Yet, we do expect perfection from AI. When a Gen AI chat interface delivers erroneous information or a generated image includes an extra finger or two, many see this as a fundamental flaw of AI and are ready to dispel the entire technology. One bricked phone doesn’t cause distrust of all smart phones, but one AI “hallucination” can spark headlines about the end of trust in technology. 

Why? Because AI feels different. Its intelligence, pervasiveness, uncanny ability to feel like a peer, and complexity shapes our perfectionism bias towards AI.

The Perfection Expectation

For years, businesses have been using AI and machine learning in innovative and efficient ways, such as product recommendations, credit scoring, fraud detection, and customer care. Those systems quietly shaped modern life, yet most people never saw them in action so they rarely judged their mistakes.

Gen AI changed the game with its ubiquitous availability to anyone with a smartphone or a computer. ChatGPT alone reports that it has 800 million weekly users. Browsers feature Gen AI search summaries. Business productivity tools offer Gen AI capabilities for copy and image creation. Whether or not you wish to engage with AI, it’s here. 

Generating content at such a massive scale is bound to have some hallucinations and errors. Fundamentally, LLMs predict what’s probable, rather than actually produce factually accurate information. They match patterns and predict the next statistically plausible word or pixel. Due to human-based reinforcement learning, AI capabilities continue to advance as engagement skyrockets, but errors still occur. 

Mistakes can feel personal, too. Gen AI is made to converse like a human; it’s an always-on work assistant and confidante. Subconsciously, we hold it to human-level accountability. When Gen AI hallucinates a citation or generates an image with too many fingers, it can feel like a betrayal of the overall promise of AI.

But for AI engineers and scientists like myself, errors are an expected step in the development of newer technologies like Gen AI and agentic AI. What’s critical is how AI is ethically built and used and how errors are mitigated.

Why Black-Box Technology Makes Businesses Nervous

Despite these concerns, businesses are pursuing the promise of AI-driven applications: greater personalization, efficiency, easeful service, and innovation. Any worries about AI imperfection center around risk, ROI, and expensive real-world consequences: initiatives that fail due to lack of reliability and therefore adoption, lawsuits caused by hallucinations or bias, and reputational damage.

This concern is fueled by the black-box nature of generative and agentic AI. The outcomes of rule-based AI, such as when matching an intent to a response in an AI-based customer care platform, are pre-programmed, transparent, and auditable. Traditional predictive AI, such as for sales forecasting, relies on structured data and statistical models. AI decisions can be traced with relative confidence.

With trillions of data points as inputs, deeply layered non-linear classifiers, and a large number of tunable parameters, Gen AI functions as a black box lacking complete explainability. It’s virtually impossible to trace an output to an exact training source, especially as many platforms build upon proprietary and open-weight models like Gemini, GPT, Claude, Mistral, and Llama, which may disclose benchmarking and testing procedures, but not data sources.

Agentic AI will further compound this complexity and frustrate the desire for glass-box explainability. Agentic architecture essentially operates like a distributed computer network, such as with microservices. Multiple AI agents autonomously coordinate and communicate to perform multi-step tasks. Gen AI is used to interpret the main goal, reason through trade-offs, and dictate precisely how actions will be executed. 

Due to this black-box nature of Gen AI and agentic AI, many organizations are increasingly scrupulous of their AI vendors, demanding a higher level of infallibility and accountability than they might from other technologies. 

Bringing Transparency to the Black Box of AI

As AI matures, so will our expectations. Experience will teach both businesses and consumers that imperfection doesn’t signal failure — it means opportunity for refinement. What matters most is knowing that people and processes are in place to identify, measure, and mitigate those imperfections responsibly. 

For now, businesses building and using AI must contend with these high expectations and even welcome the scrutiny. In this wild west moment, industry-wide guardrails are still forming. Regulatory bodies are scrambling to keep up with advances and organizations must do their best to form their own ethical frameworks, internal governance, and oversight processes. 

At Interactions, we believe that transparency is key to building more trust in AI. Our openness regarding our ethical AI practices — such as documenting data provenance, testing and benchmarks, human-directed training, accountability and integrity regarding how issues are identified and mitigated, and standards for vetting third-party models and solutions — creates visibility into what feels unknowable. 

To learn more about our work in shaping the future of responsible AI, visit our AI Trust Council site.

The post Why Transparency, Not Perfection, Defines the Future of AI appeared first on Interactions.

Consumers have embraced AI for research, productivity, and conversation; ChatGPT alone has more than 700 million weekly users. Yet a majority of consumers polled by Pew say they’re extremely or very concerned about misuse of personal data (71%), inaccurate results (66%), and bias (55%). 

On the business side, 78% of organizations used AI in 2024, a sharp increase from 55% in 2023. At the same time, Gartner recently declared Gen AI to be in the “Trough of Disillusionment” in its Hype Cycle framework, as business leaders express disappointment with ROI. On the trust side, 55% of AI experts lack confidence that U.S. companies will develop and use AI responsibly and 55% doubt the U.S. government will effectively regulate AI.

There are many reasons for this AI trust crisis. AI is changing how we work, how we learn, and how we access information. Bad actors, poor training, and poor implementation can lead to the issues listed above. And, yet, to the average user, it can seem like no one is slowing down to address the current issues before steaming ahead with new advancements, like agentic AI.

In such an environment, every AI constituent — companies that produce AI platforms, organizations that utilize them, and end consumers — are at the mercy of the worst actors unless ethical and experienced AI companies take a stand. 

Now Is the Time to Commit to Responsible AI

Interactions has led AI innovation in customer care since 2004, a lengthy tenure in a space that’s filled with startups leaping into the opportunities afforded by AI advancements. We’ve been honing our platform — a sophisticated blend of predictive, generative, and agentic AI with human intelligence — for two decades to reach 97% accuracy rates and a product that delivers effortless customer experiences from day one. Our team boasts an impressive roster of AI experts, including many PhDs, whose work has earned 130+ patents.

The reason we’re tooting our own horn is that we believe it’s up to companies like ours to take a leadership position on AI ethics and trust. Every day, we face a gauntlet of tough questions from prospects, customers, auditors, and partners on how we handle AI training and customer data. This pressure (which we welcome!), along with our company ethos of always using AI with the utmost integrity, means that we’ve already thought through the many ethical questions that AI raises, developed governance and accountability measures, engineered critical security and data safety processes, and documented the procedures. 

This work is ongoing. New advancements, use cases, and regulatory frameworks will crop up, and we’ll continue to center trust and ethics in developing responsible AI solutions. 

7 Tenets for Ethical AI Development and Use

A one-size-fits-all approach to AI ethics isn’t feasible, as every organization will differ in how (and if) it develops and uses AI. Regardless, there are core responsible AI principles that we believe every organization should consider as they create their own AI rulebook.

• Put people first. For each new use case, examine the potential positive and negative effects on your customers, employees, and humans in general. For example, will your use of AI increase inclusivity and standardize care for all customers? How are you testing for bias and errors? How are humans involved in AI oversight processes? What value do you provide in exchange for the use of customer data?
• Take an Ethical by Design approach. This stance, inspired by Secure by Design principles, accounts for AI risks from the moment each new AI app, use case, or platform purchase is conceived. All decisions should be weighed against your company’s AI guidelines before any risks become reality and you head down roads that are costly to reverse.
• Demonstrate integrity through transparency and accountability. Define and assign accountability at every stage, maintain accessible documentation on your responsible AI procedures, and always be ready to demonstrate that you are following your own rules. Regulatory requirements, third-party audits, and probing customer questions should be considered welcome opportunities for demonstrating your trustworthiness, rather than intrusions.
• Engineer for the strictest standards and protect by default. By building to the most stringent standards — whether they apply to your region or industry — enables you to proactively protect your customers and future-proof your business against scrambling to meet the many evolving regulations like the EU’s AI Act and various U.S. legislative actions. Similarly, offering high levels of data protection by default takes the onus off your customers and shows your commitment to their privacy.
• Select the right AI for the right task. When newer AI technologies like generative AI and agentic AI capture our imaginations, they can quickly become the go-to goal for company innovation. However, using the right AI for the right task is both an ethical and a risk consideration. For example, using Gen AI when a deterministic solution is more appropriate can introduce unnecessary bias and error risks, as well as require more computational power. This isn’t to say that Gen AI doesn’t have its place. Rather, its use should be carefully considered in relation to your needs, goals, and risk tolerance. (To read more about how to identify the right use cases for agentic AI, check out this complimentary Gartner® report.)
• Consider the company you keep. Every company has either become or will become an AI customer, as AI features become default across the technology landscape. This is true of AI producers, as well. Create procedures to properly vet your AI partners for trustworthiness and integrity. Not only do you want to be sure these companies have similar ethical stances, protective processes, and AI stack vetting procedures, you want to know that they’ll do what they say when it comes to working with your and your customers’ data.
• Prioritize learning, questioning, and taking a stand. AI, much like other big-hype technologies like cloud, is incredibly complex, making it difficult to keep pace with its acceleration. Constant learning, curiosity, discussion, and vigilance are key. Form an AI council to navigate thorny AI questions, balancing the risks and rewards of AI implementation with executive priorities. This council should include members with expertise in business strategy, data, risk, ethics, IT, and customer experience. Additionally,  there should be continuous education for all employees on both new AI initiatives to improve adoption, but also on AI threats from bad actors, good AI practices (like keeping proprietary data out of public models), and the benefits of powerful, reputable AI platforms.

The AI Trust Council

The seven tenets outlined above aren’t just theory — they’re the foundation of how we build and deploy AI at Interactions. We’ve codified this approach to trustworthy, ethical AI by forming the AI Trust Council. 

Established by Interactions experts who are deeply engaged with and knowledgeable about AI and its impacts. The Council’s mission is to help our company, customers, and the general business community stay educated, keep asking the tough questions, and to above all prioritize trust and humans as AI continues to evolve. From discussing the ethics of new use cases, to monitoring regulatory changes, to publishing practical guidance, the AI Trust Council shows how responsible AI principles can be embedded into day-to-day decision-making.

For Interactions, the Council is both an internal compass and an external voice — a way to ensure we stay true to our commitments while helping others navigate the complex and fast-moving AI landscape.

The post Reversing the AI Trust Crisis: 7 Tenets for Ethical AI Use appeared first on Interactions.

Regulatory compliance is often perceived as a headache — expensive, restrictive, innovation-killing red tape. But for fast-moving technologies like AI, the reality is the opposite. 

Not just new regulations such as the EU AI Act, but also existing (and perhaps seemingly unrelated) regulations such as GDPR and CCPA create frameworks for building trustworthy AI systems and protecting both the data they are built upon and the people who use them. Further, these laws don’t just protect consumers; they also unlock the confidence needed for widespread adoption. Responsible AI doesn’t happen by accident. The rise in AI-related incidents and the fall in AI trust makes that clear.

• AI incidents increased 66% in 2024 vs. 2023. If the current pace of incidents in Q1-Q2 2025 continues, they could increase again by 42% this year. 
• 69% of leaders cited concerns about AI data privacy in Q2 2025, leaping from 43% just two quarters before.
• Public trust in AI companies to protect personal data fell from 50% in 2023 to 47% in 2024.

Rather, responsible AI emerges when organizations embrace clear rules that protect both people and data, define accountability, and set shared expectations. 

Forward-thinking companies that embrace compliance and a proactive governance-by-design stance — and that demand this same commitment from their AI supply chain partners — gain strategic advantages. Compliance becomes strategic and foundational, rather than a checkbox to be marked off, making it easier to adapt to future regulations, minimize risk, and demonstrate transparency and ethical AI standards. 

In a world where trust is paramount, proactive compliance is a powerful green flag.

Regulations on the rise

At Interactions, we practice what we preach by welcoming regulatory guidance. Every new layer of clarity, accountability, and consumer protection makes it easier to not only innovate responsibly but also convey our strategy and corporate intent. We believe the future of AI won’t be built in spite of regulations, but because of them.

And more regulations are coming. Many are regional: In 2024, 131 state-level AI-related bills were passed in the U.S. One of the most impactful may be Colorado’s Artificial Intelligence Act (CAIA), set to become effective on June 30, 2026. The CAIA is the first US regulation to focus on high-risk AI systems, just as the EU AI Act does, which further codifies responsible AI use for any organization serving customers in EU member states. Like GDPR, it’s set to standardize best practices across the world.

While it’s still being phased in, the EU AI Act is the first multi-national, comprehensive AI framework. It requires transparency for limited-risk and general-purpose AI, such as intelligent virtual assistants, while imposing stricter obligations on high-risk systems (e.g., product safety, healthcare, and law enforcement) and outright prohibiting “unacceptable” systems, such as social scoring that reinforce bias and discrimination.

Looking back to look ahead

To consider how the EU AI Act can revolutionize AI use, let’s look back at the impact of GDPR. 

The General Data Protection Regulation (GDPR), along with the similar California Consumer Privacy Act (CCPA), is one of the best things that ever happened to Interactions. These specific, stringent frameworks empower us to protect our customers’ data with the highest standards and transparency. Let me explain: 

GDPR established the idea of Controller and Processor. CCPA defines the analogous roles of Business and Service Provider. The Controller (or Business) is the organization that makes decisions about how to process the data. The Processor (or Service Provider) follows the Controller’s written instructions on what to do with their data and is not allowed to deviate from those instructions. In our case, our customers are the Controllers, and Interactions is the Processor. 

Since Interactions abides GDPR, CCPA, and a myriad of supporting legal agreements, we are contractually bound to do what we’re told with your data. For example, if a customer prefers for Interactions to not use their PII-scrubbed call data to train our general models that may benefit our other customers, we won’t do that. When a customer requires their non-PII data, such as call recordings, to be immediately deleted after processing or 30 days later, we comply. We have no choice — and we like it that way. 

Instead of attempting to navigate a fragmented web of regional and customer-specific security requirements, we adopt the highest global standards as our baseline, which simplifies operations and future-proofs our technology. This baseline includes:

• Handling customer data according to our legal agreements, applicable regulations, and ethical commitments
• Isolating model training environments per customer requirements
• Deleting training data and making it impossible to reverse-engineer, or invert, the data from the model
• Automatically redacting PII from all interaction records

This stance also protects our customers, simplifies compliance and audits, and ensures that we can meet your needs wherever you operate — and wherever they may grow next. For example, a recognizable global retailer piloted Interactions within a single country. Success led them to expand their use of Interactions across 28 countries, providing a unified global customer experience. Regulatory compliance with standards like GDPR and CCPA eased this expansion.

The future of AI trust and compliance

Ultimately, no single company will shape the future of AI trust and accountability alone. As businesses’ technology supply chains increase in complexity, it’s imperative that every link — every vendor — follows secure and principled AI and data practices. 

Regulations like GDPR and the AI Act codify these principles as guardrails, enabling companies to adopt AI with less risk and more confidence. Rather than limiting innovation, these clear, safe, and principled frameworks allow companies to innovate with greater speed and focus. This foundational confidence then empowers businesses to pass on trust to their own consumers.

Join Interactions in our quest to shape the future of responsible AI by visiting our AI Trust Council site. 

The post More Regulation, More Trust: The Case for Stronger AI Guardrails appeared first on Interactions.

That’s where the concept of agentic AI comes in. Agentic AI refers to the broader discipline of building intelligent systems that can operate with varying levels of autonomy, perceiving their environment, making informed decisions, and taking meaningful actions. AI agents are the software applications that bring these capabilities to life in customer service.

At Interactions, we believe the smartest way to bring agentic AI into your organization is with purpose. Not by deploying it everywhere, but by starting where it makes the most sense. In this blog, we’ll walk through what makes AI agents different, where they create the most value, and how to approach deployment in a way that works for your business and your customers.

Not All AI Agents Are Created Equal

There’s a lot of noise in the market right now. Nearly every tech vendor is claiming they offer AI agents, but many of these “agents” are really just rebranded virtual assistants. They might have a conversational interface, maybe even some generative AI behind the scenes, but they don’t meet the bar for true agentic AI.

A real AI agent does more than respond. It reasons. It perceives what’s happening, understands the context, and makes informed decisions about what to do next. It’s proactive, not reactive. It’s adaptable, learning from what works and adjusting based on new inputs. These are the capabilities that separate true agentic AI from the rest.

The Right Use Case Matters More Than the Technology

One of the biggest mistakes companies can make when introducing AI agents is trying to apply them everywhere. Just because the technology is exciting doesn’t mean it’s right for every situation. In fact, overusing AI agents can quickly drive up costs and create inconsistent results.

That’s why Gartner® describes what they call the “AI Agent Zone.” This is the sweet spot where AI agents shine – with interactions that are too complex. Think about scenarios like troubleshooting a device, upgrading a service, or making a return. These are often the moments where customers feel stuck or frustrated, and they’re exactly where an AI agent can step in to help.

When deployed in the right zone, AI agents can deliver fast resolutions, reduce transfers, and ease the burden on your contact center. More importantly, they can improve the customer experience without adding unnecessary cost or complexity to your operations.

It’s All About the Tiered Approach

To get the most out of AI agents, it helps to think in terms of a tiered service model. At the bottom tier, you have no-touch self-service. This includes things like help articles or basic chatbots that handle simple tasks. It’s fast and inexpensive, but limited in what it can do.

The middle tier is low-touch service. This is where conversational AI comes in, helping customers navigate more involved issues or route them to the right place. It’s a good balance between automation and personalization.

The top tier is high-touch support, where human agents or subject matter experts handle complex or sensitive issues. This is where empathy and expertise really matter.

AI agents belong between the middle and top tiers. They’re ideal for interactions that go beyond what a basic bot can handle but don’t necessarily need a human. By placing AI agents at this mid-to-high point in your service model, you create a smoother path for customers and reserve your human team’s time for the moments that truly need them.

Start Small, Learn Fast, and Scale Intelligently

Another one of our key takeaways from the Gartner research is to avoid diving in too deep, too fast. Instead of launching AI agents across every touchpoint, start with a few high-impact use cases. Choose scenarios where you’re confident the technology will deliver value and where customer expectations are already high.

Many vendors offer AI agents at a “basic” level of capability to help organizations get started. These agents still bring powerful features like perception, decisioning, and actioning, but they’re easier to deploy and measure. Starting small gives your team the chance to test performance, fine-tune workflows, and build internal confidence before rolling out at scale.

This step-by-step approach doesn’t just reduce risk, it sets you up for long-term success. By proving value early, you can secure buy-in, make data-informed decisions, and gradually move toward more advanced use cases as the technology and your organization mature.

Agentic AI Is Already Reshaping Customer Experience

According to Gartner, “By 2028, 60% of organizations will have deployed at lease one CSS agentic AI use case in production, up from less than 1% in 2024.” The shift is already underway, and companies that wait too long may find themselves playing catch-up. But the organizations that succeed won’t be the ones that moved first. They’ll be the ones that moved smartest.

With agentic AI, it’s not about checking a box or chasing a trend. It’s about using automation where it actually improves outcomes and letting humans focus on what they do best. That’s the kind of CX transformation that lasts.

At Interactions, we’re helping businesses lead this shift with Conversational AI that doesn’t just talk. It listens, understands, and acts. And it does so in a way that puts customers first, reduces effort, and builds trust across every interaction.

Gartner, Ensure Benefits Outweigh Costs of AI Agents in Customer Service and Support, By Pri Rathnayake, 20 April 2025. Gartner is a registered trademakr and service mark of Gartner, Inc. and/or its affiliates in the U.S. and internationally and is used herein with permission. All rights reserved. This graphic was published by Gartner, Inc. as part of a larger research document and should be evaluated in the context of the entire document. The Gartner document is available upon request from Interactions LLC.

The post The Smart Path to AI Agents: Why Strategic Use Beats Splashy Hype appeared first on Interactions.

But here’s the catch. AI doesn’t just work on its own. Not if you want it to work well, at least.

At Interactions, we believe the most effective AI isn’t just powered by data and models. It’s shaped by people. That belief is at the core of our newest ebook, The Human Side of AI: Behind Every Conversation is a Team Focused on Your Success, which explores what it really takes to design, build, and continuously improve AI that actually helps your customers AND your business.

The Minds Behind the Magic

While many companies only talk about AI in terms of capabilities and features, we focus on even more: the people who bring those capabilities to life. That includes:

• Conversation designers who craft dialog that sounds natural, uses your brand voice, and guides customers to fast, accurate outcomes.

• Developers and system integrators who make sure your IVA plays nicely with your back-end systems, CRM, APIs, and more.

• Client success teams who aren’t just monitoring performance – they’re proactively looking for ways to improve your KPIs and support your roadmap.

• Support teams who ensure the technology performs reliably and troubleshoot before you even know there’s an issue.

These people bring creativity, experience, and strategic thinking to every part of the AI lifecycle. They’re the reason your IVA can navigate complex business rules, respond with empathy, and evolve over time.

Real Conversations Require Real Human Insight

One of the biggest takeaways from the ebook is that AI doesn’t replace humans, it works best when it’s guided by them.

Take conversation design. While AI can process and understand inputs, it takes human expertise to decide how to respond, what tone to use, and how to handle tricky or emotional scenarios. A return question in retail is different from an appointment issue in healthcare. A payment arrangement in energy might require nuanced handling depending on regulations and internal policies. Our teams don’t guess at these things. They study the data, listen to real calls, collaborate with your business stakeholders, and build experiences that reflect the real world, not just the ideal one.

For example, one energy client had highly complex internal rules around missed payment arrangements. Our team worked closely with them to essentially turn their APIs into a knowledge tree, allowing the IVA to handle countless scenarios with precision. The result? A smoother experience for customers, and a huge drop in calls to the contact center.

Optimization Doesn’t Stop at Go-Live

Another misconception we tackle in the ebook is that AI systems are “set it and forget it.” In reality, the best AI solutions are constantly evolving. Our teams regularly monitor customer interactions, analyze performance, and surface new opportunities to increase self-service, reduce friction, and improve outcomes.

The ebook shares several stories where small shifts—like changing phrasing or modifying a call flow—led to major improvements in metrics like containment, agent transfers, and customer satisfaction.

Another example? A retail client was using internal jargon in their IVA. Customers calling about returns were asked if their issue involved a “claim” from a “non-brand store”, phrasing that caused confusion and led to more live agent transfers. Our team listened to the calls, made the language more natural (“return” and “third-party store”), and saw both confusion and transfers drop significantly.

(We don’t give away all of the examples here, but they’re worth the read in the ebook…. hint hint.)

Finding a True Partner, Not Just a Vendor

Ultimately, what this ebook highlights is the value of having a true partner when you implement conversational AI. Someone who knows your industry, understands your goals, and brings the right mix of technical skill and human judgment. Someone who doesn’t just deliver a product but stays involved, offering ideas, uncovering patterns, and helping you adapt as your business evolves.

At Interactions, our customers don’t just get a platform. They get a team. One that’s as invested in their success as they are.

Want to see how it all comes together? Download The Human Side of AI to meet the people behind the experience and see how human insight, creativity, and care are what make great AI possible.

The post The Human Side of AI Still Matters (More Than Ever) appeared first on Interactions.

But not all customers want the same thing. In fact, generational differences in support preferences are more pronounced than ever. According to the 2024 Gartner State of the Customer survey, 52% of Baby Boomers still prefer calling customer service, while just 28% of Gen Z feel the same. Meanwhile 18% of Gen Z favor web chat, compared to just 7% of baby boomers. To meet differing customer expectations, customer service and support technology leaders increasingly recognize the importance of enabling connections across multiple channels.

That divide presents a real challenge: how do you design a support strategy that works for everyone?

That’s where the new Gartner® report, Unleash Omnichannel Customer Service to Improve CX, comes in. It explores the critical need for organizations to build omnichannel service models that can flex to meet a wide range of customer needs, without overwhelming internal teams or tech stacks. It’s not just about offering more channels; it’s about making those channels work together in a way that feels seamless to the customer.

Why Omnichannel Matters Now

Customer expectations aren’t just evolving, they’re accelerating. And with them, the pressure on service leaders to deliver efficient, personalized, and consistent experiences has never been greater.

Yet many companies still struggle with disconnected tools, outdated routing, and support models that haven’t kept pace with how people actually engage. The result? Missed opportunities, higher effort, and customer churn.

The Gartner® report highlights why a successful omnichannel strategy must be:

• Customer-led – designed around how people prefer to engage, not around internal constraints
• Integrated – so customers don’t get stuck repeating themselves or lost in handoffs
• Flexible – to support new channels, technologies, and service models as they emerge
• Measurable – with clear ways to track impact and improve over time
  

It also outlines key steps leaders can take to move beyond fragmented experiences and build a service strategy that is both efficient and truly responsive to customers.

What This Means for You

At Interactions, we’ve seen firsthand how powerful a well-executed omnichannel strategy can be. We help brands blend AI and human support across every touchpoint, chat, text, voice, and more, so customers get what they need without friction. We don’t believe in patchwork solutions. We believe in experiences that feel natural, personalized, and aligned with what today’s customers expect.

If your goal is to improve satisfaction, reduce operational drag, and create experiences that actually work for every generation, this report is a must-read.

Our takeaway: Omnichannel customer service isn’t optional, it’s essential. This Gartner® report lays out how to do it right.

Gartner, Unleash Omnichannel Customer Service to Improve CX, 28 February 2025, Francesco Vicchi GARTNER is a registered trademark and service mark of Gartner, Inc. and/or its affiliates in the U.S. and internationally and is used herein with permission. All rights reserved.

The post Customer Expectations Are Changing. Is Your Service Strategy Keeping Up? appeared first on Interactions.

This challenge is the culmination of several large issues facing financial services companies today:

• The competitive gates have been thrown wide open with neobanks, direct banks, digital wallets, P2P payments, and digital banking in general changing how people manage, spend, and save their money.
  
• Economic pressures are sending banks on cost-optimization missions. The majority of banks (85%) surveyed by KPGM in Q4 2023 were planning to cut at least 5 to 10% of costs over the upcoming 36 months. About a third (36%) were aiming to cut 10 to 20%, while 15% wanted to cut more than 20 percent.
  
• Customer satisfaction isn’t optimal. Recently, the J.D. Power 2024 U.S. Direct Banking Satisfaction Study found customer satisfaction declined by 27 points year-over-year for checking customers. According to an interview with The Financial Brand, the key culprits for this decline were weaker customer service and slower problem resolution. Yet with just an average score of 688 out of 1,000, direct banks (those that do business digitally and have a banking charter) still score higher in customer satisfaction than traditional banks.

Cutting contact center costs: Why containment isn’t the answer

Optimizing costs in the contact center is often seen as a key solution to these challenges. More automated interactions will naturally lead to higher containment. Labor costs can decrease, while customers are still being helped. Win-win solution, right? 

While lower containment rates look sexy on a quarterly report, unfortunately they don’t equate to the ease-filled, engaging experiences that financial consumers seek. A customer who ends a call or chat, frustrated and angry that their issue hasn’t been resolved, is contained. If they are sent to self-service articles that don’t really help, are told to visit the website to conduct a simple transaction, or give up on a long wait queue? Also contained. 

In an increasingly global and digital world, competitors are just waiting to scoop up your unhappy customers, and those customers are just as ready to spread negative reviews and poor experiences online. 

Containment simply isn’t enough to base contact center success upon. Increasingly, financial services and other companies are tracking Customer Effort Score, a quick survey question that evaluates how much effort someone had to exert to solve their issue. This score can be augmented by other KPIs, like first call resolution, abandonment rate, CSAT and NPS — and, yes, containment.

Personalization and ease without skyrocketing labor costs

So how do you boost your Customer Effort Scores and associated metrics while decreasing costs, all without relying too much on containment? The short answer is better, more human-like customer service automation. 

The long answer is leveling up your conversational AI tech so that it can understand more and do more. A best-in-class solution empowers customers to self-service more intents, therefore increasing ease, satisfaction, and containment. It can even help live agents when chats and calls must be escalated and provide data that helps your company unearth hidden issues and serve customers better.

What are the key capabilities you should look for when upgrading your conversational AI to contain costs and boost CX?

• Superior understanding via state-of-the-art Conversational AI that is being continuously evolved with the latest AI advancements.
• The ability to have humans augment AI by interpreting challenging utterances in the background, thereby containing more calls and rendering AI “mistakes” invisible to customers. 
• Machine learning that instantly ingests these human-tagged utterances to improve the system over time.
• Integrations with key banking, CRM, payment, appointment, and other systems to fuel personalization and self-service task completion. 
• An omnichannel experience that remembers interactions, so that customers can move fluidly between chat and voice for assistance. 
• Agent assists via information gathering that’s shared by the tech and Gen AI-driven contact summaries, suggested responses, and knowledge extensions.
• Advanced analytics that allow you to dig deep into your customer interactions to add intents and discover trending customer service topics.
• The regulatory, privacy, and security features that financial services companies require.
• A vendor that acts as a partner, regularly meeting with your company to dive into analytics together, plan for business and regulatory changes, and keep moving your tech forward in its abilities. 

When more interactions are not just contained, but satisfactorily concluded via superior automation, your financial services company will enjoy multiple wins: seamless scalability without increased labor costs, better insight into customer needs, and more engaged customers.

To learn more about how a best-in-class conversational AI is built, check out our ebook Under the Hood. Or, to learn more about Interactions helps companies like yours balance cost optimization with customer ease, visit our Financial Services web page.

The post Cost Cutting vs. CX: How FinServs Can Find the Right Balance appeared first on Interactions.

Dissatisfied customers can cost your business more than just goodwill. A study from the Kelley School of Business found that for every one-point increase in customer satisfaction on the American Customer Satisfaction Index (ACSI), utilities reduced their operating costs by $29 million. On top of the financial gains, improving the customer experience can enhance your brand image and foster loyalty.

One of the most effective ways to reduce customer complaints and improve satisfaction is to implement conversational AI—a modern solution that goes far beyond the capabilities of legacy IVRs or basic chatbots. The most advanced platforms combine human-like understanding with automation to deliver fast, personalized, and intuitive interactions. Here’s how conversational AI helps resolve five of the most common pain points in utility customer service.

Pain point #1: Long wait times

Utility companies frequently experience surges in customer contacts—for example, when hot weather overloads electrical systems or high winds dictate pre-emptive power outages. Even when you enlist extra seasonal agents to temporarily ramp up service, high call volume can mean long wait times for customers. Interactive voice response (IVR) solutions are often limited in their ability to understand customers’ accents, phrasing or intent, and frequently route callers to live agents for even simple tasks.

Conversational AI helps alleviate this by automating routine inquiries through voice or digital channels. Customers can check balances, get outage updates, or explore payment options instantly—without waiting for a live agent.

A truly modern solution offers omnichannel support, so customers can start a conversation via voice and continue it through SMS or web chat without repeating themselves. Seniors may prefer a voice call, while Gen Z may want to text. Look for a solution that provides persistent context so customers can start a conversation in one channel, then switch to another and pick up where they left off without re-inputting or repeating information. 

Pain point #2: Difficulty accessing account information

Few things are more frustrating for your utility customers than struggling to access their account information. Since most customers don’t call utility companies often, they can easily forget passwords, usernames or “secret questions” used to verify identity. Being asked to reset passwords and input multiple codes for two-factor authentication can be exasperating. Fail too many times and customers get locked out of their accounts, increasing their annoyance. 

Conversational AI can help simplify identity verification with natural language prompts, contextual understanding, and step-by-step guidance. Whether it’s resetting a password, updating contact information, or verifying account details, AI-powered conversations make the process faster and less stressful—without needing human intervention for most cases.

Pain point #3: Limited service outside business hours

Customers don’t only need help from 9 to 5. While traditional chatbots or IVRs can handle limited after-hours requests, they often fall short when customers have more specific or urgent needs.

Conversational AI offers 24/7 support that feels intuitive and helpful. It can facilitate paying bills, scheduling, confirming or canceling appointments, reporting outages and more—all in a friendly voice using language tailored to support your brand image. Sophisticated IVAs incorporating automated speech recognition can interpret customer requests despite unusual phrasing, mumbling, accents or poor connections, saving the day (or night) for the customer.  

Pain point #4: Complex billing and payment processes

Paying a utility bill shouldn’t be complicated—but it often is. When billing systems and customer support tools aren’t integrated, customers face hurdles that can lead to missed payments or frustration.

An automated attendant that understands natural language and can respond with a high degree of accuracy regarding intent can guide customers through making payments, checking their billing history and more. You’ll also want to choose an IVA that offers multiple language options to fit your customers’ needs. 

Of course, there are times when human help is needed, especially when untangling complex billing concerns. An IVA designed with human-in-the-loop technology can bring in agents when it doesn’t understand an issue. The best solutions blend human and machine intelligence, instantly summarizing the call for the agent, asking the agent what to do next, and responding to the customer without missing a beat. 

Pain point #5: Poor communication during service outages

Outages are stressful, and customers need fast, accurate updates. Unfortunately, traditional support systems often fall short—forcing customers to navigate endless menus or receive irrelevant responses from basic bots.

Conversational AI enables proactive, intelligent communication. It can recognize a customer’s phone number, location, and history to deliver real-time updates about service status, restoration timelines, or safety alerts. Even during peak volume, conversational AI platforms can handle thousands of simultaneous interactions—many of which customers won’t even realize are AI-powered thanks to the natural flow and tone of the conversation.

The right partner makes all the difference

Not all conversational AI platforms are created equal. The wrong solution can introduce more problems than it solves. For best results, choose a partner with deep experience in your industry, proven implementation success, and a commitment to collaboration and innovation.

Your contact center is critical to maintaining customer trust and loyalty. With the right conversational AI strategy, you can transform your service experience—making it faster, more human, and more effective at scale. Discover how Interactions’ Conversational AI for Utilities can resolve customer complaints with empathy, efficiency, and technology designed for real-world complexity.

Dive deeper into transforming utility customer service–download our free ebook, Modernizing the Contact Center: A Guide for the Utilities Industry, to explore how conversational AI can improve efficiency, reduce costs, and deliver better experiences, no matter the channel.

The post 5 Common Customer Complaints Utility Companies Face (and How Conversational AI Can Solve Them) appeared first on Interactions.

This is where Customer Effort Score (CES) comes in. CES is one of the most powerful indicators of customer loyalty, yet many companies still rely on outdated, sample-based survey methods that fail to capture the full picture. It’s time to rethink how we measure effort—because if you can’t measure it, you can’t fix it.

The Problem with Traditional CES Surveys

Historically, CES has been measured through a simple survey question:
“How easy was it to get your issue resolved?”

While this method provides valuable insights, it has serious limitations:

• Low Response Rates: Most customers don’t take the time to fill out surveys, meaning businesses only capture a small fraction of the customer experience, and likely, only the ones that left customers frustrated and dissatisfied.
• Delayed Feedback: By the time survey data is analyzed, the customer has already formed a negative perception—or worse, churned.
• Limited Scalability: Contact centers struggle to implement CES surveys across all interactions and channels, leading to gaps in insight.

This outdated approach leaves businesses flying blind when it comes to understanding the true impact of customer effort.

Predicting Effort in Real Time with AI

To solve these challenges, we leverage AI-powered analytics to predict Customer Effort Scores in real time, without requiring surveys.

Here’s how it works:

• AI analyzes 100% of interactions (calls, chats, emails), identifying effort indicators from speech patterns, sentiment, and resolution success.
• Predictive models assign an effort score to each interaction, even when no survey is completed.
• Actionable insights help businesses proactively reduce effort, whether through better self-service, agent training, or process optimization.

This scalable, real-time approach allows companies to pinpoint and eliminate friction points before customers are driven away.

Reflecting True Customer Experience Amidst Tech Disruption

As businesses increasingly integrate Gen AI, LLMs, and AI Agents into their customer service operations, predictive CES becomes crucial for evaluating the true impact on CX. While these emerging AI technologies promise to offer higher automation and efficiency, predictive CES reveals whether they genuinely reduce customer effort or create new friction. For customer service leaders, CES provides the data-driven insights needed to ensure AI adoption enhances, rather than undermines, the customer journey and service staff’s daily productivity.

Why Low-Effort Experiences Are the Future of CX

Customers aren’t just looking for resolutions—they want effortless resolutions. The easier a company makes it to get help, the more likely customers are to stay loyal and recommend the brand.

A recent discussion between Adobe and Interactions reinforced this point, highlighting how AI-driven solutions can reduce effort and enhance customer engagement. By removing friction, companies can build stronger relationships and drive long-term loyalty.

Companies that optimize for low effort:

1. Reduce churn by removing frustrating experiences
2. Increase automation success by refining self-service options
3. Boost agent efficiency by identifying and fixing high-effort interactions
4. Improve overall CX strategy with data-driven insights

At Interactions, we believe that effort is the defining factor in customer experience, and that traditional CES measurement methods simply aren’t enough. By adopting predictive analytics and AI-driven insights, brands can finally understand, measure, and eliminate effort at scale.

Your customers shouldn’t have to work hard to do business with you. By leveraging Predictive Customer Effort Score, companies can ensure every interaction is effortless, intuitive, and frustration-free. Because when effort goes down, loyalty goes up.

For a deeper dive into how AI can revolutionize CES measurement, check out the Predictive Customer Effort Score eBook from Interactions.

The post Why Customer Effort Score is the CX Metric You Can’t Afford to Ignore appeared first on Interactions.