Innovation & Tech

AI shopping agents expose the limits of Net Promoter Score

AI shopping agents mediate discovery and purchases, obscuring traditional CX signals. Brands must connect agent, digital, conversational and in-person data to measure loyalty.

By Amara Osei

4 min read

How do brands measure customer experiences when AI does the shopping?
How do brands measure customer experiences when AI does the shopping?AI-generated

What's News

  • Three named AI agents: Meta Muse, OpenAI Dots and Instinct AI.
  • Late 1990s: Digital disintermediation helped Amazon and Walmart gain ground on traditional retailers.
  • NPS proxies future business; behavioral loyalty measures track engagement recency, interaction quality and spending consistency.
  • Six agent tasks: internet research, calls, comparison shopping, reservations, deliveries and customer support.
  • Four interaction types to connect: agentic, digital, conversational and in-person.

Three AI shopping agents—Meta Muse, OpenAI Dots and Instinct AI—are set to mediate how consumers discover products and make purchases, according to Fast Company’s analysis. Shoppers would tell a digital proxy what they want rather than navigate websites and mobile apps. The analysis says: “You tell it what you want, and it does the work.”

The shift would upend traditional approaches to marketing, personalization, e-commerce and customer experience, the analysis says. It also exposes gaps after an agent contacts a company. Brands may not know why it chose Store A instead of Store B, whether it will explain the decision, or how it traverses digital assets.

Brands also need to know how much information an agent shares during those interactions, according to the source. The core problem is visibility. Once agents handle discovery, buying and support, the data that reveals whether an interaction produced the desired outcome becomes harder to see.

How did customer experience change before agents?

During the late 1990s dotcom era, companies moving from physical stores to e-commerce scrambled to develop online CX strategies, the analysis says. Search engine optimization, cookies and other technical innovations helped them infer consumer intent, journeys and experiences. The source notes that brands still try to understand those signals.

Digital disintermediation gave Amazon and Walmart an edge over department stores, specialty retailers and shopping malls that failed to adapt, according to the analysis. Smartphones later shifted consumers onto another channel. Successful organizations adopted unified customer profiles across devices, geolocation and tracking, plus more precise personalization.

Agentic commerce could drive another CX revolution across marketing, sales and service, the source says. Consumers will spend less time and effort shopping, buying and interacting with brands. The analysis says brands will need new tools to establish satisfaction, loyalty and brand equity.

The traditional omnichannel journey—reviews, online shopping, chatbots, voice bots and frontline employees—will increasingly run through the agent, according to the analysis. The customer will direct less of the path, while the proxy handles more of the work.

What can a shopping agent do?

Personal agents use the same underlying technology that drives large language models such as Gemini and Claude, according to the source. They go beyond answering questions or recommending products. The agent interviews users about preferences, remembers past experiences and then takes the initiative.

It can scour the internet, make calls, comparison shop, reserve items, order deliveries and call or chat with customer support, the analysis says. The agent then manages several steps at once, while the consumer supplies the goal. That shift puts more of the journey under the agent’s control.

Why does Net Promoter Score miss part of loyalty?

The customer-experience business historically relied on surveys and scores such as Net Promoter Score, according to the analysis. NPS acts as a proxy for whether a customer will do business with a brand again. It has always been an indirect measure, even when customers interact directly.

Behavioral measures offer a sharper view, the source says. They track how recently a customer engaged, the quality of that interaction and how consistently the customer spent with the company. Those indicators can reveal loyalty more directly than a survey score.

Neither approach tells the whole story when an agent mediates the interaction, according to the analysis. Brands still need to distinguish a completed transaction from a customer whose underlying needs were met. Without connecting signals across the relationship, they risk losing sight of the customer overall.

How should brands adapt?

Companies will need to optimize their business models for agent-mediated commerce without sacrificing CX quality or efficiency, the analysis says. One practical step is incentive design. Today, a customer who provides a rewards number can earn rewards; customer agents should have comparable reasons to disclose preferences and details.

Brands should also build corporate agents that communicate richly and honestly with customer agents, according to the source. The recommendation rests on information exchange. The more context the agents share, the better they can make informed, relevant decisions for the shopper.

The source’s operating principle is: “The key is to listen everywhere.” That means treating agentic commerce as a contextual channel rather than an isolated touchpoint. Brands should connect agentic interactions with digital, conversational and in-person activity.

Together, those signals can form journeys and outcomes, according to the analysis. They can expose insights, interests and motivations otherwise hidden inside what the source calls “the agentic black box.” Brands would then assess relationships across channels rather than infer satisfaction from a transaction alone.

Will agentic commerce end customer-experience control?

Brands may view agentic commerce as a threat to their control of customer experience, according to the analysis. The source compares the moment with music’s move from physical discs to platforms such as iTunes and Spotify. Some record labels resisted surrendering their physical relationship with fans before embracing digital media.

The analysis argues that brands can turn the shift into an insight engine rather than face an existential threat. Those that use agent channels to build relationships with new and existing customers can gain a clearer view of what drives them.

For brands, the test is whether they can connect agent signals with digital, conversational and in-person outcomes before mediated shopping makes customer behavior less visible. That connection is the analysis’s route from a black box to measurable loyalty.

Source: Fast Company

Share this article:

More from Amara Osei

Amara Osei

Show full bio

Senior reporter covering companies, deals and strategy in entrepreneurship, startups and venture funding at Business Bearings.

2,422 articles

Related articles

« Previous article