Since the launch of ChatGPT, customer experience software vendors who have leveraged large language models (LLM) are supplying their customers with capabilities not seen before. Customers are seeing AI Agents able to handle 70% of tickets with human-like precision and provide real-time and personalized answers while being available 24/7 across almost all old and new channels. This is not the old chatbot where it overpromised and under-delivered. AI Agents are providing delightful and low friction experiences for their existing and future customers while reducing the costs of providing such services. AI Agents are faster, more consistent, more patient, and sometimes more resourceful. Some can even complete the actions on behalf of customers now.

They don't always have the answers but like Google's AlphaGo that improved in a short period to beat the Grand Masters in Go, these AI Agent systems are also improving at a remarkable pace. Like Amazon next day delivery that reset the expectations, customer support will go through a renaissance with AI.

The early adopters of LLM-based AI Agents are reporting strong results. They see clear ROI and the process of getting there has been easier than ever. It still requires decent handholding from vendors and resellers because these are very new. However, the advantage is meaningful. New winners will emerge and gain market shares because they are able to provide fast and immediate self-service that accurately answers questions and takes care of needs without long waiting time nor bewildering menu options. If there is one area amongst all the AI initiatives where any organization should consider, customer services is a clear place to start. It has the most proven number of examples that it can consistently improve customer happiness, improve output, increase revenue, reduce costs and provide strong ROI.

Based on the 2,340 customer adoption examples across the 50+ vendors providing various AI Agent solutions, we observe a few trends. First, the adoption is broad across 70 industries and all sizes. Technology companies including software, services and hardware represent the largest concentration of adopters at 26%. These players understand the fundamental shifts that are happening in AI and they are early converts. Other industries such as retail goods (9%), financial services (8%) and hospitality (3%) are right behind.

Many have first deployed this capability first as a co-pilot for their own agents and now deploying the full autonomous agents. Given any enterprise adoption and internal approval takes at least 6 months, if you as the leader make a decision in Q1 2025, it won't be until late 2025 that your company begin to see the benefits. Many enterprises with much more complex workflows have also chosen to roll their own – we think those will take longer and bump into more challenges. The technology is moving too fast for internal or even external consultants to keep up. We recommend that unless you really have the resources to build your custom AI Agent plus the patience to consistently improve what you have built, there are many great vendors available to you.

The LLM-based AI landscape will only accelerate the change. There are already 50,000+ new AI Agents companies across all business departments and human tasks. In customer support, we see a future where AI will be able to handle not just texts, emails, social media chats, and calls, but soon appear on Zooms and videos as a human-like avatar with voice as well.

LLM-based AI Agent will usher a new era in customer experience.