Despite growing investment in AI and automation, enterprise satisfaction with AI solutions and service providers lags other technologies, according to a report by Information Services Group (ISG).
The report highlighted a significant gap between the expectations of AI and automation and the reality of their performance.
While enterprises are leveraging AI across various use cases, from chatbots to data analytics, many remain skeptical about the technology’s actual impact.
In fact, generative AI (GenAI) earned the lowest score among emerging technologies, rated at just 68.46 on a 100-point scale, with overall enterprise satisfaction averaging at 71.5.
While AI and automation led outsourcing deals in 2024, aimed at improving efficiency and cost savings, these technologies are not yet meeting their promises, the report noted.
Business Process Outsourcing (BPO) services earned the highest CX scores, due in large part to their focus on execution, delivery and meeting client needs.
The study noted AI and automation have played key roles in improving operational efficiency, scalability and decision-making in BPO, though concerns about data security remain.
“AI and automation projects can benefit from BPO’s best practices, such as proactive communication, flexible service delivery and tailored solutions,” said Heiko Henkes, managing director and principal analyst with ISG.
He recommended service providers engage clients early to understand their unique challenges and customize AI solutions accordingly.
GenAI Implementation Hurdles
Henkes said enterprises often face significant hurdles when implementing GenAI, especially due to unstructured data.
“GenAI requires well-organized data aligned with business requirements, but many firms lack this foundation,” he said.
This misalignment, coupled with the complexity of cloud transformations, makes it difficult for GenAI solutions to deliver meaningful results.
Talent shortages, inflated expectations and the rapid pace of technological change further complicate implementations.
Henkes said to enhance CX, service providers should offer comprehensive guidance throughout the GenAI journey, helping firms simplify deployments, align solutions to business needs and prepare for successful outcomes.
“Providers can improve CX by offering robust AI training, continuous post-deployment support and transparent communication to set realistic expectations,” he said.
He added that given economic uncertainty and concerns about return on investment (ROI), providers should work cost-effectively and showcase small, achievable successes to build trust.
“Notably, a slight increase in onshoring activities has been shown to significantly boost CX,” Henkes said.
Phased Approach to AI Adoption
While AI and automation offer substantial efficiency gains, their complex implementation can undermine customer satisfaction, as reflected in lower CX scores for these technologies.
Henkes recommended enterprises take a phased approach to AI adoption, focusing on areas with the greatest potential for efficiency gains.
Developing a maturity chart that maps potential use cases against benefits and urgency will help prioritize where AI can deliver the most value.
“To mitigate CX challenges, enterprises must prioritize clear communication about goals, timelines and risks,” he said.
He noted low-code/no-code platforms and modular AI tools can reduce technical barriers, making it easier for firms to integrate AI, while strong post-implementation support and workforce training are crucial for maintaining customer satisfaction and addressing issues early.
Aligning Expectations
Henkes said to align AI and automation with client expectations, service providers should begin with a thorough business requirement assessment before proposing AI solutions.
“Offering pilots or proof-of-concept projects is an effective way to demonstrate the value of AI before full-scale implementation,” he said.
He added it’s important to evaluate whether AI is the most appropriate solution for a specific business challenge.
“If a simpler, more efficient method exists, it should be considered first,” he said. “In cases where AI can deliver greater benefits, it should be prioritized.”
Service providers must focus on delivering measurable outcomes, such as cost savings or operational efficiencies, that clearly show the value of AI.
Moreover, conducting a thorough assessment—comparing the business requirements, technology being leveraged, and expected benefits—before building an AI implementation plan can help prevent inefficiencies and unmet expectations.
“Providing clear documentation, post-implementation training, and ongoing support ensures that enterprises can fully leverage the technology,” Henkes said.
