In this episode of , Kevel CEO James Avery sits down with Matt Nichols, partner at Commerce Ventures, to explore the transformation of retail media, the impact of AI on shopping behavior, and the supply chain models reshaping commerce.
Venture capital is pouring into retail media infrastructure as the market reaches a tipping point.
In this episode of Unlocking Retail Media, Kevel CEO James Avery, sat down with Matt Nichols — partner at Commerce Ventures, one of the most active investors in the retail and retail media ecosystem. Drawing on his 20 years in venture capital and his experience running a direct-to-consumer jewelry business that was later acquired by Berkshire Hathaway, Nichols examines the evolution of retail media since 2019, the role of AI in transforming shopping experiences, and the supply-chain innovations reshaping how retailers compete.
Here are the five biggest takeaways from their conversation.
"Retail media has just gotten so big that, I think, for the largest players, you can't afford to fully outsource it."
— Matt Nichols, Commerce Ventures
When Nichols first reached out to Avery in December 2019, the retail media landscape looked very different. At that time, retailers essentially had two choices: build everything from scratch like Amazon, or completely outsource to platforms like HookLogic or CitrusAd.
Since then, the sheer scale of the retail media opportunity has transformed. After watching Amazon and Instacart generate massive revenue from retail media, executives at low-margin retailers can no longer treat it as a secondary initiative. Nichols observed that, although the pressure to own and customize retail media offerings has intensified, building everything in-house still remains unrealistic for most.
A middle ground — where retailers maintain control and customization while relying on specialized tech providers — has become the dominant model for building differentiated retail media networks.
Artificial intelligence isn’t just another efficiency tool — it’s fundamentally changing how consumers search and how retailers need to think about advertising.
“I think the biggest change with AI is your ability to utilize large language models to figure out all of these associations between products that a human probably could never have figured out,” Nichols mused. Historically, retail media has depended on keyword-based targeting. Retailers had to anticipate every characteristic of a product and tag it accordingly. When a consumer searched, those tags determined what would appear. While this system worked, it was limited by our ability to predict every potential association.
AI upends this model entirely. Large language models can identify relationships between products based on patterns found across the internet — associations too complex and tedious to tag manually. As a result, searches can deliver relevant results based on context and intent, not just keyword matches.
"Searches that consumers do on the site deliver what they're fundamentally looking for based on all these connections that live out on the internet and not just the ones that you've been able to capture in tags," Nichols explained.
For retail media, this shift brings both opportunity and disruption. The opportunity lies in improved consumer experiences and higher conversion rates. The disruption is that traditional keyword-bidding strategies must evolve. Brands will need to focus less on specific search terms and more on outcomes — for instance, selling more ketchup, whether shoppers search for “ketchup”, “hot dog toppings”, or “picnic items.”
Many industry observers believe AI platforms will only capture low-consideration purchases — toilet paper, basics, commodity items. Nichols disagrees, predicting a world with text- and query-based shopping experiences via ChatGPT.
“Maybe it’s similar to what we have today,” he said. “But the output actually inside of the ChatGPT experience probably looks more like a Shopify site.”
The key difference between AI platforms and past attempts at shopping innovation (like Google or Instagram Shopping) is interactivity. In earlier models, consumers had to be shown exactly the right product at exactly the right moment, with no room to ask additional questions. If more information was needed, they’d have to leave the platform entirely.
AI platforms eliminate that friction, Nichols explained. Consumers can ask follow-up questions, compare options, inquire about specifications, and complete purchases without ever leaving the environment. This dramatically increases the likelihood of conversion across a broad range of product categories.
If the shopping experience inside ChatGPT begins to resemble a full e-commerce site, retail media fits naturally within that environment. Sponsored products, promoted listings, and paid placements become part of the AI-driven shopping journey rather than an awkward add-on.
For years, retail innovations of the past focused on pushing products as close to consumers as possible through massive fulfillment networks and same-day delivery. Nichols believes the next wave of innovation is headed in the opposite direction: keeping products close to manufacturers.
He points to retailers like Shein, Temu, and Quince as proving the value of lean inventory models. By producing goods on demand or in small batches, they reduce excess inventory, minimize markdowns, and adapt quickly to changing customer tastes.
This approach, Nichols observes, is becoming more critical as consumer preferences shift faster than traditional inventory planning can keep up. Retailers relying on old models risk constant mismatches — too much stock leads to discounts, too little leads to missed revenue.
For retail media, lean supply chains create new advantages. Marketplaces can broaden assortment without inventory risk, and brands can test new products or ad strategies with minimal upfront cost. The agility of lean supply chains aligns naturally with the real-time optimization that retail media demands.
Retailers aren’t slowing down on launching marketplaces — they’re accelerating. According to Nichols, the logic is simple: when a retailer already has consumer traffic, expanding into third-party products is an untapped revenue opportunity.
"I think if you're big enough to drive the audience to your site, ultimately someone came to your site 'cause they wanna purchase," Nichols said. "If you can augment your first party products with a relevant set of third party products that can live up to your brand reputation…there's almost no downside."
Most retail sites suffer from low conversion because shoppers can’t find the right product. Marketplaces fix this by expanding assortment without requiring retailers to hold additional inventory.
“There’s a reason it’s been such a powerhouse for growth for Amazon,” Nichols explained. “You don’t take the inventory, you’re relying on someone else and you can scale really quickly.”
For retail media, marketplaces are even more valuable. More third-party products mean more competition, more sponsored placements, and more advertising demand. "Retail media shines when it's a third party product on some retailer site,” Nichols noted. “And it's all about you as that brand pushes your product to the top.”
As marketplace technology becomes easier to implement, more retailers will adopt the model — increasing ad inventory, intensifying bidding among brands, and ultimately boosting retail media revenue for retailers who own the platform.
The retail media landscape has matured significantly since 2019. What started as a land grab has become a more strategic battle over who owns the technology, the data, and ultimately the advertiser relationship.
Nichols is optimistic about the opportunity but realistic about the hurdles. AI is reshaping how people shop, potentially pulling transactions off retailer sites. Supply chains are shifting toward lean, demand-driven models. Marketplaces are growing quickly, adding both scale and operational complexity.
For retailers, succeeding in this next phase means owning the most critical parts of their retail media stack while partnering where specialized technology is needed. It means preparing for AI-driven shopping that may happen off-site, while strengthening on-site AI capabilities to keep traffic and conversions. And it means using marketplaces to expand assortment without compromising brand standards or advertiser trust.
"The pressure they're putting on their teams has only increased to accelerate the efforts in these areas," Nichols said of retail leaders. The opportunity is enormous — but realizing it requires focused investment in the right infrastructure.
For more insights like these, tune in to the full episode of Unlocking Retail Media, the podcast where Kevel CEO James Avery sits down with industry leaders and innovators shaping the future of retail advertising.
Listen to this episode with Matt Nichols here.
