Trends in CX and AI

In this Q&A, we sat down with OvationCXM's head of product, Alan Finlay for a discussion on the 2025 trends in CX and AI.

Q: What do you predict for digital transformation trends in banking in 2025?

Alan: I have a few predictions for digital transformation in 2025. First, I believe internal teams will stop operating as separate entities, throwing tasks "over the wall" to each other. Instead, we will see teams come together as a unified front to deliver cohesive experiences.

Second, there will be a continued push toward omni-channel experiences. This includes communication channels and digital portals where customers can access their updates. These elements will become more unified and streamlined.

Additionally, I expect a bigger focus on the end customer experience, particularly on the corporate side. While consumer banking has a long history of high-quality mobile apps and user-first experiences, corporate banking is catching up. In 2025, I foresee a stronger emphasis on creating seamless digital experiences for business customers, alongside individual consumers.

We’ll also see organizations prioritizing the connection of disconnected data silos. Many banks have data scattered across multiple systems, and by unifying these systems—especially with the application of AI—there’s a significant opportunity to extract more value and enhance customer experiences.

Finally, I think generative AI will play a larger role. Although enterprise adoption has been slow, especially in banking, internal teams will begin to align on controls and security measures, enabling wider adoption. This may start with internal use cases before expanding to customer-facing applications.

Q: How will AI and machine learning reshape CX in the next five years?

Alan: Over the next five years, AI will bring significant transformation to customer experience (CX). While we may not see a total overhaul, there will be large-scale improvements in user experience (UX), workflows and even the roles people perform.

The most critical factor will be high-quality, accessible data. Data is the fuel that powers AI for CX, and companies will continue to invest in refining it. In the software space, I expect to see a shift toward vertical-specific solutions. Specialized AI tailored to specific industries and problem sets will outperform generalized models in terms of quality, speed and cost-efficiency.

From a customer’s perspective, these advancements will result in faster, better and more unified experiences. We will see more high-quality automated responses, real-time interactions across channels, and personalized engagements based on customer preferences and behaviors.

AI will also facilitate collaboration across companies with shared customers, enabling seamless, unified experiences. Whoever owns that unified experience will ultimately win the long-term customer relationship.

Another notable shift will be in how teams spend their time. Tasks like researching updates, following up and drafting communications will decrease as AI takes over. This will free up time for employees to focus on solving complex problems and improving systems. It’s a win-win, allowing teams to focus on high-impact work while automation handles routine tasks.

On the software side, we’ll see more conversational interfaces and natural language-driven designs, replacing grids and forms. Real-time helpers embedded within applications will simplify user onboarding and reduce the need for extensive training.

Finally, AI will start taking more direct actions. Today, it excels at predictions and generating responses. In the future, it will progress to executing tasks, completing the swing from insight generation to action.

Q: What is agentic AI, and how will it be used in 2025?

Alan: Agentic AI involves autonomous agents capable of making decisions, planning actions and adapting to changing situations with minimal human oversight. It’s a step beyond traditional AI, where agents are tasked with achieving specific outcomes and can independently decide how to reach those goals.

In the early days of AI, humans programmed systems to follow patterns. Over time, these systems evolved to generate complex outputs like entire books, images and videos. Agentic AI builds on this foundation by enabling agents to act autonomously, completing tasks without needing humans to guide every step.

This approach has enormous potential. Companies could deploy hundreds or thousands of these agents, each designed to perform a specific task. These agents could collaborate to achieve broader organizational objectives, simplifying the design and maintenance of complex systems.

In 2025, we’ll likely see early adopters experimenting with agentic AI, though large-scale adoption—especially in financial services—may take longer. Enterprises will need to address security concerns and ensure agents align with regulatory requirements. That said, the foundation laid in 2025 will set the stage for more widespread use in the years to come.

Q: What is an example of agentic AIs working in the real world?

Alan: One example that often comes up is in travel. You ask, “Where do I want to go? What are the dates? What are my options?” Instead of just showing itineraries, the agent books everything for you: hotels, flights and plans. You’re set to travel.

In our CX world, it could look like an email that comes in with a question, for example:
"What’s the latest on this piece of information?"

An agent might query the data, find what’s needed—say, “how much you’ve processed over the last 30 days”—extract that and generate a direct email response:
"You’ve processed $30 today and $100 in the last 30 days. Do you have any other questions?"

These agents work together across specific paths to orchestrate multi-touchpoint experiences.

Right now, most organizations think of journeys as structured: A leads to B, B leads to C. But life isn’t linear—things happen dynamically and in parallel. Imagine defining a desired outcome while AI agents adapt, assess and take actions dynamically.

For example, agents could send emails, onboard customers, make decisions and handle tasks in real time. Even if the path isn’t pre-built, the customer still reaches the desired outcome. It’s a dynamic orchestration of multiple agents rather than a rigid workflow.

Q: Why should financial services focus on customer journey orchestration in 2025?

Alan: As the sector grows—offering more products and services—the operations often remain disjointed. Enterprise-level companies end up with silos of teams working together but without true integration.

Scaling by adding more teams or people only makes the problem worse. That’s why organizations of all sizes are prioritizing unified, orchestrated customer journeys as a key objective for the next year.

Instead of dealing with disconnected experiences and hoping for the best, companies are shifting to:

1. Real-time visibility into every customer’s journey.
2. Insights on friction points and churn risks—powered by real-time data.

Without observability, you’re guessing based on anecdotal evidence like a few emails or survey responses.

Throwing more people at the problem doesn’t work long-term—it breaks at scale. This issue is compounded by the evolving expectations of end users. Even on the corporate side, where adoption often lags, decision-makers are consumers in other parts of their lives. They expect the same ease, visibility, and collaboration they experience elsewhere.

To meet these expectations, companies must connect disconnected teams, systems and organizations into a unified flow. That’s how you deliver a seamless, purpose-built customer experience.

Those who ignore this shift risk losing customers to competitors offering seamless onboarding and servicing. You’ll see churn and attrition among customers seeking better, integrated solutions.

Q:  Why should companies prioritize investing in ecosystem connections in 2025?

Alan: Investing in ecosystem connections is vital—though not every business operates in an ecosystem that demands high-touch visibility and collaboration across organizations. But for those that do—like enterprises or financial services working with multiple partners—seamless ecosystem connections are the key to building high-quality, long-lasting relationships at scale.

When your company is part of an ecosystem and depends on partners or vendors to deliver on your brand promise, you need to ensure they treat your customers as you would. This requires real-time visibility and collaboration. Without it, it feels like dealing with four or five different companies, leading to fragmented and disconnected experiences.

Without these connections, silos emerge, and you’ll inevitably end up with finger-pointing:
"That's not our problem; it's theirs."
Nothing frustrates customers more than being told to contact someone else because you can't solve their issue.

To truly own the customer relationship within an ecosystem, you need visibility and collaboration layers that enable a unified experience. This investment ensures your customers feel supported, regardless of which partner or team is involved.

Q: How can companies overcome customer reluctance about generative AI use in their customer support?

Alan: Overcoming this reluctance starts with quality. High-quality output is essential. The short answer is: If you provide fast, personalized, and accurate responses, people will embrace it. But realistically, AI isn’t perfect. So companies should adopt a cautious, gradual approach.

Start Small:
Begin with internal use cases before deploying customer-facing AI. For example:

• Summarize long documents or complex content.
• Provide AI-generated search results with links to validate information.
• Generate draft responses to save time while allowing human edits.

This builds internal confidence in AI and helps teams understand how a human-AI model fits into their workflows.

Create an AI Champion Role:
Assign someone internally to oversee AI use cases and the quality of outputs. This person would:

• Review results and gather feedback.
• Update content and refine prompts.
• Drive adoption while ensuring quality.

This role could eventually lead to a team that transforms how your company uses AI.

Customer-Facing Deployment:
When you move to customer-facing use cases, ensure continuous monitoring of data and outputs. If there are bad or missed responses, identify whether the issue lies with missing content, inaccurate information, or incorrect prompts. Constant review and iteration are key.

Engage Customers Directly:
Ask customers for their opinions on AI. When done right, many users will appreciate the speed, accuracy, and ease of use. Your internal teams will benefit, too, as AI saves them time to focus on higher-value work.

That said, there will always be customers who prefer human interactions, and that’s okay. Give them the option. Let customers choose their preferred path—whether that’s interacting with AI or with a human—and you’ll achieve better outcomes and satisfaction for everyone involved.

Also, avoid the temptation to replace everything with AI right away. Many AI-only companies focus exclusively on conversational AI, which can be incredibly powerful, but also potentially intimidating. What works best, we've found, is when AI integrates seamlessly into existing user interfaces (UIs) and workflows that people are already comfortable with.

In the CX space, if you completely overhaul how you interact with customers, it can be challenging to get people on board with the change, as it disrupts their daily workflows. Instead, consider adopting a CXM platform where AI is integrated natively. This way, as you go through your daily processes, the AI is there, prompting, helping, and guiding you without a complete transformation overnight.

I believe this will be a major transformation over the next decade, but in the short term, integrating AI into existing systems is the best way to help people ease into it.

Q: What tools are needed in financial services—or any complex organization, really—to fast-track innovation and efficiency in 2025?

Alan: In 2025, innovation and efficiency will largely be driven by AI. While tools like routing, chatbots and communication workflows have become table stakes and are still important, they won't drive the next wave of efficiency.

AI will play a major role in unlocking efficiencies, and the key to this is data. The better your data and the more structured it is, the more value you’ll unlock. So, here’s where I think organizations should focus:

1. Accessible Data
   Many organizations have great data, but it’s often locked away in systems that aren’t easily accessible. Data that’s “at rest” doesn’t do much for you. What you need is data that’s accessible through APIs—data that can be accessed by other systems to unlock valuable use cases.
2. Connection Points Between Systems
   Once you have accessible data, you need a way to connect it across all systems. Your CXM platform needs to integrate with all your data platforms, including communication systems. The more connection points you have between systems, the fewer tabs you need open and the more unified your users' experience will be.
3. Orchestration Layer
   The final piece is an orchestration layer. With data and systems connected, you need a layer that can bring everything together. This layer will allow you to design, build, and implement multi-touchpoint experiences, eliminating disconnected touchpoints and providing visibility across all systems, tools, and data sets.

This orchestration layer enables automation, streamlines processes, and provides insights that were previously unavailable, offering a new roadmap for driving efficiency gains and continuing to improve workflows.

Once you have your accessible data, you don’t necessarily need to build all the integrations yourself. There are plenty of tools and pre-built connectors available that allow you to build a connection once and then connect all your systems off the shelf. These tools help speed up the process of managing and moving your own data.

When it comes to orchestration tools, there are systems available that can help manage touchpoints across all these disconnected systems. Some of these orchestration tools even come with built-in data connectivity, so you won’t need to create all those connection points yourself. Often, these tools already have pre-built connectors to the systems and data storage that you use, making it easier to get everything talking to each other in a no-code or low-code way. This simplifies the orchestration process and speeds up implementation.

Q: What’s the fastest way for a company to optimize customer onboarding and support?

Alan: I don’t necessarily think “fast” is always the best approach. I’d say it’s important to start small. You don’t need to change everything at once—and, in fact, you probably don’t want to disrupt everything right away.

If we’re talking about optimizing complex, multi-touchpoint processes like onboarding, I think the fastest way to add value is by introducing an orchestration tool. This connects the dots between different processes, without having to change any underlying systems.

At OvationCXM, when helping clients, we look at processes like onboarding and map out the ideal customer journey. On day one, we don’t automate anything. Instead, we focus on identifying and mapping out the different touchpoints and unifying them across teams, systems and organizations. After running customers through this manual process, we gain insights into where friction points exist, where customers churn and what areas are prime for automation.

With that data, we can create a roadmap for automation, layering it on top of the visibility layer. This helps us optimize and scale processes efficiently.

Regarding customer support, I believe the fastest way to make a real impact is by bringing everything into one CXM platform. This allows you to centralize all customer information, communications, workflows, automation, cases and knowledge in one place. Having everything natively integrated eliminates the need to pull from different systems just to answer a simple question.

For example, if a customer chats with you and then calls, you want both interactions visible in the same place so your team can provide a seamless experience. By consolidating communication channels in a CXM platform, you’ll see an immediate lift in customer satisfaction (CSAT) and increase collaboration across teams. This centralized approach will create a noticeable improvement in support efficiency and customer experience.

Q: So, what’s ahead for OvationCXM in 2025?

Alan: For OvationCXM in 2025, we’re focusing on a few key themes.

The first is customer journeys. We’re looking to provide more flexibility in how you build and design these journeys. We’ll be adding new actions that unlock new experiences, including API-based actions, connector-based actions, and forms with contextual data. This will empower things like digital onboarding experiences and a host of other new actions that will emerge throughout the year.

The second theme is the customer portal. We want to extend the customer journey by providing a way for customers to track their progress and set expectations. Whether it’s through a separate portal or a widget inside an existing one, the goal is to give customers real-time insights into where they are in their journey, what comes next, and how long it will take. This will help them stay on track, complete necessary tasks, and get notified when they need to take action. It’s all about eliminating disconnected emails, chats, and phone calls by centralizing everything in one place, so both the customer and internal teams can collaborate more effectively on complex processes.

The third theme revolves around the ecosystem. We’ll be expanding our ecosystem to connect any data, system, or organization that needs to share visibility and collaborate in real time. This will involve an increase in pre-built system connections, which will help orchestrate experiences across various systems and organizations. It’ll also help automate previously manual tasks, and we plan to add several major new organizations to the OvationCXM network. This will create greater collaboration within the financial services industry, enabling real-time connections and easier collaboration across different players in the ecosystem.

The fourth theme is AI. We’re going to improve our core AI offerings, with a big focus on generative AI. This will include use cases like experience summarizations, holistic customer and ecosystem insights, and generative AI chatbots and voice bots. We’ll also be adding co-pilots for internal agents to make it easier for them to do their jobs. Additionally, we’re focusing on deeper journey insights and recommendations, as well as connecting more data sources to AI agents to provide richer, more context-aware answers.

In terms of core functionality, our focus will be on things like authentication, data structures, email and communications, notifications, and analytics—basically, the foundational features that power the entire platform and ensure a smooth experience for users.

Finally, security will be a major focus for us. We’re aiming for SOC 2 compliance certification in Q1 and will continue to build on that, as security will always be a key ongoing focus.

Q: What are the biggest pain points you hear from companies you want to transform customer onboarding and support? 

Alan: We work with a lot of enterprises, and one of the things we do is journey mapping. We dive deep into understanding core processes and workflows, identifying what’s working and what’s not.

A recurring theme we hear is the disconnected nature of these experiences, especially in large organizations. Often, multiple teams—three, four, five different ones—are involved in onboarding a customer. In financial services, this is even more complex, with several different organizations needing to be engaged, particularly around payments.

Right now, the process is often something like this: I create a case, assign it to someone, and then I have no visibility into when it will get done, whether it's done, or what the next steps are. There's no real collaboration between these teams or organizations. The work gets done eventually, but if you ask, “Where is this customer in the experience?” the answer is usually vague. You might know it’s somewhere within the first 30 or 60 days, but you can’t pinpoint the exact status or identify where the customer might be stuck. There’s no unified layer that connects all these touchpoints.

One of the biggest pain points we hear repeatedly is: How can I create a seamless experience? How can I link all these disconnected actions and get visibility into how long a customer spends at each stage of their journey? This insight allows organizations to identify friction points and understand where to focus improvement efforts.

Because if you don’t know where your biggest problems are, how can you know what to improve first? You can guess based on customer frustration, but with a unified layer, you can rely on data to tell you exactly where to focus. That’s how you prioritize the areas that will have the biggest impact and provide the highest return on investment for the year.

When organizations try to tackle the problem of onboarding, for example, they often approach it in various ways, but it's usually not ideal.

Some organizations use shared Excel documents and Outlook inboxes, sending emails back and forth. They just share an inbox and respond as the emails come in. At the end of the day, it works—but it’s far from a great experience.

A step up from that might be creating separate queues for different teams. So, you might create cases for the legal team, the operations team, and the customization team. While this helps orchestrate the work, the problem is there’s no consistent visibility. It works until you try to gather any insight from the process. When you try to understand things like where people are in the journey or how much time has been spent, it’s nearly impossible.

Ultimately, it’s not that the customer will never get onboarded—it’s just a bad experience. This creates challenges internally, with operations teams lacking visibility, and the end customer has no idea what's going on behind the scenes. For leadership, there’s no clear insight into the most critical process of bringing on new clients.

We’ve seen this issue tackled in many ways. But the toughest part is getting out of the status quo—realizing that the old methods are no longer good enough. The key is to adopt a system that allows you to orchestrate these processes, gain visibility into them, and create great customer experiences, instead of just settling for what’s been working, even if it’s not optimal.

I think what we're seeing in 2025 is a shift in mindset. Organizations are starting to recognize that the old way isn't sufficient anymore. The technology and systems are there. The competition is there, too. If you don’t evolve, your customers will gradually migrate toward the companies that are providing seamless onboarding experiences, especially in industries like financial services.

Startups, with API-first and tech-driven approaches, are entering the space with beautiful, seamless onboarding flows. If you don't keep up with that, you risk losing customers due to ease of use.

So, now is the time to move away from the status quo that has worked for the last 20 years and start thinking about what's next. The systems will work for now, but in the next 12 to 24 months, they won’t cut it anymore. More and more organizations are realizing that and are investing in 2025 to shift from outdated methods to systems that will actually drive business and deliver better customer experiences.