AI is well on its way to transforming the contact center, but no more so right now than with Agentic AI. I was the first analyst to write about this emerging trend last year, and my prediction about this being a leading trend for 2025 is very much coming to pass.
The concept isn’t quite mainstream yet, so let’s cover that ground before reviewing the implications for contact center leaders.
As with all technologies, AI is jargon-heavy, and the terms AI Agent and Agentic AI are often used interchangeably. Both refer to a form of automation that most of us think is still a few years out. But first we need to zero in on why AI is getting the attention it is.
What is Agentic AI?
A primary reason why AI is transformational – everywhere, not just for contact centers – is the speed of change.
We all know that AI is coming at us quickly, but less understood is how AI itself is enabling this.
AI can analyze data faster and cheaper than humans. This means that AI models are now being trained and built by AI rather than by human coders, developers, programmers, etc.
This reality is quite recent, and the good news for contact centers is that innovation is coming now at a faster pace than ever before.
A primary reason why AI is transformational – everywhere, not just for contact centers – is the speed of change.
Agentic AI is simply the latest development on the customer experience (CX) innovation spectrum, and just as this starts to take hold, the next wave of innovation won’t be far behind.
As such, contact center leaders need to view Agentic AI as the next CX driver. Implicit here is the need for an agile approach for how AI is transforming the contact center – this isn’t just about one wave of innovation to embrace – there will be continuous waves to manage.
Now let’s bring the focus back to Agentic AI, as this is a step-level change in terms of where and how automation brings value for the contact center.
To briefly explain, Agentic AI pertains to a new class of bots that can autonomously manage a customer inquiry end-to-end, with minimal or no human supervision.
What takes this beyond what existing bots can do is the ability to use human reasoning to make decisions, apply judgment, and even make recommendations to customers.
This set of capabilities has not existed before, and raises the bar for what’s possible with self-service and contact center automation.
Aside from the fact that this can provide a new – and very different – form of self-service, Agentic AI represents a major improvement over IVR-based options. In this regard, Agentic AI is an opportunity for all contact centers to up their game and could possibly be the best use case for investing in AI.
How Far To Go With Automation?
This really is the fundamental question, both with AI in general, and Agentic AI in particular. There is no question that contact centers need more and better forms of automation, and nothing does that better than AI. As customers make greater use of digital channels, their options for contacting businesses expand, and with that, the volume of inquiries will only increase.
...Agentic AI pertains to a new class of bots that can autonomously manage a customer inquiry end-to-end, with minimal or no human supervision.
Contact centers don’t have the luxury of hiring more agents, and legacy forms of self-service are woefully inadequate to do the job. Automation can be used in many ways to help manage growing call volumes.
This is where it’s important to understand the various use cases for AI.
Some of these are customer-facing, and this is where Agentic AI can take on certain types of inquiries, and automate them to a greater extent than IVR ever could.
Not only can they be routine inquiries like address changes or checking on the status of an order, but AI Agents can handle more complex situations, such as canceling a hotel reservation and making a new booking. These are high-impact use cases in terms of improving CX when done well.
Other use cases are behind the scenes, where there is no direct customer engagement. But the Agentic AI-driven automation makes for more streamlined workflows and fast execution of tasks that indirectly drive better customer outcomes.
A good example would be where a customer speaks to a human agent to request a refund. Once the agent approves it, the follow-on tasks are delegated to an AI Agent that could then autonomously manage things to complete the interaction.
If AI is not central to your thinking about modernizing your contact center, you will soon be at a competitive disadvantage that will be difficult to overcome without AI.
While these AI use cases sound promising, the logical extension for AI Agents is to fully automate customer service, and take on a growing share of the workload.
But while that is certainly a desirable scenario, Agentic AI is still far too new to take on the complex inquiries that human agents are often very good at handling.
To be fair, there are other scenarios where human agents fall short – remember, humans aren’t perfect – raising the thought that perhaps an AI agent could have done a better job.
In this regard, automation should be viewed as a way to help human agents better manage their workloads, rather than to replace them. To some extent, AI will replace human agents, but mainly in cases where agents cannot or will not adapt to working with AI to help make them better.
There will always be the temptation to evaluate AI on its potential to replace agents for cost reduction, but that only impacts operations. As businesses make CX a strategic priority, executive leaders will place more value on AI’s impact on customers. And that’s where Agentic AI can be a big step forward.
What CX Leaders Should Be Thinking About
This brings us full circle to the title of this article: are human and AI Agents better together? The answer can go either way, and it really depends on your business priorities.
If AI is viewed solely as a tool for cost-cutting, then the answer is no. In that case, AI-led automation will largely be about reducing your pool of agents, since they represent the largest cost factor for contact centers. Of course, a lot of things will have to go right for this plan to work, especially in terms that would not be detrimental to CX, such as:
- AI Agent customer interactions that feel human-like and conversational.
- The AI Agents stay in their lanes and don’t venture off-topic.
- Responses and outcomes are accurate and resolve the issues.
- Execution of the tasks are done in a timely manner.
As fast as AI is evolving, the above scenario is wildly optimistic, and based on current realities, CX leaders should be thinking differently.
If you fall anywhere on the spectrum that embraces AI for making CX better - instead of only improving operations - here are three things you should be thinking about to make the right decisions for your contact center’s AI journey.
1. AI is not a passing fad
Few technologies have been so heavily hyped, so it’s easy to think you can ignore AI until the hype gives way to a disappointing reality. You never really understood it in the first place, so better to let others take the risks, and just wait things out until the next technology wave comes around.
There is some validity to this thinking, but AI is delivering tangible results for contact centers today. Rather than being a passing fad, AI is poised to fundamentally re-set the entire model around customer service. If AI is not central to your thinking about modernizing your contact center, you will soon be at a competitive disadvantage that will be difficult to overcome without AI.
2. Agentic AI is a good form of automation
There are many ways to automate effectively with AI, but in terms of net benefit, none has more potential – right now – than Agentic AI. CX leaders don’t need to know the fine points about how Agentic AI works, but they do need to know about its iterative nature.
Initial deployments should be for fairly basic inquiries that AI Agents can handle end-to-end with little risk.
...position Agentic AI as a tool to augment human agent performance, rather than as a threat to replace agents.
By training these agents on your customer data, they will rapidly learn and be able to handle more complex tasks effectively. They are still far from perfect, but if properly managed, they will keep improving, and with AI’s native ability to scale, this type of automation will become your best solution to manage growing call volumes.
3. AI Agents and humans can be better together
This represents the ideal outcome where contact centers get the best of both worlds. Human agents spend less time dealing with mundane inquiries, and with real-time coaching from AI assistants, CX will improve, as will agent engagement.
Similarly, as AI Agents get better, they will handle more of the workload, helping to optimize operations. So long as CX is maintained – or improved – this will prove to be a better plan for AI automation than using AI Agents to replace human agents.
The key here is to position Agentic AI as a tool to augment human agent performance, rather than as a threat to replace agents. Only by seeing Agentic AI as a complement to their role, will human agents come to embrace this, and see the benefits of working side-by-side: instead of being a competition between man and machine.