Azoma secures dunnhumby ventures backing for agentic commerce platform

Azoma, the London-based Agentic Commerce Optimisation platform, has announced a strategic investment from dunnhumby ventures, the venture and innovation arm of the retail data science firm. The deal is intended to accelerate Azoma's product roadmap and deepen integration with dunnhumby's retailer client base, as AI-powered shopping agents become a mainstream route to purchase for consumers.

As part of the arrangement, Leo Nagdas, Head of dunnhumby ventures, has joined Azoma's board as an adviser. Nagdas founded dunnhumby's Retail Innovation Network, an open innovation programme connecting retail technology leaders, and brings with him access to dunnhumby's proprietary data science assets and its global network of retailer and supplier partners. The investment amount was not disclosed.

What the data shows

Azoma published findings from its analysis of tens of millions of AI-generated shopping responses in the second quarter of 2026. The data revealed that earned and social media sources account for 86.5% of the citations driving Alexa for Shopping's product recommendations, and 76% of those underpinning Walmart's Sparky assistant. ChatGPT draws more heavily on retailer-owned content, at 37.1% of citations. The implication is significant: brands have minimal direct control over the sources that most AI agents use when recommending products to shoppers.

"AI shopping assistants are rewriting the path to purchase, and today retailers and CPGs have no reliable way to connect what an agent recommends to a sale," said Nagdas. "Azoma is a leader in the emerging agentic commerce space, and the team combines real depth in AI with a practical grasp of how retail actually works."

Azoma's platform covers measurement and optimisation across retailer-specific agents, including Alexa for Shopping, Walmart Sparky and AI Lazzie, as well as general-purpose assistants such as ChatGPT, Gemini, Claude, Meta AI and Siri. The company frames its discipline as Agentic Commerce Optimisation, positioning it as a commerce-specific evolution of the broader Generative Engine Optimisation and Answer Engine Optimisation categories that have emerged around AI search.

Market context

The investment lands at a moment when the retail and consumer packaged goods sectors are grappling with the commercial implications of AI-intermediated discovery. Traditional search-engine optimisation and retail media strategies are built around a consumer who browses and clicks; agentic commerce assumes a consumer who delegates the choice to an AI assistant entirely. That shift puts pressure on brands to influence upstream signals, including third-party editorial, social content and product-data feeds, rather than their own web pages alone.

Azoma's named client roster, including L'Oréal, Unilever, Mars, Beiersdorf and Reckitt, suggests the platform has traction with large, multi-brand CPG groups that operate across many categories and cannot afford to be invisible to agent-driven recommendation engines. Competitors in the adjacent GEO and AEO space are multiplying rapidly, but few have oriented explicitly around the point of purchase rather than general AI-answer visibility.

CEO Max Sinclair described agentic commerce as "a $9 trillion opportunity," a projection that carries the forward-looking uncertainty typical of emerging-category sizing. dunnhumby's involvement is strategically significant: the firm sits on one of the richest retail loyalty and transaction datasets in Europe through its Tesco Clubcard heritage, and that data could materially strengthen Azoma's ability to close the loop between agent recommendations and measurable sales outcomes.

Regulatory read-across

Brands and retailers deploying AI shopping agents in the EU will face transparency obligations under the EU AI Act, particularly around how automated recommendations are disclosed to consumers. The UK's Competition and Markets Authority has also signalled interest in AI-driven retail personalisation and the potential for algorithmic systems to disadvantage smaller brands or new entrants. Azoma's measurement layer could serve a compliance function as well as a commercial one, providing an audit trail for how products are surfaced by third-party agents, though the company has not yet articulated a regulatory-readiness use case publicly.