Measuring customer loyalty in the age of AI needs a fresh take on NPS

Net Promoter Score (NPS), the trusted metric for measuring customer advocacy and loyalty, has long provided critical insights into customer experiences and organizational health. However, as AI and GenAI become integral to business operations, organizations must rethink how they leverage NPS to capture the nuanced realities of intelligent systems and services.

At NTT DATA, Inc., we have maintained a longstanding, globally harmonized approach to measuring NPS. For nearly a decade, we’ve systematically collected customer feedback through standardized questions, creating a strong foundation for informed, customer-centric decision-making.

Now, as our portfolio increasingly incorporates AI and GenAI solutions — and as we continue to deliver measurable success for our clients — we recognize the need to evolve our feedback framework. This evolution will capture the complexities that come with deploying emerging intelligent technologies.

• ALSO READ → A force for good: How AI and GenAI are reshaping our world

Why traditional NPS falls short for AI and GenAI

AI-driven services — ranging from conversational assistants to predictive analytics, sophisticated modeling, agentic solutions and more — introduce distinct challenges to measuring customer experience. Traditional NPS, anchored in a likelihood-to-recommend question, effectively captures advocacy but not the deeper dimensions of algorithmic trust. Moreover, it often fails to capture additional critical dimensions that are unique to AI, such as explainability, consistency, measurability and transparency.

GenAI, in particular, introduces specific experience factors, from occasionally unpredictable outputs (“hallucinations”) to the need for strategic alignment with organizational values, business goals and robust ethics and governance. To realize the full value of customer feedback in this new era, organizations must expand NPS measurement beyond traditional methods and embrace an expanded model — one that incorporates contextual, qualitative and continuous evaluation.

Building an effective NPS strategy for AI and GenAI

Strategic touchpoint selection 

Effective NPS measurement in AI requires careful timing of data-collection surveys across the technology lifecycle — post-interaction, post-deployment, post-adoption and at pivotal points like major feature rollouts. For this reason, NTT DATA’s Client Experience team ensures that all our client relationship and engagement feedback programs include “likelihood to recommend,” which is used to calculate NPS. This nuanced approach provides deeper insights into how users’ experiences evolve alongside the AI systems themselves.

Audience segmentation 

Capturing meaningful insights demands segmentation by user roles and use cases. Executives, developers and business users all have unique perspectives on AI’s value. The C-suite alone requires sampling among titles ranging from the CEO and the CIO to the CISO, the CDO and the CFO. Tailored segmentation allows organizations to differentiate between responses, enabling more precise and actionable strategies.

Qualitative feedback: Understanding the “why” 

To gain a deeper understanding of NPS drivers in GenAI, in particular, organizations must pair quantitative scoring with qualitative analysis. Incorporating open-ended questions, analyzed through natural language processing, provides critical insights into user trust, satisfaction and specific experience-related concerns such as latency, explainability and privacy.

While expanding customer feedback loops has traditionally required significant resources, the current AI and GenAI landscape offers a unique opportunity to capture and analyze rich qualitative feedback more efficiently than ever before, effectively using AI to justify and streamline this process.

A few best practices for tailoring NPS for AI and GenAI

Move beyond the traditional NPS question

The traditional single, direct NPS question is a straightforward metric designed to gauge customer satisfaction and loyalty by assessing customers’ willingness to endorse a product or a service. However, going beyond that to ask specific, AI-relevant follow-up questions — such as “How confident are you in the recommendations provided by your AI tool?” or “How useful was this generative model in achieving your objectives?” — yields richer and more actionable insights.

Focus on lifecycle tracking and evolution 

Regularly track NPS throughout the AI product lifecycle — from pilot projects to full-scale deployments — to identify changing user perceptions and proactively address shifts in sentiment.

Integrate feedback loops 

Successful organizations link customer feedback directly to AI model enhancements. At NTT DATA, we view NPS detractor feedback as a trigger for targeted improvement initiatives, developing robust “get to green” plans that lead to measurable outcomes and reinforce customer trust.

Challenges and cautions to NPS implementation for AI and GenAI

Organizations can face significant challenges when extending traditional NPS to AI and GenAI, technologies that are evolving quickly. Also, client maturity rates vary widely. Indexing and interpretation must hold more breadth, combined with qualitative checkpoints in addition to scoring. Common missteps include:

• An overreliance on quantitative scores without sufficient qualitative context
• Treating AI-driven experiences similarly to legacy offerings, missing critical experiential nuances
• Neglecting feedback from high-impact but lower-frequency stakeholders, such as executive sponsors or strategic account decision-makers

Real-world benchmarks and insights

NTT DATA’s NPS score has shown a consistent upward trend over the past five years — rising from the high 30s to over 60, which places us significantly above the average of 40–44 for the broader IT services industry.

We’ve extended the success we’ve achieved with our NPS program in North America to every region around the world. Globally, we focus on ensuring we hear from a strategic sampling of accounts, distributed across industries and other company demographics to represent the full business.

Read more about some of our AI and GenAI clients around the world:

• Modernizing the L’Oréal shopping experience with AI: A breakthrough innovation in the beauty industry
• A global beverage leader revolutionizes retail trade spending with GenAI: Easy access to data and insights empowers decision-makers
• Partnering with the Crown Prosecution Service to deploy GenAI: Supporting the CPS in their ambition to reduce delays across the UK justice system
• Transforming US housing provider Balfour Beatty’s work-order assurance process with GenAI: A significant boost to the efficiency and reliability of work-order reviews

Leading customer-centric AI adoption

For effective NPS implementation in the era of AI, we recommend deploying tailored measurement tools and analytics capabilities that deliver richer, more actionable insights, including:

• Customized NPS survey templates
• Dashboard frameworks for structured visualization and tracking
• Feedback rubrics to categorize, interpret and act on qualitative data

As organizations invest more in AI and GenAI solutions, adapting traditional customer-experience metrics like NPS becomes essential.

At NTT DATA, our established approach to customer measurement — combined with a commitment to ongoing innovation — positions us to help our clients unlock the full potential of intelligent technologies. By strategically adapting NPS methodologies to reflect the nuances of AI-driven services, these organizations can achieve greater transparency, strengthen customer trust and build lasting competitive advantages.