AI for better CX

Executive Summary

Brands have been using artificial intelligence (AI) in customer interactions for almost a decade, with intelligent chatbots being a prime example. The rapid ascent of generative AI has opened management's eyes to the technology's vast potential to improve customer experience (CX). Hyper-personalization, automated note-taking and call processing, and real-time prompts for responding to customer queries are just some of the ways this technology could help CX teams. However, the report questions whether they are using it to good effect.

A survey of 5,000 consumers and 500 executives across seven countries revealed that while brands are actively developing CX use cases for AI and see it paying off in areas like personalization, customer support, and loyalty, most consumers are not seeing improvements in personalized experiences attributed to AI. Consumers also feel that brands are not providing enough access to humans in automated interactions.

Key experts interviewed for the report include John Aylward, Morton Bell-Izzard (Exelon), Abhii Parakh (Prudential Financial), and Stacy Sherman (Doing CX Right).

Our Findings in Short

• Brands must do more to pass on their internal CX benefits from AI to customers: While executives report improved CX and loyalty metrics, consumers are frustrated by a lack of access to human agents (cited by 47%).
• Companies need to overcome data constraints to do better on personalization: Despite executives seeing benefits, consumers are more likely to say personalization has detracted from their CX. Data privacy (65%) and data quality (46%) are major pain points for companies.
• Consumers are not ready to give up human contact: Consumer satisfaction is higher in human-led interactions (88%) compared to AI-driven ones (60%). Consumers prefer easy access to humans.
• The hunt is on for new ways to measure the impact of AI: Companies are adapting existing metrics (53%) or developing new ones (33%) to measure AI's impact, driven by the need for ROI evidence.

Key Statistics:

• 88% of consumers are satisfied with interactions handled by human agents vs. 60% satisfied with AI-driven interactions.
• 33% of brands are looking to develop entirely new metrics to measure AI's impact on CX.

AI in CX: What Do Brands Gain?

Most executives surveyed believe their businesses are benefiting from using AI in CX. In the past two years:

• 72% of executives say their companies' CX metrics have improved, with 52% attributing it to AI.
• 66% say customer loyalty metrics have improved, with 60% believing AI played a significant part.

Brands are seeing a positive AI impact on CX:

A bar chart shows the perceived impact of AI on CX and loyalty metrics. For CX, 37% reported a considerable impact, 31% a moderate impact, and 23% a great impact. For loyalty, 32% reported a considerable impact, 29% a moderate impact, and 20% a great impact.

While executives report positive impacts, quantifying AI's impact remains a challenge. Some companies are seeing gains in employee productivity. Abhii Parakh of Prudential Financial noted that AI saved their marketing team up to 75% in time for first drafts of marketing content and reduced customer research synthesis from three weeks to three minutes.

The biggest AI gains are in productivity:

Customer support is the most common AI use case (75% of executives). Exelon uses AI to help representatives handle calls more efficiently, summarize calls, and provide data at the right time. Personalization is another major use case (71%), followed by data collection and analysis (69%). Executives believe AI delivers the most CX benefits in these areas.

A bar chart illustrates AI use cases and areas of greatest benefit. Customer support leads in both main use cases (75%) and greatest benefit (47%). Personalization follows with 71% main use and 44% greatest benefit. Data collection and analysis shows 69% main use and 37% greatest benefit.

How AI helps to keep the lights on:

Exelon used AI and predictive analytics to identify customers who might struggle with energy bills during the COVID-19 pandemic. This enabled a targeted outreach program with personalized assistance recommendations. Exelon plans to use AI to streamline the application process for energy assistance funding.

Consumers' Verdict: Mixed

Consumers surveyed are less positive about AI's impact on CX than brand executives. While generally relaxed about AI-only interactions, their comfort levels vary by task type. Nearly two-thirds are comfortable with AI-enabled purchase transactions and product inquiries, but fewer are comfortable when AI handles complaints.

Most consumers are relaxed about machine-enabled interactions:

Pie charts show comfort levels: 64% for purchase transactions, 63% for product inquiries, 58% for changing account information, and 52% for resolving a complaint.

Consumers' most common automated interactions involve banking, financial services, retail purchases, and health products. 60% report satisfaction with automated interactions over the past two years, but 88% prefer human-led interactions.

Consumers strongly prefer human-led interactions:

A bar chart compares satisfaction with automated vs. human interactions. Human interactions consistently show higher satisfaction across the board, with 88% reporting satisfaction compared to 60% for automated interactions.

No human involvement is the top source of consumer frustration:

The biggest annoyance for consumers is the inability to access a human agent. Brands also recognize this as the main complaint from customers about AI-enabled interactions. Morlon Bell-Izzard notes that even with AI empathy, it may not satisfy all customer needs.

A bar chart shows consumer and brand frustrations. For both, 'Inability to speak or (text) chat with a live sales/customer service agent' is the highest frustration at 47% for consumers and 48% for brands. Other frustrations include limited options, slow response, awkward language, and information loss.

AI has not transformed personalization yet:

Consumers are unimpressed with AI personalization, with 30% saying it detracted from their experience, 26% saying it improved it, and 44% reporting no impact. This is a concern as personalization is a key AI use case.

A series of pie charts illustrate the impact of personalized AI interactions on customer experience. The largest segment (44%) indicates 'Neutral (neither improved nor detracted from it)', followed by 'Somewhat detracted from it' (20%) and 'Significantly detracted from it' (10%).

Data privacy rules limit companies' use of AI for personalization (65% of executives). Stacy Sherman emphasizes the importance of transparency in communication about AI use to build customer trust and relationships.

Data, Skills and Metrics Are Holding Companies Back

Data issues, inadequate data quality, and privacy rules are significant challenges for brands using AI in CX operations.

Data and skills issues are the biggest problems for brands when using AI for CX:

A bar chart shows chief pain points: Inadequate data quality (46%), Skills gaps in existing staff (46%), Compliance with data privacy rules (39%), Security concerns (35%), and Limited customer acceptance of AI (34%).

Companies need to ensure data availability and quality for AI use cases, such as predictive models for customer complaints. Gaps in AI-specific skills are a long-standing issue, exacerbated by generative AI. Recruiting skilled experts (41%) and training existing staff (30%) are top priorities.

Customer feedback is executives' top priority for improving their use of AI:

A bar chart shows priorities for improving AI use. Top priorities include improving customer feedback mechanisms on AI use (49%), improving compliance with data privacy rules (43%), and investing in the recruitment of skilled AI experts (41%).

There is work to do on measuring AI's impact:

Nearly one-third of executives prioritize developing new metrics to measure AI's impact on CX. Currently, 53% are adapting existing CX metrics, while 33% are developing new ones. 14% have no metrics yet.

A bar chart illustrates how companies measure AI's impact. 'We have adapted/are adapting our existing CX metrics' is at 53%, 'We have developed/are developing completely new metrics' is at 33%, and 'We do not yet have metrics that measure the AI impact' is at 14%.

Companies like Prudential Financial are exploring established metrics as proxies, while Exelon is devising new metrics. Measuring AI's ROI is considered challenging but an area of opportunity.

Is the Human Role in CX Secure?

Executives are unsure if CX operations will ever be fully automated by AI. About 26% believe the human element will always be retained.

Spending on CX will target both AI and human improvements:

A stacked bar chart shows future investment balance: 44% anticipate an equal mix of AI-driven and human-driven improvements, 23% lean towards mostly human-driven, and 22% towards mostly AI-driven. Retail is an exception, with 52% prioritizing AI-driven improvements.

44% of brands will place equal investment priority on AI-driven and human-driven improvements to CX.

A hybrid approach requires customer-facing staff to be upskilled to work effectively with AI. Priority training areas include handling customer complaints about chatbots, understanding AI prompts, and handling data privacy complaints.

Handling customer complaints is a focus of AI training:

A bar chart details AI training priorities. Top areas include handling customer complaints about quality of interactions with AI chatbots (43%), understanding AI prompts (43%), and handling customer queries or complaints about data privacy issues (40%).

Companies can use AI for training simulations, allowing employees to practice interactions and build confidence without fear of repercussions.

Employee adoption of AI:

AI can help employees overcome fear. Companies should acknowledge and address the emotional and psychological barriers employees have regarding AI. Transparency about AI usage and its benefits for skill development is crucial.

Abhii Parakh notes that initially skeptical employees became excited about AI tools after using them, showing a shift in comfort levels.

AI in CX: Cracking the Code

While generative AI has expanded CX use cases and improved KPIs, ongoing consumer frustrations indicate that brands still need to optimize AI use in CX. Key recommendations for brands include:

1. Make it easier for employees to work with AI: Reduce friction between human and machine elements and improve AI training for customer-facing staff.
2. Use the power of pre-emption: Leverage AI's predictive capabilities to help customers pre-empt problems and avert issues in individual interactions.
3. Trust in personalization: Enhance personalized experiences using AI, but address data privacy concerns through greater transparency to increase consumer acceptance.
4. Find new metrics: Develop new ways to measure AI's impact on KPIs as spending grows and executives seek evidence of benefits.

About the Research

The report is based on surveys of 5,000 consumers and 500 executives across seven countries, conducted in October and November 2024. Executive respondents worked across six industries and held senior roles. Consumer respondents were all over 18.

The surveys were conducted by Longitude, a Financial Times company, on behalf of Verizon.

The executive survey demographics:

• Countries: USA (70%), Australia (8%), Japan (7%), UK (6%), France (3%), Germany (3%), Netherlands (3%).
• Sectors: Healthcare (20%), Financial services (20%), Insurance (20%), Retail (20%), Consumer goods (12%), Utilities (8%).
• Roles: C-level (20%), C-1 (20%), C-2 (40%), Other (20%).

The consumer survey demographics:

• Countries: USA (40%), France (11%), Germany (11%), UK (11%), Australia (10%), Japan (10%), Netherlands (7%).
• Ages: 18-24 (14%), 25-34 (18%), 35-44 (17%), 45-54 (17%), 55-64 (17%), 65+ (16%).
• Gender: Female (51%), Male (49%).