{"id":34582,"date":"2026-09-29T14:48:00","date_gmt":"2026-09-29T12:48:00","guid":{"rendered":"https:\/\/askme.it\/insights\/nps-doesnt-measure-automated-customer-service-and-it-never-did\/"},"modified":"2026-03-26T12:23:39","modified_gmt":"2026-03-26T11:23:39","slug":"nps-doesnt-measure-automated-customer-service-and-it-never-did","status":"publish","type":"insights","link":"https:\/\/askme.it\/en\/insights\/nps-doesnt-measure-automated-customer-service-and-it-never-did\/","title":{"rendered":"NPS doesn’t measure automated customer service. And it never did"},"content":{"rendered":"
\n

The Net Promoter Score asks one single thing: how likely are you to recommend this company to a friend or colleague? The answer depends on everything: the product, the price, the purchase experience, the after-sales service, the brand communication. When a customer gives a 6 instead of a 9, nobody knows for certain whether the problem lies in product quality, delivery times, or the last customer service interaction. Customer service is left with a number that doesn’t tell them where to act.<\/p>\n

With AI automation, this problem becomes even more complex.<\/p>\n<\/section>\n

\n

Gartner’s prediction: NPS doesn’t work for customer service<\/h2>\n

Gartner had predicted that over 75% of organizations would abandon Net Promoter Score as a measure of customer service and support success by 2025. The reason identified by Gartner is structural: the broad language used in NPS captures the customer’s intention based on their assessment of much more than just customer service. This makes it difficult to identify which actions the service team should take to improve performance.<\/p>\n

Gartner describes NPS as one of the most widely used KPIs globally but also as the one that systematically fails to produce actionable insights for customer service leaders. The prediction about NPS abandonment reflects a trend already underway in the most advanced customer service teams, even though C-suite support for the metric remains strong: 58% of customer service leaders are required by their leadership to measure NPS, and CEOs frequently use it in investor communications.<\/p>\n<\/section>\n

\n

With AI, the problem multiplies<\/h2>\n

In a customer service system where most interactions are handled by AI agents, NPS captures even less of the service’s contribution to the overall customer experience. If a customer interacts with an AI assistant to resolve a technical issue, then gets transferred to a human agent for a more complex matter, and then receives an automatically generated follow-up email, the subsequent NPS reflects the sum of all these experiences, without distinguishing which part worked and which didn’t.<\/p>\n

Gartner identifies the need to measure customer experience at the individual interaction level as the starting point for actionable metrics. In a hybrid human-AI system, this means capturing specific feedback for each type of interaction: automated, assisted, and fully human. The satisfaction drivers are different in all three cases, and improvement actions require separate diagnoses.<\/p>\n<\/section>\n

\n

The metrics that work: CSAT, CES, and VES<\/h2>\n

Gartner recommends evaluating customer service performance at the interaction level through three metrics as alternatives to NPS. The first is Customer Satisfaction Score (CSAT): it measures satisfaction specific to the interaction just completed, with a direct question about the quality of the experience. It’s actionable because it’s linked to a precise moment and a specific channel.<\/p>\n

The second is Customer Effort Score (CES): it measures how much effort the customer had to exert to resolve their issue. In a system with AI, CES is particularly relevant because the primary goal of automation is to reduce friction, and CES measures it directly. Gartner identifies reducing customer effort as one of the priority objectives for customer service leaders in 2026, alongside first contact resolution and delivering greater value.<\/p>\n

The third is Value Enhancement Score (VES): it measures whether the interaction helped the customer get more value from the product or service. This metric is particularly relevant in the context of transforming customer service from a cost center to a business driver, which Gartner identifies as one of the main trends shaping the future of the function by 2028.<\/p>\n<\/section>\n

\n

91% of leaders must demonstrate impact on satisfaction<\/h2>\n

In a survey conducted between September and October 2025, Gartner found that 91% of customer service leaders are under pressure from senior management to implement AI not just for efficiency but to directly improve customer satisfaction. Improving customer satisfaction is the top priority for 2026, followed by operational efficiency and self-service success.<\/p>\n

This pressure demands metrics that effectively demonstrate the link between AI and customer satisfaction. CSAT measured before and after introducing an AI assistant on a specific channel shows the impact directly. CES measured for automated interactions versus escalated ones shows whether the system is actually reducing friction for customers interacting with AI. These metrics speak the language of customer experience, not just operational efficiency.<\/p>\n<\/section>\n

\n

AI as a feedback analysis tool<\/h2>\n

The irony of the metrics debate in AI customer service is that AI itself can solve part of the problem. Gartner identifies real-time sentiment analysis during interactions as one of the high-value AI use cases in customer service: tools that read the tone and content of conversations to understand how the customer feels without waiting for them to fill out a survey.<\/p>\n

This capability transforms satisfaction measurement from a point-in-time event to a continuous process. Instead of asking the 1-2% of customers who respond to surveys how it went, an AI system can analyze 100% of interactions and produce insights on experience quality in real time. Customer service leaders who use this capability have access to a level of granularity that no traditional survey system can match, and they can identify problems before they become visible trends in aggregate metrics.<\/p>\n<\/section>\n

\n

The starting point<\/h2>\n

Gartner is direct about the correct sequence: first build the internal case to eliminate NPS from post-interaction customer service surveys, then replace it with interaction-level metrics like CSAT, CES, and VES that produce actionable insights. If management still requires NPS measurement, do it in a way that satisfies expectations without overvaluing its importance for operational customer service decisions.<\/p>\n

In a context where AI handles a growing volume of interactions, the metrics driving improvement must be granular enough to distinguish between what works in automated self-service, what works in escalation, and what works in fully human interactions. NPS doesn’t make this distinction. The alternative metrics Gartner recommends do.<\/p>\n<\/section>\n","protected":false},"excerpt":{"rendered":"

Gartner had predicted that over 75% of organizations would abandon Net Promoter Score as a customer service metric by 2025. With AI handling most interactions, the NPS problem has become even more acute. Here’s what to measure instead.<\/p>\n","protected":false},"featured_media":34584,"menu_order":0,"template":"","insights_category":[549],"insights_tags":[633,687,689,759,791],"class_list":["post-34582","insights","type-insights","status-publish","has-post-thumbnail","hentry","insights_category-ai-and-customer-service","insights_tags-automation","insights_tags-customer-experience","insights_tags-customer-service-en","insights_tags-kpi-en","insights_tags-nps-en"],"acf":[],"_links":{"self":[{"href":"https:\/\/askme.it\/en\/wp-json\/wp\/v2\/insights\/34582","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/askme.it\/en\/wp-json\/wp\/v2\/insights"}],"about":[{"href":"https:\/\/askme.it\/en\/wp-json\/wp\/v2\/types\/insights"}],"version-history":[{"count":1,"href":"https:\/\/askme.it\/en\/wp-json\/wp\/v2\/insights\/34582\/revisions"}],"predecessor-version":[{"id":34583,"href":"https:\/\/askme.it\/en\/wp-json\/wp\/v2\/insights\/34582\/revisions\/34583"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/askme.it\/en\/wp-json\/wp\/v2\/media\/34584"}],"wp:attachment":[{"href":"https:\/\/askme.it\/en\/wp-json\/wp\/v2\/media?parent=34582"}],"wp:term":[{"taxonomy":"insights_category","embeddable":true,"href":"https:\/\/askme.it\/en\/wp-json\/wp\/v2\/insights_category?post=34582"},{"taxonomy":"insights_tags","embeddable":true,"href":"https:\/\/askme.it\/en\/wp-json\/wp\/v2\/insights_tags?post=34582"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}