AI in CX isn’t plug-and-play—here’s what leaders should know

In this special guest post, Jackie James from Quadient explains why successful AI in customer experience depends on taking a staged, strategic approach rather than handling implementation as a one-time fix.

When the topic is AI in customer experience, the conversation often jumps straight to the technology. Which tools should we use? How fast can we scale? What’s the next big innovation? In my experience, that’s not really where a successful transformation begins. At Quadient, my role is to manage the technology stack for our global contact centers, from AI-driven self-service solutions such as web portals and bot programs to the internal tools our employees use every day. And if there’s one thing I’ve learned, it’s that implementing AI in customer experience isn’t really about speed or hype, but sequencing, focus, and commitment.

AI in CX is a journey. It’s not something you switch on and walk away from; it requires thought, iteration, and the right support. And if I had to share the biggest lessons from our own experience, they would be these.

Getting started matters more than getting everything perfect.

When I first stepped into this role, what felt most important was taking that first step in our AI journey. That matters because it’s very easy to get stuck in planning mode. There is so much noise around AI, and now with generative and agentic AI adding even more buzz to the market, it can feel like you need to have the full long-term vision mapped out before you begin.

Strategy matters, but momentum is just as important. Even if you start small, it’s crucial to start somewhere. You learn by doing and understanding where customers and teams are ready, and where existing processes will support or slow down progress. One of the biggest mistakes is waiting for the perfect launch instead of building early learning into the process.

Sequence matters.

If I could start over, I would be even more deliberate about focusing on digital support first, and then layering bots on once that digital customer base is established. This is one of the most practical lessons for any leader thinking about how to implement AI in customer experience.

Driving digital and bot adoption simultaneously can be challenging. Digital adoption is not always immediate. Some customers take time to shift their habits, and if you launch bots on a digital platform before you have enough customers actively using it, you won’t see a return right away.

What I’ve found is that it works much better when each step is handled as its own component. First, help customers make the transition to digital support and build engagement. Then introduce automation into an environment where customers are already engaging. That makes it much easier to optimize each part of the journey and to understand what is driving value.

Take a data-driven approach.

One of the most important parts of any CX transformation is knowing where AI will have the greatest impact. Not every process should be prioritized at the same time, and not every workflow will deliver the same return. If you want to implement AI in customer experience to create real value, you have to focus on the right areas.

That means looking closely at the data to identify the processes that will deliver the greatest ROI. Where are the highest volumes, the most repetitive requests, and the biggest friction points? Where can digital self-service or automation improve the experience while also making the operation more efficient? Here is where AI changes from a technology to an operational conversation. It forces you to be honest about which journeys are working, which are not, and where investment is really going to move the needle.

Another lesson that shaped my thinking is that AI success depends just as much on people as it does on technology. An example that stands out for me is the work we did in France, where call containment improved from 10% to 44%. From the outside, it can look like a straightforward automation success story, but the real challenge was internal change management. We had to build trust, prove the product worked, and scale models that had already succeeded in other countries.

What really drove the improvement was reskilling the workforce to focus on bot optimization. We dedicated employees to reviewing transcripts, identifying workflow gaps, and understanding where customers still needed help. That gave us visibility into where journeys were breaking down and where the next improvements were needed. AI customer experience transformation is not just about deploying automation, but about building the capability to refine it continuously.

AI is not plug-and-play.

One of the biggest misconceptions is that AI is something you activate once and then leave alone. It doesn’t work that way. AI requires attention, training, refinement, ownership, and resources because implementation is only the beginning.

This is true whether you are working with self-service bots, digital channels, or broader AI CX transformation across the customer journey. The organizations that will get the most from AI are not the ones launching the most tools, but those that commit to continuous improvement. That means using conversation markers to see where customers drop out, simplifying steps where journeys become too complex, and identifying when customers shift channels because the experience is not working. In the long term, the bigger challenge will not be scaling automation, but scaling talent.

The beauty of the right platform is that once you have built strong process flows, it can be relatively straightforward to scale them across countries or business lines with some localized adjustments. The right platform can make it relatively easy to extend process flows across countries or business lines with only minor adjustments. What is hard is preparing people for a support environment that automates simple tasks, leaving agents to handle more complex issues. That changes how contact centers need to think about training and skills.

The real takeaway from this journey is that AI is not a one-time implementation; it is an ongoing discipline. It takes leadership, a clear strategy, focused execution, and dedicated resources to make it work for your customers, your employees, and your business. If I had to sum up how to implement AI in customer experience, I would say start somewhere, be intentional about the order of operations, let data guide your priorities, and never treat AI as something that can run on its own.