more fragmented, organizations increasingly need tools that help agents manage volume, variability, and multi-system complexity.

Companies using AI-assisted workflows see meaningful improvements in speed and accuracy: gains that are essential for global operations where consistency is paramount. Customers care less about where an agent sits and more about whether their issues are resolved quickly, clearly, and accurately.

Onshoring Politics vs. Service Economics

Political interest in reshoring contact center and IT support work has grown, especially as concerns about data security and AI-driven fraud increase.

• In 2025, Congress renewed calls for reshoring and limiting offshoring through the proposed “Keep Call Centers in America Act,” which also signaled a desire for more oversight of AI-driven interactions.
• In March 2026, the FCC issued a notice seeking comments for proposed rules on reshoring contact centers, requiring standard American English by agents, and on combatting illegal calls from other countries.

But the practical underlying economics haven’t changed.

There’s a reason countries like the Philippines have a $30 billion call center outsourcing economy: cost matters. And as companies absorb higher costs from tariffs, supply-chain instability, and rising labor rates, those pressures inevitably push them toward more affordable service models.

Customers also expect more from service interactions: faster resolution, clearer answers, and greater personalization. Expectations are higher, but the price they expect to pay is not. Offshore BPOs allow companies to meet those expectations and manage cost.

With AI, offshore centers can now deliver higher-quality interactions at lower cost, making them more strategically valuable, not less. This is why offshoring persists. It’s not politics. It’s affordability, capability, and CX economics.

The future mix will be:

• Onshore for regulated, supervisory, or specialized work.
• Offshore for scalable, AI-augmented service.
• AI plus self-service for high-volume repetitive tasks.

The result isn’t a contraction of offshore BPOs. It’s a rebalancing of which organization handles what, driven by capability, cost, and risk profiles rather than politics alone.

Operationalizing The AI Value Proposition

Brands aren’t buying people per hour anymore. They’re buying measurable outcomes. Clients now expect:

• First contact resolution (FCR) improvements.
• Lower error rates.
• Stronger fraud detection.
• Multi-system consistency.
• Predictive insights.
• Rapid adaptation to policy changes.
• Hyper-accurate compliance tracking.

These are not “seat” deliverables. These are capability deliverables. And delivering them requires more than just having AI tools licensed somewhere in the stack.

AI doesn’t eliminate offshore BPOs. It elevates expectations and their ability to meet them.

In many contact centers, in-house and BPO alike, agents work in tech environments that grew over time like a game of Jenga: a CRM here, a ticketing system there, loyalty and POS systems built for other parts of the business. All held together by fragile APIs and swivel-chair work.

But ripping and replacing those systems is slow, risky, and expensive. This is where the concept of operationalizing AI becomes essential. It means layering intelligent orchestration on top of that reality, not waiting for a perfect replatform.

In practice, that looks like:

• An AI “overlay” that is system-agnostic, pulling data from old and new tools without needing deep integrations.
• Real-time guidance that tells agents what to do next, rather than forcing them to hunt for answers across five different screens.
• Automation that quietly handles routine lookups, after-call work, and documentation so humans can focus on complex, emotional interactions.

For large, outsourced environments, this has to work across dozens of client systems, in high-turnover teams, and across multiple regions and languages.

AI needs to be:

• Deployed quickly.
• Governed continuously.
• Delivered to agents at the second they need it, even in outsourced environments where systems vary, integrations are limited, and workflows change frequently.

Operationalizing AI enables the above deliverables. And in complex BPO ecosystems it is becoming the true competitive differentiator. Offshore BPOs that embrace this are positioned to deliver all of these outcomes more effectively than ever before.

AI doesn’t eliminate offshore BPOs. It elevates expectations and their ability to meet them.

Final Takeaway: Offshore BPOs Don’t Disappear — They Transform

Despite political shifts.

Despite the rise of AI.

Despite self-service being a cost center.

Offshore BPOs remain essential, but they must evolve. The ones that survive will be those that shift from selling labor to selling intelligence, capability, and AI-enhanced outcomes.

The future belongs to BPOs that can pair AI’s precision with humans’ judgment, empathy, and accountability.

This combination will define the next generation of global service delivery and the next generation of customer trust.

Telecommunications spending has become significantly more complex for modern contact centers as customer expectations, technologies, and communication channels continue to evolve.

What was once a predictable, per-minute model has expanded into a mix of email, live chat, SMS, chatbots, and more. While this may enable better customer experiences (CXs), it also creates hidden inefficiencies that many organizations overlook.

To manage these costs effectively, contact centers must adopt a new approach.

Challenges of Managing Complex Costs

Modern contact centers offer flexibility and increased capabilities, but they also introduce a broader, more dynamic cost structure. Multiple channels of communication can mean multiple contracts stored across several different locations in a variety of formats.

Contact centers often struggle with two things: (1) a lack of clarity and (2) simplicity when managing telecom expenses, including managing licensing.

When an organization doesn’t have a clear understanding of its telecom rates and renewal timelines, it can lead to inefficiencies and unnecessary expenses.

For example, missing a contract renewal deadline may result in an automatic extension at an unfavorable rate. Yet some organizations still rely on spreadsheets or other traditional tracking processes.

These approaches are complex, time-consuming, costly, and not suited for modern operations. Service changes can occur faster than companies can update their records, making it difficult to track what information is accurate. These inefficiencies can add up quickly and ultimately drain budgets and productivity.

Additionally, a lack of organized, consolidated data also limits visibility. When contracts are scattered across inboxes and physical locations, and invoices arrive in multiple formats, identifying cost-saving opportunities becomes difficult.

For instance, disorganized data that spans several facilities with different operational needs could result in an organization being billed for services it no longer uses.

This challenge can be compounded when a company has separate teams handling contracts, operations, and payments who do not frequently communicate with one another.

Turning Data into Cost-Saving Opportunities

Effective telecommunications cost management starts with consolidating all data, including invoices, services, and contracts, into a centralized system.

This provides better visibility, accuracy, and strong operational control. It enables contact centers to move away from a reactive cost control strategy to more proactive management of their telecom expenses.

This holistic view will also reveal where charges are accurate, where they might be duplicated, and where services can be optimized. It allows organizations to begin identifying underutilized or redundant costs and contracts that no longer align with business goals.

Decision-makers can then use this information to pinpoint cost-saving opportunities (see BOX below) that would have otherwise remained hidden. Thus, resulting in a stronger alignment between financial and operational objectives.

By taking a close look at existing operations, identifying and addressing inefficiencies, it’s possible for companies to see a 2%–3% annual reduction in their telecom spending.

Taking Action

Engaging with a consultant can be helpful when facing the daunting task of assessing and overhauling telecom expenses, particularly for large organizations that might have siloed teams handling contracts and expenses.

These experts will bridge the communication gaps between departments, working with all relevant stakeholders to understand organizational goals.

They also have access to the latest data management, analysis, and security tools (also see BOX below) to organize and protect information, deliver actionable recommendations, and uncover cost savings.

Telecom expense complexities will never disappear, particularly for companies operating contact centers that rely on these technologies to remain operational. Technology will continue to advance, and customer expectations will evolve.

Taking control of expenses now with the right tools, processes, and expertise will unlock future opportunities for efficiency rather than frustrations.

Additionally, the long-term benefits extend beyond financial gains. More visibility into ongoing telecom costs can give finance, facility management, and operations teams greater confidence in their future planning and decision-making.

Customer interactions can be likened to railroad cars in a train being pulled by a locomotive. Each interaction, like the train, is on a journey, and each customer has a destination, like the cars that are carrying the products.

On locomotive-powered trains, couplers or drawbars enable the railroad cars to move forward by transmitting the energy from the locomotive(s) propulsion systems. Contact centers perform a similar function with customer interactions. The agents (human and now AI) pull the customers along.

But how the trains and contact centers perform depends on many factors: like the power of the companies that are moving them, operational environments, and the tightness of their connections. They share the challenges of avoiding splitting apart or, worse yet, derailments.

To find out how the customer interaction train is moving, to see how clear the track is, but also to look forward and respond to any troublesome issues, we had virtual conversations with our panel of industry supplier experts. They are:

• Sarita Fernandes, Vice President, Product Management, Avaya
• Lisa Orford, Global Vice President, Product Management, Contact Center, 8×8
• Sarika Prasad, Director, Product Marketing, Five9

Q. What changes, trends, and developments have you seen emerge over the past 12 months that impact customer contact, and what are their drivers?

SARITA FERNANDES

 

 

 

Over the past 12 months, customer contact has become much more dynamic. Organizations are moving away from fixed workflows and toward systems that can respond in real time, based on what’s happening in the interactions.

Customer expectations are a big driver. People expect companies to understand their histories, intent, and context across every channel. At the same time, enterprises are trying to connect fragmented systems and make faster decisions with better data.

As a result, customer interactions are becoming a much more valuable signal for each business. Each conversation reflects customer sentiment, intent, and the state of the relationships at that moment.

That shift is increasing demand for technologies that can work with live data. Approaches like model context protocol (MCP)-enabling platforms allow AI systems to access and act on real-time context, making interactions more relevant and responsive.

We’re also seeing more focus on how AI and human agents work together. The most effective models bring both into the same workflow, with AI supporting speed and scale while humans focus on judgment and customer experience (CX).

This is what we call “tandem care”, a model of service where AI agents and human agents collaborate in real time, each amplifying the strengths of the other to improve outcomes.

This coordination allows for seamless transitions between self-service and human agent assistance, ensuring that full context and progress are preserved while delivering an effortless customer and employee experience.

LISA ORFORD

 

 

 

The biggest shift I’ve seen is that customers have stopped being patient. They expect you to know who they are, remember what happened last time, and resolve their issues fast, regardless of which channel they’re using. And when you can’t deliver that, they leave. It’s that simple.

What’s driving it? AI has reset expectations. People have experienced genuinely helpful automated interactions, so now they hold every contact center to that standard.

The organizations that are struggling most right now are the ones with fragmented systems; they can’t deliver a connected experience because their data is scattered across tools that don’t talk to each other.

The contact center used to be a cost center you managed. Today it can make all the difference in a competitive market. The leaders who recognize that are investing in platforms that unify the CX. The ones who don’t are losing customers to competitors who have.

SARIKA PRASAD

 

 

 

Over the past year, contact centers have shifted from AI hype to operational reality. Teams are no longer just experimenting with AI. They are implementing it at scale and across the full customer journey.

AI solutions now go well beyond managing simple FAQs. They power intelligent triage, real-time quality management, and end-to-end self-service. Voice AI has matured dramatically. Interactions that once felt robotic now feel fast, natural, and contextually aware.

At the same time, younger customers are raising the bar. Customers expect AI-powered service as standard, with a clear path to a human agent when needed.

“As AI continues to evolve, its usage is going to go much further, becoming more mature, reliable, and more widely adopted…”
—Sarika Prasad

Yet technology alone is not enough. CX has become the defining battleground for brand loyalty.

• Businesses must create experiences that feel effortless, personal, and deeply human.
• There is a growing recognition that AI must augment, not replace, human empathy.
• Trust and governance around AI deployment are now boardroom priorities, not afterthoughts.

This shift is being driven by a fundamental change in mindset. Contact centers are a growth engine, not just an operational cost. When organizations get this balance right, combining smart automation with genuine human connection, they will not just resolve issues. They will build lifetime customer loyalty.

Q. What is the state and direction of AI in the contact center? Is it mature and reliable, or does it still have issues? Are contact centers realizing benefits from it and if so, where?

Further, what is AI being used predominantly for? Assisting live agents to improve the CX? Or to deflect and/or shorten live agent contacts and reduce the need for staff?

SARITA FERNANDES:

AI is delivering real value today, especially in areas like agent assist, summarization, and automating routine interactions. Many organizations are already seeing improvements in efficiency and productivity.

Where it still falls short is consistency. In a lot of environments, AI doesn’t have full visibility into the customers, the interaction histories, or what’s happening across systems. When that context is missing, the outputs can feel disconnected.

That’s why more of the focus is shifting to how systems connect and share information in real time. With MCP, AI can access live, contextual data during the interactions, improving both accuracy and relevance.

“…AI and human agents should be working from the same information. Clear handoffs and shared context go a long way in keeping the experience consistent.” —Sarita Fernandes

Contact center agents can now be more proactive with customer responses, as they’re able to leverage the power of AI to easily access a vast array of data points, such as customer profile and demographics, behavioral data, transactional data, and more.

Right now, the most common use cases are clear. AI is helping agents with real-time guidance and next best actions. It’s handling routine inquiries, directing requests to the right places, and turning conversations into structured data that the business can actually use and learn from over time.

As that happens, the connection between AI and human agents becomes more important. When both are working from the same context, transitions are smoother, and the experience stays consistent.

All of this also depends on having the right governance and data controls in place, so organizations can trust how AI operates and how customer data is being used.

LISA ORFORD:

AI is delivering real value, but not universally and not by accident.

Where it works well: agent assistance. Real-time coaching, post-call summaries, and surfacing the right information at the right moments. Agents handle calls faster, with more confidence, and customers feel the difference: faster resolution, less repetition, and fewer transfers.

Where AI falls flat is when organizations deploy it on fragmented infrastructure. A chatbot that doesn’t know what happened on the phone last week isn’t helpful; it’s just another thing customers have to work around. Those failures aren’t AI problems; they’re data problems.

“AI has reset expectations. People have experienced genuinely helpful automated interactions, so now they hold every contact center to that standard.” —Lisa Orford

The honest truth is AI will handle an increasing share of routine contacts. That’s not a threat; it’s an opportunity if you’re prepared for it.

It means live agents can focus on the interactions that actually benefit from human judgment and empathy. The contact centers winning right now are the ones using AI to make human interactions better, not just to reduce headcount.

SARIKA PRASAD:

AI in the contact center is maturing quickly, reshaping how we know it today.

AI is not just an automation process making back-office operations smoother; it is now a key business lever at the front lines of the customer CX journey. It’s highly effective at handling high-volume interactions, improving routing, and supporting agents with actionable insights in real-time.

That said, adoption rate and meaningful value are not the same thing. While over 80% of contact centers have adopted AI, according to our 2025 Business Leaders CX report, many are still closing the gap between deployment and real operational impact.

Unfortunately, urgency is outpacing readiness. Consequently, and not surprisingly, organizations rushing to implement without the right data infrastructure, integration maturity, or governance foundations are finding it difficult to scale beyond early pilots.

As AI continues to evolve, its usage is going to go much further, becoming more mature, reliable, and more widely adopted, making automated CX operations more efficient and seamless.

AI has multiple use cases within the contact center and across the CX journey, being most widely used to gather key information, quickly resolve routine tasks, and seamlessly route interactions to humans or other digital channels, when needed. So, organizations must get the implementation right to benefit.

The dominant model emerging is not AI replacing agents. It is automating where possible, with humans handling what matters most. AI handles volume. Humans handle value. Complex, emotionally nuanced conversations remain firmly in human hands.

Finally, how contact centers measure success is shifting. Handle time gives way to Customer Effort Score, Net Promoter Score, and real-time sentiment as the metrics that truly reflect AI’s impact on CX.

Q. What are your recommendations when choosing, deploying, and using inbound and outbound customer contact applications?

SARITA FERNANDES:

The biggest thing is to look at how everything works together, not just the individual tools, and to select a platform that’s open and does not lock you into a vendor.

A lot of organizations solve for specific use cases, but the systems don’t share context or connect well. That creates friction quickly, both for agents and customers.

Interoperability is key. Systems should be able to connect across channels, workflows, and AI models. Flexibility matters too, especially as the AI landscape continues to evolve. Supporting both private and public models helps avoid having to start over later.

Real-time context is another big factor. Applications perform better when they can understand what’s happening in the moment and adjust. More dynamic workflows tend to handle real-world variability better than rigid ones.

It’s also important to think about how interaction data is used. Every conversation is a signal. When that data is captured and used effectively, it helps the business make better decisions and spot trends earlier.

On the operational side, AI and human agents should be working from the same information. Clear handoffs and shared context go a long way in keeping the experience consistent.

Finally, deployment flexibility and governance still matter. Most organizations are balancing cloud and on-prem environments while meeting security and compliance requirements. The solution should support that without adding unnecessary complexity.

If those pieces are in place, it becomes much easier to build a system that can adapt and improve over time.

LISA ORFORD:

The biggest mistake I see is purchasing point solutions that solve today’s problem but create tomorrow’s headache. If your inbound, outbound, and AI systems don’t share the same customer data, every interaction starts from scratch. Unified data isn’t a feature: it’s the whole game.

Sequence your AI investments carefully. Automation works best on top of mature operations with real interaction data.

If you deploy AI before you understand your contact patterns, you’ll get low containment rates and frustrated customers. The data you collect in Year One is what makes AI actually useful in Year Two.

Finally, measure what customers experience, not just what agents do. Volume metrics tell you how your operation is running.

First contact resolution (FCR), customer effort, and loyalty tell you whether it’s actually working. Build your evaluation framework around outcomes from the start: that’s the only way to know if the technology is earning its keep.

SARIKA PRASAD:

Simply choosing a platform and setting it and forgetting it is no longer enough when managing customer contact applications.

For businesses to see true success, they need to have full visibility into their contact center operations, leaning on that data to develop informed, strategic plans based on their functions or product/service lines for CX decisions and deployment. As examples:

• Organizations that typically manage more complex interactions, such as financial services, healthcare, or government, need to focus their investment and automation strategy around information security, seamless escalation, and human agent empowerment.
• Organizations in the retail space must prioritize solutions that are built for managing high-volume inquiries, quickly and effectively.

To ensure customer contact platforms continue to align with company business goals and customer expectations, I recommend conducting regular audits, gathering real-time analytics, and regularly sourcing customer feedback and agent insights.

The more data leaders have into resolution, consumer behavior, and agent experience, the more successful they will be in optimizing their contact center strategies and turning CX into a revenue driver.

The hardest part of deploying AI agents in a contact center isn’t the technology. It’s the moment when a successful pilot fails to translate into production.

I’ve seen that moment from both sides. At Autodesk, I inherited a large customer service organization with CSAT in the low 80s and led it to an industry-high 95%.

That result didn’t come from technology alone. It came from prioritizing customer experience (CX) first and letting efficiency follow.

The leaders who succeed will be those who move deliberately from pilot to production, treating AI agents as strategic assets, not cost levers.

This article is that playbook with these five steps to follow.

Step 1: Reject deflection; adopt enhancement strategies.

Most AI contact center deployments begin with a deflection goal: to keep customers away from human agents as long as possible. Containment rate becomes the North Star. Every interaction resolved without human involvement is counted as a win.

This framing is flawed. Deflection optimizes for the cost of the interaction. But enhancement optimizes for the outcome the customer needs.

Towards reaching a 95% CSAT score at Autodesk, we mapped every category of customer contact and asked these questions: “How can we reduce customer friction?” and “Where can automation deliver value while maximizing CX?”

For many high-volume, low-complexity contacts, such as order status, account updates, and basic troubleshooting, AI can deliver faster, better experiences than routing customers to queues and waiting for humans.

But for anything involving escalating frustration, account risk, or complex multi-system issues, the human touch is critical.

The gap between a successful pilot and a failed production rollout is usually a measurement problem.

The deployment question should be: “Where does AI create a genuinely better experience and where does it create friction that we are willing to accept because it reduces cost?” That second category should be small and it should be called out in your planning process.

Klarna’s own CEO eventually put it plainly: the key distinction in customer satisfaction lies in the type of task. Basic tasks are often handled more efficiently by AI. Complex problems still require human interaction (CX Today).

That insight should have been the starting point, not the lesson learned 18 months and a public reversal later.

Step 2: Segment your interaction portfolio before writing a single use case.

Before deploying any AI agent, you need a clear map (see FIGURE) of customer interactions. Segment them across two dimensions: interaction complexity and emotional intensity.

• Low complexity, low emotion. These are your best AI candidates: password resets, order status, balance inquiries, appointment scheduling, basic policy lookups. The customer wants speed and accuracy, not empathy. AI can outperform humans here when implemented well.
• Low complexity, high emotion. Proceed with care. A billing dispute is technically simple, but the customer calling about it may be stressed or at risk of churn. AI can start the interaction, but the escalation path to a human should be frictionless and swift. Klarna’s AI chatbot took up to 20 seconds to answer simple FAQs, not to mention the runarounds afterwards. That latency alone destroyed the experience.
• High complexity, low emotion. AI can assist but should not lead. Agents with AI co-pilot tools, such as real-time knowledge surfacing, case summarization, or next-best-action prompts, perform measurably better here than either AI alone or unaided humans.
• High complexity, high emotion. This is a human-first zone. Customers in financial distress, experiencing product failures with downstream consequences, or navigating multi-channel escalations need a skilled, empathetic agent who can reason through a tricky situation. No current AI agent does this well, and attempting to automate these interactions is where CX is severely compromised, leading to customer churn. This segmentation is not a one-time exercise. As AI capabilities evolve and your interaction mix shifts, revisit and revise it.

Step 3: Define what “production-ready” means before piloting.

The gap between a successful pilot and a failed production rollout is usually a measurement problem. Pilots optimize for what’s easy to track: containment, handle time, deflection. Production is judged on what actually matters: customer satisfaction, repeat contacts, retention, lifetime value.

That misalignment is fixable, but only if you define production success before you begin the pilot. This requires early alignment across Finance, CX, and Operations on the metrics that will govern scale decisions.

Four thresholds should determine whether an AI agent is ready for production.

• CSAT parity. AI interactions must meet or exceed human baseline CSAT for the same use case. This is non-negotiable.
• Repeat contact below baseline. If customers come back more often after interacting with AI, the system is deferring problems, not solving them.
• Controlled escalation rates. Every AI-to-human handoff carries cost and friction. Track escalation by interaction type. Rising rates signal poor scoping or routing.
• Seamless human fallback. Customers must be able to reach a human quickly and without losing context. In practice, many interactions will still require this handoff, often at critical moments.

Gartner estimates that by 2029, agentic AI will resolve 80% of common service issues autonomously, reducing costs by 30% (Gartner, March 2025). Getting there isn’t a single leap. It requires disciplined, iterative expansion where each deployment earns the right to scale.

Step 4: Govern your AI agents like managing your best employees.

When I managed large customer service orgs at ADP and Autodesk, we did not deploy a new rep into a live customer interaction without training, quality monitoring, escalation protocols, and coaching feedback loops.

AI agents deserve the same governance structure, or an even more rigorous one, because they operate at a scale and speed no individual human agent can match.

The agents who will thrive in an AI-augmented contact center are the ones who can handle the interactions AI cannot: the complex, the emotional, the novel, the high-stakes.

Governance in practice means four things:

• Behavioral guardrails. Define explicitly what your AI agents are authorized to do, say, and offer. Define what they are not. AI agents that stray outside their defined scope, providing inaccurate information or handling interaction types they were not trained for, will create liability and erode trust at scale.
• Quality reviews. Sample AI-handled interactions the same way you sample human agent interactions. Score them on the same rubric.
• Feedback loops into retraining. Unlike traditional software, AI agents learn and improve when their failures are fed back into the model. This requires a process: someone reviews queues, analyzes escalation patterns, and model updates are tested before redeployment.
• Human supervisor visibility. Supervisors need real-time dashboards showing AI agent performance alongside human agent performance. Both human and AI should be managed with the same operational rigor, not as separate domains.

The organizations that are getting AI right are building an AI workforce management (WFM) discipline inside their contact center operations. It is not glamorous. But it is what separates a sustainable deployment from a high-profile reversal.

Step 5: Bring your human agents into the deployment, not the aftermath.

One of the most consistent findings in recent research is that contact center agents are more open to AI than leadership assumes.

Research from Cresta found that 65% of agents want real-time AI suggestions during customer interactions. Organizations that reduced new agent onboarding time by 50%-plus did so by embedding AI assistance into the training process.

The agents who will thrive in an AI-augmented contact center are the ones who can handle the interactions AI cannot: the complex, the emotional, the novel, the high-stakes. Your AI deployment is a talent strategy and should be managed as such, like the following:

• Bring agents into the pilot. Ask them where AI is helping and where it is creating friction. Their feedback is often the earliest signal that something is wrong with routing logic or scope definition: signals that would normally take weeks to show up in your CSAT scores.
• Be honest with your team about what AI deployment means for their roles. Define what human agents will be responsible for as AI takes on more volume. Invest in building those capabilities and not on attrition to resolve the equation.

Putting the Playbook in Practice

The pressure to deploy AI agents at speed is real. A Gartner survey found that 77% of customer service and support leaders are feeling it from their own senior executives (Gartner, October 2025). The pressure will only intensify as agentic AI capabilities expand.

But the leaders who will succeed are the ones who defined their customer interaction portfolio before they defined their use cases:

• Who set production-readiness thresholds before they ran their pilots?
• Who governed their AI agents with the same rigor they applied to their human workforce?
• Who brought their teams along rather than presenting them with a fait accompli?

The Klarna story (see BOX) is instructive because the reversal was avoidable. The cost savings were real. The quality deterioration was also real. A more deliberate deployment strategy would have captured most of the former while preventing most of the latter.

The goal is not to choose between cost efficiency and customer experience. The goal is to deploy AI agents in a way that delivers both.

The Gartner survey cited above confirms where AI deployments should be focused. The highest-value use cases are agent assist, self-service, operations automation, and agentic AI, not, and in my view, mass agent elimination.

I spent years driving toward that 95% CSAT score at Autodesk by treating every customer interaction as a signal worth listening to. AI agents do not change that discipline. They scale it, if you govern them well enough to let them.

The technology is ready. The question is whether your deployment strategy is.