In an ever-growing agentic economy, understanding how products are purchased is becoming increasingly challenging. Therefore, brands, retailers, and technology platforms are investing in new measurement strategies to understand how AI agents are influencing shopping decisions.

New infrastructure is changing the commerce game, such as Google’s Universal Commerce Protocol and OpenAI’s Agentic Commerce Protocol. These are allowing consumers to complete the entire shopping journey, from finding a product and product comparisons to final purchase, within a single AI experience – an agentic one-stop shop for purchases. 

This is forming a new challenge in terms of measurement and growth for brands and retailers, as they need to know whether their products and services are visible in AI-driven discovery, as well as being accurately represented. Ultimately, they must understand whether that visibility is translated into traffic and sales and what can be done to improve overall performance.   

NIQ and Similarweb advance measurement for the AI shopping landscape

NIQ, in partnership with Similarweb, have announced new advancements to their agentic commerce measurement strategy, adding a measurement component to NIQ’s Commerce Intelligence offering. These updates are vital as AI agents move past product discovery and become part of the actual purchase. 

The latest capabilities are designed to connect AI-driven product discovery with retail sales data, giving brands a better view of whether visibility within AI platforms is translating into traffic, conversions, and ultimately sales.

Troy Treangen, Chief AI and Product Officer at NIQ, commented on the importance of retailers and brands understanding where and how AI is influencing their business. Treangen said NIQ clients also want to know “how quickly that influence is growing and what they should do about it.”

It appears agentic commerce measurement strategies are becoming a key tool for understanding these changes, helping businesses track how AI is influencing product discovery and identifying where that influence is growing in order to respond. 

Five key areas of measurement 

To measure agentic commerce successfully, five key areas should be considered, from consumer intent to AI-driven conversion. 

Firstly, consumer intent relates to what consumers ask AI assistants, including which questions, needs, and prompts are influencing the agentic commerce decisions. Then there is agentic shelf visibility, which looks at where products appear when AI assistants recommend options to consumers and how well a brand features compared to competitors. 

The readiness of product content is another crucial factor where brands and retailers ensure their product has the correct information and is in the right format so AI assistants can understand and recommend them easily and accurately. AI-driven traffic should also be considered, measuring the percentage of traffic to brand and retailer product pages that originates from AI platforms.

Finally, understanding how AI-influenced engagement connects to omnichannel purchasing behaviour can help businesses determine whether AI-driven discovery truly translates into real-world sales.

Emerging measurement capabilities 

There are several developments that are introducing new avenues towards improved measurement. For instance, OpenAI’s feature, built-in to ChatGPT, allows an AI agent to signal pre-purchasing stages, such as product considerations, comparisons with competitors, and recommendations. Therefore, brands can gain more visibility into the discovery and consideration stages of the customer journey, rather than only seeing the final purchase. 

Agentic identification standards are also in development, such as Mastercard’s and Visa’s “Know Your Agent” frameworks with cryptographic verification. These systems could verify when an AI agent is actively making a purchase on a user’s behalf, therefore allowing brands and retailers to separate agent-driven purchases from traditional API transactions, before measuring each separately. 

Developments in cross-channel identity resolution are also taking place, helping businesses gain significant measurement insight through the connection of an agent-mediated purchase with a known customer profile. The transaction then gives identity signals, including email and payment method, linked to existing customer records. In turn, this allows brands to track repeat purchases and the long-term value of existing customers. 

Agentic commerce measurement still in its infancy

Although AI shopping systems can currently process transactions, many do not yet provide the same detailed data about the customer journey as traditional ecommerce tracking, such as considered products and recommendations. 

The conversation around measurement in agentic commerce is getting louder and is becoming a key discussion for marketers. These questions will be central to talks around AI and marketing measurement at upcoming DMWF events, including expert insight at the latest DMWF North America event in New York. 

Advancements, such as those made by NIQ and Similarweb, are closing this gap. Measuring exactly how an AI agent influenced a customer’s decision is becoming a reality. Approaching agentic commerce requires trust and confidence and brands must develop a measurement infrastructure before scaling up. As visibility improves, marketers will have greater insight into how AI is influencing purchasing decisions, eventually helping them understand whether AI agents drove a sale and why. 

See also: Hoka turns Strava running data into digital out-of-home advertising

Marketing Tech News, powered by DMWF Ltd, delivers the latest marketing technology news, insights and industry updates.