Being a new ICMI Contributor, many of you may not know me, but as a 20+ year independent technology analyst, I’ve been writing and speaking extensively about contact center and CX topics for many years. My initial focus as a contributor has been on strategic planning, for which I have a pretty broad perspective. This focus began here, where I outlined the need for taking a strategic approach for deploying AI in the contact center.
There are many strands to explore on this front, and I will address two over this post and my next one. Regardless of the business challenge, a strategy must consider which actions will have the greatest impact on meeting the objectives of the business.
For the contact center, two objectives should be paramount: improving customer service and optimizing operational performance. One is internal and one is external, and when it comes to AI, the benefits should be mutual. Deploying AI must be in the service of supporting both, and these two objectives are highly interdependent.
I’ll address the former in this post, then operations in my next article. To help contact leaders develop a framework for their AI strategy, here are two building blocks for setting your plan in motion.
Improving Customer Service: Priorities for AI
Setting priorities here requires careful consideration, especially if you’re early in your AI journey. Conventional approaches to customer service have been built around problem resolution, and reactively addressing inquiries as they come in. With legacy contact center technologies, agents had little more than CRM data to work with, and usually operated with great uncertainty, not just for understanding the issues, but also how best to resolve them.
The transformational nature of AI takes the contact center well-beyond this model, where the bar for customer satisfaction has always been low. As such, AI should be viewed as an opportunity to rethink how customer service can be improved, especially since customer expectations are so much higher now. This is largely due to customers becoming more technology-centric, and contact centers need to keep pace. In this context, deploying AI needs to become a high priority, since it’s the best and fastest way for contact centers to get on a level playing field with customers.
In this regard, the priority would be to deploy AI in order to lead the way for improving customer service base on today’s realities, rather than the legacy model outlined above. Key to this approach is understanding how data-centric customer experience has become with today’s digital technologies. Only with AI can all of this data be harnessed to provide the personalized forms of service that customers really value. This simply wasn’t possible before AI. The sooner you are ready to adopt this new model of customer service, the sooner your AI deployment can yield strategic value.
Improving Customer Service: Expectations for AI
With the above priority in mind, your strategy must also have realistic expectations. AI’s near-magical qualities make it look like a silver bullet for all your problems, and that trap must be avoided. While AI has been a work-in-progress since the 1950s, it is far from fully formed in 2026. Any AI deployment will be piecemeal, where small successes build into larger successes. The key idea here is that AI technologies are iterative, meaning there is constant learning, and the accuracy of applications will keep improving.
These expectations also help mitigate the risk of AI negatively impacting customer service. For that reason, initial AI deployments should be internal, usually for automating tasks and processes. As these capabilities are refined, there will be less downside extending AI to customer-facing applications.
This will also help contact center leaders determine the right amount of AI to use, the right situations and the right types of customers. An across-the-board approach to AI is not strategic. Rather, the focus should be on where the impact will be greatest with the least amount of risk. Doing so means giving customers choice as to when AI works for them and when they need to connect with a live agent. Some customers will be comfortable using AI most all the time, whereas others only in specific situations.
Conclusion
AI has only been framed here at a high level, and more discussion is needed around the various modes – voice, text, video – and use cases – routine inquiries, urgent needs, multiple queries, etc. Customer service takes all these forms, and when deployed strategically, AI can be effective for all of them. In fact, as you get further along with AI, you will likely need a specific strategy for each of them.
Just thinking about an overall AI strategy, much of this depends on your ability to map out customer journeys to find the right points of engagement with AI, but also the quality of data you have on your customers. The better they are, the better your AI deployments will perform.
As such, you should consider doing audits for each of these before framing your AI strategy. In many cases, customer data is highly fragmented, incomplete, inaccurate, difficult to access and stored across a messy mix of formats. All of these will impact the ability of AI to bring the transformative change needed to meet today’s customer expectations.
These factors will be the underpinning of your AI strategy, since AI is entirely data-dependent, and customer journeys help identify the pain points for your customers. The starting point; however, should be setting priorities. AI is not a quick fix for all your customer service challenges. Paired with that, you must have realistic expectations for where and how AI can impact customer service.