Just like enterprises were trying to make sense of Generative AI (gen AI) 12 months ago, agentic artificial intelligence (AI) is now entering that same exploratory phase. There is a mix of excitement, experimentation, and some AI-washing too.
We’re also now seeing transformation initiatives suddenly rebranded as “agentic”, with minimal real differentiation. That is, benefits profiles from these “rebranded” interventions are not significantly different from what they used to be. It’s exactly what happened during the initial wave of gen AI euphoria. But amidst the noise, something substantial is emerging.
Reach out to discuss this topic in depth.
What we are hearing from enterprises and service providers
The sentiment is clear: agentic AI holds promise, but its adoption even in mature Business Process Outsourcing (BPO) areas like Customer Experience Management (CXM) is still in early exploration.
There’s curiosity and experimentation, not yet scaled confidence. To determine whether enterprise functions truly require agentic AI-powered interventions, without being influenced by service provider proposals, buyers need to ask:
- Does the underlying process require agents to act?
- Does it require thinking or decision-making?
- Does it need context and historically learned behaviors to operate?
The underlying process and its associated complexity should guide whether a simple AI or an agentic AI solution is best.
Where Agentic AI can shine in CXM
Agentic AI is tailor-made for action-heavy use cases, such as sales support desks, ticketing and booking front desks. Sample use cases in the CXM space that seem to be emerging are:
- Agents that book/reschedule/cancel appointments via natural language processing
- Airline/hotel agents handling complex rescheduling and refund workflows
- Reservation agents using customer history and loyalty data to make tailored bookings
- Agents that work with multiple systems to respond to user inquiries without the need for hand-offs
From a deal commercials standpoint, there are two clear impact areas of agentic AI:
- Delivery models
Just like gen AI, the rise of agentic AI will reshape delivery models, drastically redefining shoring norms and ‘standard’ pyramids. This in turn will impact the Total Contract Value (TCV) of CXM deals signed with Full Time Equivalent (FTE)-based/transaction-based pricing
- Pricing models
Will agentic AI change pricing models? Yes, though the popular perception that value-based pricing will increase seems unlikely. Based on recent engagements with enterprises, we believe what they really need to figure out, is if they are ready to contract with task-based pricing. The idea of all tech costs (Application Programming Interfaces (APIs), models, infrastructure etc.) being bundled in and having a “one price per completed action” seems compelling, though whether it gets adopted or not will boil down to (a) how accurate the TCV estimation is, and (b) how favorable is it compared to traditional FTE-based/transaction-based pricing
Given the pace at which AI is progressing, it is likely that Agentic AI could be a standard part of CXM deals in 3 years with a high degree of confidence in answers to the above questions. But if you are an early mover who is trying to understand how your peers are adopting Agentic AI, feel free to contact us.
If you found this blog interesting, check out our Agentic AI: True Autonomy Or Task-based Hyperautomation? | Blog – Everest Group, which delves deeper into another topic regarding Agentic AI.
If you have any questions or want to discuss the evolution of Agentic AI in more depth, please contact Sahil Shah ([email protected]) and Rahul Gehani ([email protected]).
