Contact Center Agents' Perspective on AI

Transcript

[00:00:01] Steve MacDonald: Welcome to the Contact Center Perspectives podcast. I’m Steve MacDonald, your host. Today, we’ve got a really interesting conversation with Vladyslava Vozniuk. Vladyslava, you’re the Customer Service Manager for Usersnap, and we will be talking about AI. It’s a big conversation—a big topic—written about everywhere.

[00:00:21] Steve MacDonald: It’s almost like we’ve heard too much about AI. But what we rarely hear about is the agent’s perspective on AI—what they think, how they’re looking to adapt, and how they plan to work with it within our organizations. So that’s what we’re going to talk about today.

[00:00:40] Steve MacDonald: First, I’d love to talk about you. You started your career in customer success and customer support, then moved into digital marketing, and later returned to customer support. You wanted to be closer to the customer, and that’s great. In customer support, with all those interactions, we gain tremendous insight and value that can be shared across different departments—marketing, sales, and beyond.

[00:01:08] Steve MacDonald: Having worked outside of customer support and in other departments gives you a really interesting background for our conversation today. But before we dive in, Vladyslava, could you share a bit more about your background?

[00:01:25] Vladyslava Vozniuk: Sure! I have a really diverse background. I studied international relations and diplomacy, but at one point, I needed a summer job, so I decided to try customer support. At first, I thought it was just for the money, but then I realized the amount of soft skills you develop as a customer support representative is incredible. It’s hard to compare with most other jobs.

[00:02:06] Vladyslava Vozniuk: From problem-solving to empathy, intuition, persistence, and confidence—all of that can be gained through a customer support role. At some point, I decided to work in digital marketing, which was also really exciting. I believe it’s a customer-facing role in some ways because you learn about brand positioning and how to communicate with customers.

[00:02:22] Vladyslava Vozniuk: But I missed human relationships and truly helping people rather than just selling something to them. That’s precisely what customer success is—a combination of customer support and sales. You build relationships with customers, and while you do upsell, it’s only when it makes sense for the customer, not just for the company. That’s what I love about it.

[00:02:48] Steve MacDonald: That sales role is often overlooked. Unfortunately, contact centers and customer support are frequently seen as cost centers rather than value centers. But we’re constantly gathering valuable information, cross-selling, upselling, and answering questions from customers who are in pre-purchase mode. The impact on sales is direct and significant.

[00:03:11] Steve MacDonald: Today, we’re talking about AI and the agent’s perspective. It’s important to understand how critical the relationship between agents and customers is. To start, what are the challenges and opportunities of integrating AI?

[00:03:34] Vladyslava Vozniuk: Well, as you mentioned, one challenge is that human agents in customer support centers are often seen as costly, while AI is promoted as a cheaper and more efficient alternative. But that’s not entirely true because AI cannot match the value human agents provide. Honestly, I don’t think AI will be able to match it in the next hundred years.

[00:03:59] Vladyslava Vozniuk: That’s why I believe the only right approach to customer support is a hybrid model—combining AI and human agents. This way, you get efficiency and 24/7 availability, which many customer support centers already provide, along with human empathy, intuition, and the ability to handle complex tasks that AI simply can’t manage.

[00:04:25] Vladyslava Vozniuk: The rapid advancements in AI have made people overly excited, leading them to overlook its drawbacks. That’s one of the biggest challenges right now. But I feel like people are beginning to realize this, which is why many customer relationship management platforms are developing hybrid systems—where AI serves as the first level of support, and human agents handle the second level.

[00:04:53] Steve MacDonald: From the agent’s perspective, are they seeing AI as a helpful tool? Or are they worried it might take their jobs? What’s their overall sentiment—positive or negative?

[00:05:09] Vladyslava Vozniuk: There’s definitely a lot of nervousness about AI, especially in support centers. The first thing many agents think when they hear about AI is job insecurity. But I think a lot depends on how AI is communicated to them.

[00:05:29] Vladyslava Vozniuk: If leadership is transparent about their AI implementation plans—explaining and demonstrating how responsibilities will be assigned—then that nervousness tends to fade. It’s all about perspective. AI isn’t here to replace agents but to enhance their performance. When agents understand this, they become more open to adopting new technology.

[00:06:00] Steve MacDonald: That hesitancy you mentioned at the start—how do agents react when leadership reassures them that AI is here to enhance, not replace, their roles? Are they convinced? Or do they remain skeptical?

[00:06:17] Steve MacDonald: Do they say, “I hope that’s true,” but remain hesitant? Or are they confidently moving forward, believing their roles and the industry are secure?

[00:06:38] Vladyslava Vozniuk: I can’t say with 100% certainty because I can’t read their minds. But I can share what I’ve observed.

[00:07:06] Vladyslava Vozniuk: At the beginning of 2024, we had a discussion about AI and asked agents if they planned to integrate it into their workflow. They were hesitant. However, in a recent meeting, it turned out they had already started working with AI.

[00:07:31] Vladyslava Vozniuk: I think their perspective has evolved. In our performance meetings, I always highlight the strengths that make human agents irreplaceable. I believe that over time, their perception of AI is shifting positively. And based on my experience, I’m optimistic that they see AI as a tool rather than a threat.

[00:07:47] Steve MacDonald: Yeah, as time goes on and agents still have their jobs while using AI, that initial concern should start to fade.

[00:07:47] Steve MacDonald: I’d love to dive deeper into the unique value of human agents. We’ve touched on this already, but there are so many perspectives. What do you believe is the true value of human agents—something AI simply can’t replace?

[00:07:58] Vladyslava Vozniuk: Yeah. So, I’ll share my perspective as an employee of a product company. As I already mentioned, I rely on intuition for fundamental things that I think everyone would agree on because they’re completely true. When customers are frustrated, for example, they don’t want to hear just a dry response like, “Sorry for your inconvenience.”

[00:08:22] Vladyslava Vozniuk: They want to know that someone on the other end actually understands and cares about them. I even notice this in myself—I prefer human support when I have a real issue, not just a simple question, because I want to know there’s a person on the other side.

[00:08:39] Vladyslava Vozniuk: In my opinion, it’s reassuring to know that someone truly cares about your problem, at least to some extent. When it comes to product companies—like Usersnap—the voice of the customer is one of the most important aspects of our strategy and roadmap planning. Agents help us with this significantly because they collect customer feedback, prioritize what matters most, and forward it to our teams—specifically the product team. That’s how we build our roadmap. No one wants to spend a quarter or two developing features only to realize customers aren’t actually interested in them.

[00:09:21] Vladyslava Vozniuk: Also, with technical products, there are a lot of nuances. For example, when we have a bug, multiple customers might report the same issue but describe it in completely different ways. It helps our developers tremendously when agents identify these connections and escalate the issue with all relevant conversation links.

[00:09:45] Vladyslava Vozniuk: That way, developers can see how customers are affected and understand the process more clearly. This consolidated information allows us to solve issues more efficiently. That’s why I also question the efficiency of AI in some cases—it really depends on the industry and the specific situation.

[00:10:04] Vladyslava Vozniuk: Additionally, intuition plays a role. Depending on how customers behave and their tone of voice, agents can determine whether they are frustrated, seeking empathy, or simply looking for a quick resolution.

[00:10:21] Vladyslava Vozniuk: These small agent insights are crucial. Development resources are limited, and teams have to balance building features with resolving customer issues. Understanding which issues to prioritize wouldn’t be possible without agents.

[00:10:37] Steve MacDonald: In our previous conversation, you talked about the importance of human connection in building trust and loyalty. Some customers specifically want to talk to a live person, and there are clear benefits to that. Can you tell us more about this?

[00:10:56] Vladyslava Vozniuk: Yeah, that’s absolutely true. I see this reflected in our G2 reviews. Users frequently mention our agents by name and appreciate the human aspect of our support. Sometimes, I see customers joking with our agents in chats, and the agents joke back—it’s great! That’s something I personally wouldn’t be able to do as naturally.

[00:11:31] Vladyslava Vozniuk: During check-ins, customers often mention how much they appreciate our support. They even say hi to Simon and other agents, which really reinforces that human connection.

[00:11:44] Vladyslava Vozniuk: I was reflecting on this topic recently and realized that I always opt for human support myself. Curious if others felt the same way, I looked up some data and found a study by Carter from the summer of 2024. It showed that over 80% of users have major concerns about AI in product companies and actually prefer human support. So, it’s not just my personal opinion—there’s data to back it up.

[00:12:15] Steve MacDonald: That’s amazing because it validates what we all feel. AI can be frustrating because it often doesn’t truly understand what we want. As customers, we don’t want to be trained to communicate better with AI—we want a real conversation. The empathy and connection agents provide are invaluable. I love the idea of agents joking with customers and even being mentioned by name in G2 reviews. That shows the depth of connection a live agent can build.

[00:12:48] Steve MacDonald: That’s one of my biggest takeaways from what you just shared. Now, I’d love to talk about training and collaboration—how you work with agents, how AI fits into that, and how your company approaches it.

[00:13:04] Vladyslava Vozniuk: Yeah, just like chatbots can be trained, agents can be trained as well. But what AI lacks is two-way feedback. In our work, we use a Slack channel for most updates. We have an update channel and a document with guidelines on how to handle various situations and workflows.

[00:13:29] Vladyslava Vozniuk: What I find really valuable is that agents can always provide feedback. First, they can ask for clarifications. When you train a chatbot, you just feed it a source, and it generates responses—even if it doesn’t fully understand. AI might even hallucinate information. But agents will always clarify, which helps us identify issues in our workflows or even with our product features.

[00:14:07] Steve MacDonald: Yeah, it’s great when agents can contribute their perspective on training and materials. If something isn’t working well, they can give feedback, which helps improve service overall. That collaborative process makes a big difference.

[00:14:24] Steve MacDonald: Now, let’s look toward the future. We’re all talking about hybrid models—AI and agents working together. Based on where you are now, what’s your vision for where this is heading?

[00:14:40] Vladyslava Vozniuk: I see AI as the first-level responder, handling basic questions and issues. Then, agents should function as second-level support, taking on more complex cases.

[00:14:58] Vladyslava Vozniuk: At Usersnap, sometimes agents need to go into the platform, test integrations, or troubleshoot features to understand what’s blocking a customer. AI can’t do that. That’s why agents should handle the deeper investigative work, collect feedback, and also engage in proactive support—which is crucial.

[00:15:23] Vladyslava Vozniuk: Proactive support is one of the most important aspects right now. Reaching out to customers and identifying issues before they even notice them is a game-changer. That’s how I see the future, but so far, I haven’t found the perfect tool that fully enables this.

[00:15:43] Steve MacDonald: So, I’ve got an area to talk about that is traditionally seen as a big negative in the industry with human agents—the turnover issue. You put all this time and effort into training them, and then they learn as they work, becoming more intuitive and better at what they do. They can take on bigger and more complex problems and cases. But with AI, there’s no turnover rate.

[00:16:07] Steve MacDonald: Do you see AI helping to address the issue of agent turnover because it’s allowing them to avoid the mundane, lower-level customer service interactions and instead focus on more complex, interesting cases? What do you see regarding AI integration and its impact on agent turnover?

[00:16:34] Vladyslava Vozniuk: Yes, that’s interesting. I’ve never considered it from this perspective, but you’re completely right. The turnover issue exists with my team, though it hasn’t been such a big issue—maybe because of our well-structured training workflows. Our company has never really suffered from it.

[00:16:55] Vladyslava Vozniuk: Generally, it’s always complicated when an agent leaves, and you have to train someone from scratch. AI can act as a buffer because it can handle fundamental issues and questions, allowing human agents to focus more on the investigation—essentially stepping into the customer’s shoes to understand the problem deeply.

[00:17:20] Vladyslava Vozniuk: This way, agents get extra time to catch up on technical aspects and different features without compromising the support quality, thanks to AI.

[00:17:32] Vladyslava Vozniuk: AI can be a positive thing. However, I don’t think it’s valid if turnover is used as an argument against human agents. Human agents’ value cannot be compared to AI—they serve different roles.

[00:17:47] Steve MacDonald: Yeah, I love how you put it—that agents have to take on the role of the customer, stepping in to experience the same issue firsthand. AI can’t do that. You must get in there, push the button, click through the process, and see the problem. That’s a valuable perspective.

[00:18:10] Steve MacDonald: I’d love to ask you—because we’ve covered much about the agent’s perspective on AI—what haven’t we covered? What else do you think is important for us to understand about AI and its direction?

[00:18:24] Vladyslava Vozniuk: One thing I haven’t mentioned when discussing AI’s pros and cons is its significant technical limitations. For instance, we haven’t found a tool to read images, screenshots, or recordings effectively.

[00:18:45] Vladyslava Vozniuk: If you work in support, you quickly realize this is a big issue. Screenshots are the primary way to show problems with tech products, and recordings are invaluable. We often ask our agents to watch recordings and summarize them for developers.

[00:19:04] Vladyslava Vozniuk: AI can’t do this. And then, of course, there’s the issue of hallucination. Almost every AI tool I’ve tried generates its responses—especially when dealing with technical problems.

[00:19:15] Vladyslava Vozniuk: It’s great that AI can assist with technical responses, especially when not all agents have a technical background. But on the other hand, sometimes AI fabricates code, which can cause customer problems rather than solve them. So, while empathy, intuition, and other human qualities are crucial, we sometimes forget that AI isn’t technically advanced enough yet to replace agents.

[00:19:47] Steve MacDonald: Yeah, I never thought about that—shared screenshots, screencasts, or recordings that customers send in to illustrate a problem. That makes perfect sense.

[00:20:06] Steve MacDonald: I’d love to ask you—throughout your customer support career, what are you most proud of?

[00:20:06] Vladyslava Vozniuk: This is always a hard question to answer without sounding like I’m bragging, but I’m really proud of the level of customer satisfaction our company has achieved.

[00:20:34] Vladyslava Vozniuk: I already mentioned G2 reviews. Occasionally, I read them to ensure we’re on track with our support and to understand how people perceive the product. Seeing how pleased customers are with our support always warms my heart.

[00:20:34] Vladyslava Vozniuk: We won a G2 badge 2024 for Best Support and Customer Satisfaction. Our customer satisfaction rates are consistently high, and when I meet customers in person, they often mention our support as one of their favorite aspects of the product.

[00:21:09] Vladyslava Vozniuk: This makes me realize that support isn’t just about solving problems—it’s also about retaining customers and building strong relationships. That’s what I love about my job and our team.

[00:21:09] Steve MacDonald: What I take away from that is the idea that customer support is a favorite part of the product experience. We spend so much time on product development and marketing—marketing is one of the most significant cost centers in the industry. Then there’s sales support, and so on.

[00:21:20] Steve MacDonald: But the experience you create for the customer should be elevated to at least that same level of importance as the product itself. It’s one of the most significant differentiators—a massive competitive advantage for the company.

[00:21:42] Steve MacDonald: If we start thinking about it that way—because that’s what you just articulated—it aligns with what has been proven repeatedly. Many studies, including a significant McKinsey study, have shown that customer experience is the number one differentiator across all industries.

[00:21:42] Steve MacDonald: I’m glad you brought that up because achieving that kind of customer experience with AI alone is hard, right?

[00:22:06] Vladyslava Vozniuk: Yeah, true. I also just wanted to point out that that’s why customer support is not only important for retention but also for conversion. When trial customers reach out to support, it’s their first impression of the company. Apart from features and pricing, they also want to know that if any issues arise, they will be supported and receive immediate help. That’s why it’s also a great indicator of a good company.

[00:22:34] Steve MacDonald: Love that. So, we’ve talked about a lot. If there were one thing you wanted the audience for this episode to take away from everything we’ve discussed here, what would that one thing be?

[00:22:49] Vladyslava Vozniuk: The main takeaway is that AI is a great tool that will not leave our lives. I don’t think so.

[00:22:57] Vladyslava Vozniuk: But you have to remember that it’s just a tool and cannot replace all the benefits of human support. That’s something customers feel on a deep level, and studies prove it. So, you must treat AI as a tool, not an employee.

[00:23:16] Steve MacDonald: That’s a great takeaway. Thank you. Here’s one thing I’m pretty sure about—people will have questions after watching this episode. Would it be appropriate to share a link to your LinkedIn profile so people can reach out if they have questions?

[00:23:31] Vladyslava Vozniuk: Yeah, sure, of course.

[00:23:33] Steve MacDonald: Well, Vlada, thank you so much. I always interview people in customer support—from chief experience officers and chief customer officers to leaders like you in the support center—and you brought some fresh perspectives today.

[00:23:46] Steve MacDonald: The agent’s perspective and the importance of the agent often get lost because upper management usually focuses on us as a cost center—how to limit, how to contain. My takeaway is how can we give more opportunities, empower our agents, and strengthen their relationships rather than artificially box them in just to save money? That’s a very fresh perspective, and I appreciate you sharing it with us today. Thank you very much.

[00:24:19] Vladyslava Vozniuk: I’m glad. Thank you so much for inviting me as well.