By now, virtually every online retailer and marketplace has reckoned with the potential of commerce media. Today, there are hundreds of retail media networks worldwide, each promising advertisers new ways to promote their brands and products within stores and websites, at high-impact moments of consumer evaluation and purchase.
To fully unlock the value of retail media, the industry must take a page from leaders like Amazon and Walmart, who are largely on-site ads businesses. They continue to use the high margins of on-site ad profits to reinvest in lower prices and expanded supply chains. Machine learning is at the heart of those profitable advertising engines, helping match shoppers with relevant ads and break down constraints to truly scalable growth.
Deliver a truly personalized shopping experience with ads
The impact of machine learning on advertising is to deliver relevant ads and drive ad performance. Many legacy ad platforms run auctions based purely on cost per click, resulting in irrelevant ads to the consumer and capped ad revenue for the retail media business. The value of ad impressions is not maximized, and advertiser budgets are under-utilized.
Instead, retailers must ensure that their commerce media vendor supports real-time ML models, matching users with the most personalized product listing ads by normalizing bids with real time click and conversion rate predictions as the user is navigating through the site — whether that's within a search result, browsing a category page, or navigating online check-out. Higher click-through rates and conversions drive faster budget utilization, unlocking more spend and ad placements across the site.
Optimize retail campaigns for advertiser outcomes
Once shoppers are matched with the best possible products, retailers can next bring that relevance to how they optimize ad auctions over time. Rather than awarding placements solely to the highest CPC bidder, predictive models can calculate "effective" CPMs based on the likelihood of a given user driving the advertiser's intended outcome.
More sophisticated ad optimization has the added benefit of improving product discovery across the site. Better user relevance means greater ad efficiency, or a higher number of clicks across available impressions. This effectively creates more ad supply that can be monetized across a larger base of advertisers within the commerce media platform.
In this way, retailers and marketplaces can diversify and scale their advertiser base and balance their overall business. When put together, long-tail dollars can rival and even surpass big brand budgets, just like Amazon has built a significant fraction of its ads business with over two million advertisers.
Automate seller activation to scale the ad network
Key to growing any retail ads business is the ability to rapidly turn sellers into advertisers. Ideally, a commerce media platform can integrate into seller portals or other native dashboards for fast activation and streamlined campaign and order management. Built-in automation and ML must be capable of onboarding tens of thousands of advertisers in weeks or even days. This scalability ensures the ads business is healthy and growing its base of advertisers at all times.
As digital commerce grows, we're also seeing a convergence towards marketplace strategies. Major retailers such as Target and Nordstrom are crossing the chasm to turn their brand websites into broader marketplaces for vetted third-party sellers. This approach carries less direct inventory risk, and supports an "infinite aisle" experience for shoppers, keeping them on-site and boosting order frequency and value.
But successful execution requires rapidly activating sellers at scale and using ML to match potentially millions of items with the right shoppers at the right time.
Make ML a key capability within the ad stack
Retail media is not like other commerce categories. Given its higher margins, specifically around on-site ad placements, and the virtuous cycle of value creation for sellers and consumers, it has the power to transform the larger business. Retail ad profits can fuel customer acquisition, boost retention rates through lower prices, extend merchandise and supply chain offerings, and more.
But to maximize the promise of on-site media, retailers need to deliver long-term value for advertisers. That means a strong user experience and high performing audiences. A real-time ML engine can power true personalization, while the right commerce media platforms are built with automation for scaling activation and real-time decisioning.
Without this technological edge, any ads business will be limited by stagnant budgets and low product diversity, which means less relevance for consumers. It's the difference between a stubborn ceiling that continually hinders growth, and the uncapped potential of a network that can learn and grow at light speed.