by Dana Lee on 1st Apr 2026 in Video Interviews

In this episode of A Coffee With... Sam Shosanya, agency sales manager at Gumtree.com, joins ExchangeWire's head of content, John Still, over a flat white to chat retail and agentic AI.

Sam has been in the industry for about 5 years, with roles focused on reaching diverse audiences at scale and geocontextual targeting. He currently represents PHD and dentsu in his work at Gumtree. Tune in as he and John discuss search intent, the rise of agentic traffic, and the zero click web.

Logged-In First Party Data

Over 25 years, Gumtree has ingested a rich base of logged-in first party data. The platform reaches users who are in a buying mode, providing insights into varied intents and buying signals. It can also track users throughout their journey, enabling detailed audience identification. This proprietary and exclusive data set is similar to that of a walled garden.

Gumtree recently introduced a 'Pay & Ship' feature in specific verticals, allowing users to complete purchases directly on the platform. The goal is to expand this feature across all verticals, creating a transactional loop that keeps users engaged within their ecosystem.

Retail Landscape and AI’s Impact

Retail media is shifting its focus from scale to intent. Platforms with logged-in users who have a clear purpose will be more valuable than those with massive but passive audiences. As the open web shrinks, first party data becomes increasingly critical for effective advertising.

"The open web is shrinking, and most people that are on retail media platforms stay logged in. So again, that's just great intent. When people are on the platform, they're searching. And the way we see it, it's more so in a sense that it's not going to be all about scale. It's going to be about what people are looking to do on that platform and what that intent is." — Sam Shosanya

Sam describes a structure to identify different users by intent: the searcher (viewed a vertical multiple times), the intender (searched nine times), the enthusiast (replying to ads), and the seller. This audience segmentation allows advertisers to target users precisely based on their interest level and activity.

Agentic & the Zero Click Web

AI is changing search behaviour, moving from keywords to conversational queries via LLMs. This creates zero-click challenges for traditional publishers reliant on search referrals. Platforms for users with specific intent are less dependent on search engines, and thus better positioned to retain their audience directly.

The next wave in retail may involve AI agents that learn user preferences and make purchases on their behalf. By 2027, AI bots could outnumber humans online. Marketplaces will need to ensure their platforms are accessible and appealing to these automated agents.

All businesses will become ad tech businesses. As businesses with large user bases and data plateau in growth, they inevitably turn to advertising to monetise. AI companies like OpenAI may leverage their vast user interaction data to enter ad tech, creating new opportunities and challenges for the industry.

Agentic AICommerce MediaFirst-Party DataRetail Media