Newer doesn’t always mean better.
Whether it’s a parent convincing a child that they don’t need another new toy because they hardly play with the 30 they already have, or the owner of a thrift store explaining that the cracked vase and dogeared paperback book they’re trying to sell you have ‘character’ –everyone has heard some variation of the phrase throughout their lives.
And that’s because there’s wisdom in it.
While the latest shiny new gadget, fashion must-have, or top-of-the-range car certainly draws the eye, they are more often than not unnecessary and unneeded.
So, where does agentic AI fit into all this?
We aren’t suggesting that the tech taking the customer service and experience space by storm isn’t impressive, far from it; it’s more the case that it might not be the right fit for every organization.
Unfortunately, figuring out whether or not agentic AI is the best option for your contact center or customer service department can be difficult.
According to Sabio’s Chief Innovation Officer, Stuart Dorman, this is partly due to the fact that many tech vendors are pushing agentic AI as a one-size-fits-all solution.
This can make it difficult to cut through the hype, particularly for companies that are less experienced and knowledgeable about the differences between agentic AI and other AI offerings.
For Dorman, at its core, agentic AI requires three things: a conversational interface, reasoning capability, and the autonomy to act on a customer’s behalf.
He said: “Some businesses aren’t quite ready for that full stack – but that doesn’t mean they can’t start somewhere.”
The Building Blocks to AI Success
Like many tech implementations, the extent to which an organization can reap the rewards is often determined by its pre-existing foundations.
In Dorman’s experience, to fully take advantage of agentic AI, businesses must have the following:
- Strong data infrastructure
- Accessible systems via APIs
- A joined-up view of the customer journey
“If a company doesn’t have those foundational elements – say their data is siloed or inaccessible – that doesn’t mean they can’t use AI, it just means the scope will be limited,” he explained.
Indeed, Dorman is particularly keen to emphasize the importance of data, which he describes as the “fuel for AI” and can be broken down into four areas: internal data, customer data, task completion, and consumer sentiment.
Before a company seriously considers implementing agentic AI solutions, it needs to ensure that its internal data and knowledge are accurate and up-to-date. Otherwise, the AI could produce incorrect information.
“That’s the simplest way it can go wrong,” Dorman explained.
From a customer data perspective, it is crucial that the systems used by organizations to store the data, such as CRMs and ERPs, are accessible.
In addition, a company’s systems must be designed to allow AI to take action and access the necessary information to complete tasks.
From the consumer side, agentic AI solutions from companies like OpenAI and Amazon require access to personal data, which introduces major trust concerns.
Dorman outlines how this is one of the major challenges to AI implementation:
“Are you comfortable giving that kind of access to a Meta or a Google? For many people, the answer is no – and that’s a real barrier to adoption we haven’t solved yet.”
Is the Juice Worth the Squeeze?
Although it may seem like a fairly obvious assessment, organizations considering introducing agentic AI solutions – and indeed any AI solution – must have a clear roadmap for how the tech will deliver ROI.
Where the waters become muddied is that many organizations aren’t clear on what return they’re aiming for or how to measure it.
And that’s a broader issue – not just in contact centers, but across enterprise applications of AI.
Dorman also discussed how timelines for ROI are shrinking.
“What used to be 12–18 months is now often expected in six,” he said.
“For example, we’ve seen a lot of investment in back-office AI tools and co-pilots, but not a noticeable jump in productivity yet.
“So, the pressure is on to prove value – quickly – and that’s where a clear strategy becomes essential.”
How Sabio Can Help
As a CX and customer service implementation specialist and expert services partner for organisations, Sabio can assist in determining which AI tools best complement their existing infrastructure and help them achieve ROI.
The company can also support businesses not currently suited for agentic AI by experimenting with other iterations of the tech.
Dorman details how the company will often begin by leveraging conversational AI to gather insights about what customers want to achieve – such as in this case with British Airways.
That data can then guide internal priorities, such as implementing new systems or exposing existing ones through APIs.
In doing so, Sabio can help organizations reach a point where they may be ready to introduce more advanced AI capabilities, like agentic AI, later on.
“Even when it’s being used to assist human agents, we still need to make sure the knowledge base is accurate, the data systems are reliable, and the user interface (UI) is well thought out,” Dorman said.
“For instance, surfacing real-time information is great for new agents, but might annoy a seasoned rep. So there are design and usability considerations too.”
Watch CX Today’s exclusive interview with Stuart Dorman here to learn more about Sabio’s agentic AI credentials.
You can also check out Sabio’s full suite of implementations and services by visiting the website today.
