Today’s digital-savvy customers increasingly expect seamless and highly personalized engagement. In response, enterprises are looking for artificial intelligence (AI) to elevate customer experience (CX). In order to effectively raise the bar, organizations need to trade up indiscriminate use of AI tools and instead prioritize use cases that tackle high-impact customer pain points.
There’s no question that AI is already reshaping contact center and customer interactions. Research from The Everest Group shows generative AI (genAI) is a top contact center investment priority this year, with over 60% of enterprises surveyed planning to invest $1.5 million on genAI in CXM operations.[1] By 2030, The Everest Group expects AI to move from a standalone tool to an enabling technology for customer interactions, ushering in hyper-personalized experiences and predictive problem solving. The technology is also essential for empowering agents with real-time insights, emotional intelligence, and intelligent automation that will simplify and facilitate complex tasks.
The trick to realizing genAI’s full potential for addressing common CX challenges like response latency, agent inconsistency, and information accuracy is targeting the right use cases. With expectations running high and the possibility of AI fatigue setting in, Foundry reached out to the CIO Experts Network, a community of IT professionals and technology industry influencers, to see how they are prioritizing CX use cases that will benefit from AI integration.
Solving the right problems
The No. 1 rule for driving impact with AI is to start with the customer journey, not with the technicalities of a technology deployment, notes Michael Bertha, a partner at Metis Strategy. Through use of data and direct customer involvement—not internal assumptions—companies should explore what elements of their CX and customer support are causing friction. “That’s when it’s time to bring in AI,” Bertha says. “The value isn’t in the model; it’s in solving the right problem.”
IT and business leaders should identify high-frequency customer pain points—for example, high-touch, high-volume interactions such as support ticket triage or self-service support content. “Typically, this includes areas like reducing response times or improving self-service resolution rates,” notes Kumar Srivastava, chief technology officer at Turing Labs. “It is critical that those use cases be prioritized where quick wins can be demonstrated and business critical KPIs can be influenced.”
Use cases where AI creates measurable improvements in response times, accuracy, or customer satisfaction goes a long way in proving ROI and generating the quick wins that firm up executive and grassroots support. Departments with an appetite for automation tend to be strong pilot candidates as they are least resistant to change and more willing to influence others to get on board.
Starting small, scaling what works, and scraping what doesn’t is an important steppingstone to success. “Keeping the proof of concept lean yet data-rich and showcasing clear before-and-after metrics is what will garner buy in from the C-suite,” says Will Kelly, a writer focused on AI and cloud technologies.
As a general rule of thumb, use of AI to improve customer support functions is nearly always a good place to start, according to Issac Sacolick, (@nyike) president of StarCIO and bestselling author. By leveraging AI along with customer data, product information, and intelligence around customer touch points, organizations can deliver faster and more accurate customer service. IT leaders can tap into AI to improve tech-related self-service functions and enhance an agent’s ability to connect a customer issue with a tailored response. Additionally, AI agents can be developed for customer service managers, product managers, and IT to drill deeper into customer pain points to understand the most pressing issues.
“Organizations that improve these CX functions as a priority create feedback loops that improve products and services,” Sacolick adds.
Creating a structured scoring framework is a more formalized way to prioritize CX use cases that could benefit from AI. William Benjamin, principal generative AI/ML expert at Allstate, recommends enterprises evaluate potential AI use cases across these four dimensions:
- Strategic alignment: Determining whether the use case aligns with core business objectives;
- Business value: Quantifying the use case’s impact on measurable CX outcomes, including faster issue resolution, increased upsell conversions, or lower support costs;
- Technical feasibility: Assessing whether the company has the right talent, data quality, and architecture to execute successfully;
- Privacy, security, and governance: Determining whether customer data use complies with regulatory standards and internal policies.
“Strong change management, auditability, and model oversight are non-negotiable to scale AI initiatives responsibly,” Benjamin adds.
The challenges ahead
Of course, targeting the right CX use cases is critical, but not the only factor in ensuring AI success. Harnessing a cross-functional team of IT and non-IT stakeholders and collaborating with customers and employees helps build trust in the technology. In addition, expanding the pool of stakeholders is likely to uncover some not so obvious CX use cases for AI. “Collaboration, communication, and fostering a positive culture are essential keys to achieving wider acceptance of AI in CX,” says Arsalan Khan (@ArsalanAKhan), speaker, advisor, and blogger.
If past is prelude, companies have done a subpar job of articulating the value of emerging CX technologies like chatbots and the metaverse, and it appears genAI is no different. Poor customer experiences and weak scripts have frustrated users, and organizations must learn from those lessons as they move forward with AI. Building trust, delivering incremental value as opposed to “moonshots,” and building a solid data foundation are non-negotiable conditions for success, contends Peter Nichol, data and analytics leader for North America at Nestlé Health Science.
“AI should be deployed as a fit-for-purpose tool, not a universal solution,” Nichol says. “AI delivers when it’s aimed at a specific, measurable problem.”
[1] Everest Global, Generative AI and a Holistic CX Platform: A Perfect Pairing, 2025.