Dr. Luke Soon once again collaborated with me on this post.
Developed in 2003, NPS (Net Promoter Score) is a metric that has been widely adopted across the globe. It’s creators, Bain & Company, claimed that it measures customer loyalty and is a predictor of future growth. Its methodology is relatively simple: it asks respondents how likely they are to recommend a product, service, or organisation on a scale of 0 to 10. Based on their answers, respondents are categorised as Promoters (people who gave scores of 9 or 10), Passives (7’s and 8’s), or Detractors (0-6). This has provided a quick and intuitive way for businesses to gauge customer sentiment. However, as industries and technologies have evolved, the limitations of NPS have become more apparent, especially in complex customer ecosystems.
In this post, we’ll explore the shortcomings of NPS, dive into real-world examples where it falls short, and propose how AI can help transform and improve the way we measure customer loyalty.
NPS reduces a respondent’s entire experience with a brand to a single number. While simplicity is part of its appeal and executives love seeing another metric on their dashboard, it doesn’t capture the nuanced and varied reasons why a customer might feel positively or negatively about that brand.
Case Study: A large retail bank noticed consistently high NPS scores but still faced churn among long-time customers. Through qualitative interviews, the bank discovered that while customers were happy with certain aspects of their service, such as digital tools, they were frustrated by long wait times for in-branch assistance. The NPS survey failed to capture this complexity, leading the bank to misinterpret the overall health of customer satisfaction.
Knowing whether customers are likely to recommend your brand is valuable, but NPS doesn’t explain why they would or wouldn’t. Without follow-up questions, specifically asking “why did you give that score?”, NPS alone doesn’t provide clear direction for improving customer experience.
Case Study: A leading telecommunications company used NPS to monitor customer loyalty. Despite a relatively high score, the company saw a sharp increase in customer complaints about internet connectivity issues. The NPS scores gave no clues about these issues because they didn’t specifically ask customers about their experiences with connectivity, leaving the company scrambling to address the root problem.
NPS surveys are often completed by respondents with extreme opinions — they’re either very satisfied or very dissatisfied — while people with more moderate experiences might not respond. This can lead to skewed data that doesn’t represent true customer sentiment.
Case Study: An online retailer noticed that their NPS fluctuated significantly based on the timing of their surveys. If the survey was sent immediately after a seasonal sale, scores would spike because many Promoters responded. However, after a product delay, the NPS scores dropped dramatically due to Detractors, giving an inaccurate reflection of overall customer loyalty.
Different cultures have different norms when it comes to rating satisfaction. Some cultures, such as those in Latin America, are more generous with high ratings, while others, such as in a lot of Nordic countries, may be more conservative. This makes comparing NPS across regions difficult.
Case Study: A global software company tracked NPS across different markets but found the scores in Japan consistently lower than in North America. After further investigation, they realised the cultural tendency in Japan to avoid extreme positive ratings, while American customers were more likely to give high scores. This made it difficult to compare loyalty across regions and required more sophisticated approaches to interpret the data.
One of the key limitations of NPS is that it doesn’t explain why a customer gave a particular score. AI-driven sentiment analysis can be applied to the open-ended responses that often accompany NPS surveys. Natural Language Processing (NLP) algorithms can analyse and categorise feedback, providing insight into the emotions behind the scores.
Example: A global airline incorporated AI-driven sentiment analysis into its NPS surveys. When customers provided open-ended comments, AI analysed whether the sentiments were positive, negative, or neutral and clustered common themes such as flight delays or customer service issues. This allowed the airline to pinpoint which issues were driving Detractor scores and enabled them to make targeted improvements in key areas.
AI can be used to empirically analyse a customer’s experience and predict their NPS before surveys are even conducted. By analysing customer behaviour, purchasing patterns, and past interaction data, AI can predict Net Promoter scores and flag customers who are likely to be Detractors, allowing companies to intervene early.
Example: An online streaming service used AI to build predictive models based on customer viewing patterns, subscription cancellations, and previous feedback. The model identified customers at risk of becoming Detractors based on behaviours such as watching less content or cancelling premium subscriptions. The service then deployed personalised offers to re-engage those customers, reducing churn rates significantly.
AI can go beyond identifying Detractors to performing root cause analysis. By correlating Net Promoter scores with transaction data, customer interactions, and product usage, AI can uncover hidden patterns and systemic issues that contribute to low scores.
Example: A SaaS company used AI to automatically correlate low Net Promoter scores with customer interaction logs. The AI identified that Detractors frequently contacted customer support for the same technical issue, which hadn’t been captured by the initial NPS survey. This allowed the company to fix the underlying problem, leading to improved satisfaction and higher scores.
Traditionally, NPS surveys are sent at specific points in time (for example, after a purchase). AI can enable real-time NPS feedback, gathering customer sentiment dynamically across multiple touchpoints and updating NPS as the customer journey evolves increasing organisational agility.
Example: A retail chain integrated AI into its point-of-sale system to trigger NPS surveys in real time after specific interactions, such as returning a product or completing an in-store purchase. The AI adjusted survey questions based on previous responses, personalising the experience for each customer. This real-time feedback gave the company more accurate data and enabled them to be more responsive in the way that they addressed customer concerns.
Using AI, companies can personalise their follow-ups based on Net Promoter scores. Promoters can be nurtured with loyalty programmes or referral incentives, while Detractors can be targeted with tailored offers to win back their favour. AI can automate this process, ensuring timely and relevant responses.
Example: A subscription box company used AI to automatically segment customers based on their Net Promoter scores. Promoters received referral incentives, while Detractors were offered a personal discount and a direct line to Customer Service. This AI-driven follow-up process resulted in a 15% increase in customer retention among Detractors.
While NPS remains a valuable metric for gauging customer loyalty, it falls short when used in isolation. Its simplicity can obscure deeper insights and leave organisations with only a partial understanding of their customers’ end-to-end experience. By integrating AI into NPS processes, companies can transform this legacy metric into a dynamic, predictive, and actionable tool that not only measures customer loyalty but also drives meaningful improvements in customer experience.
From sentiment analysis and predictive modelling to real-time feedback and personalised outreach, AI offers numerous ways to augment NPS, turning it into a powerful tool for both understanding and improving the customer journey.
Are you ready to transform your NPS strategy with AI? The future of customer loyalty measurement awaits.
Image courtesy of Pixabay
Luke Soon is a business transformation professional with over 25 years’ experience leading multi-year human experience-led transformations with global telcos, fintech, insurtech and automotive organisations across the globe. He was the lead partner in the acquisition and build-up of the human experience, digital and innovation practices across Asia Pacific with revenues surpassing $250 million. He helps clients activate their Purpose by monetising innovation and building new revenue streams (experience equity), starting with their why. His personal purpose is to install the primacy of humanity in the experience economy and exponential age. Connect with him on Twitter and LinkedIn.
Dr. Luke Soon once again collaborated with me on this post. Developed in 2003, NPS (Net Promoter Score) is a
....READ MORE
This post first appeared on Sprinklr's blog in June, 2024. In a world where the cost of operating a business
....READ MORE
This post is the result of a collaboration with Luke Soon. Many customer experience transformations stall because leaders can’t show how
....READ MORE
