Why retail media supply-side platforms alone won’t save your ad business

Retail media is the fastest-growing sector of digital advertising, with ad spend volumes rising more than  20 percent year over  year. As retail media networks (RMNs) gain prominence, misconceptions are emerging that could hinder the potential for long-term success. Two widespread myths in particular are creating confusion for retailers looking to scale their ad businesses:

1. “There are too many RMNs,”  with estimates suggesting over 200 globally, raising concerns about fragmentation and the challenge of managing advertiser relationships.
2. “Retail Media platforms should bring demand,” implying that growth depends on third-party ad tech partners to attract advertisers.

Both of these beliefs overlook the true value and unique position of RMNs. Rather than relying on third-party solutions, RMNs can take control of their own advertising ecosystems — leveraging advanced technology and machine learning to unlock growth, enhance relevance, and build stronger direct relationships with their advertisers.

One popular but misguided approach is the creation of a retail media supply-side platform (SSP), where RMNs manage multiple demand sources to boost revenue. Let’s explore what this looks like and why it doesn’t deliver the scale or performance RMNs truly need.

What a ‘retail media SSP’ looks like

A common solution proposed for “bringing demand” is the creation of a retail media supply-side platform (SSP), which promises to manage multiple demand sources and boost revenue. At first glance, this approach seems appealing, but let’s break down how it actually works — and why it doesn’t fully deliver for RMNs looking to scale.

Here’s how a retail media SSP functions:

For the RMN/Publisher: 

1. Choosing an SSP-enabled platform: The RMN selects an onsite ads platform with SSP capabilities. The focus here is on sourcing demand, often from third-party ad networks or directly from demand-side platforms (DSPs).
   
   
2. Prioritizing bids: The RMN sets the logic for bid prioritization—using methods like header bidding or waterfall logic to accept bids from different sources.
   
   
3. Establishing deal types and advertiser tiers: RMNs choose from private or preferred deals (e.g., programmatic guaranteed or private marketplaces) and open auctions for real-time bidding (RTB). RMNs can also set preferential tiers for top endemic advertisers (those already selling products through the retailer) or open the platform to non-endemic advertisers.

For the advertiser:

1. Accessing inventory via DSPs: Advertisers use their preferred DSP to bid across available inventory from multiple RMNs. They can participate in real-time auctions or submit campaign bids manually (e.g., CPC bids for sponsored products through tools like Pacvue).
   
   
2. Bid execution and delivery: Once the bid is submitted, it is routed through an ad exchange or directly to the RMN, which serves the ad based on the pre-set rules.
   
   
3. Performance tracking and optimization: DSPs report on campaign performance and can offer recommendations for improving bidding strategies based on past data.

This focus on “bringing demand” is why many retail media networks turn to platforms like GAM or Criteo, as demand is often seen as their primary value proposition.. The problem: this won’t meaningfully grow your business. 

Why a retail media SSP alone won’t scale 

While SSPs offer some benefits, they come with several drawbacks that prevent RMNs from truly scaling. Beyond disintermediating retailers from their brand relationships and adding demand-side fees that hinder performance, here are additional reasons SSPs fall short:

• No improvement in existing inventory performance: SSPs may provide multiple demand sources, but they don’t address the underlying performance issues. Without machine learning (ML)-based relevance and real-time data on user activity, RMNs can't deliver the 1:1 personalization needed to optimize ad relevance.
  
  
• Advertisers carry the media risk: Many retail DSPs aren’t true DSPs; they function more as campaign management tools (e.g., CPC bidding). This means advertisers take on performance risk without outcome-based bidding models like target ROAS, which limits scalability.
  
  
• Risk commoditized inventory:  In programmatic RTB setups, advertisers optimize across multiple inventory sources, including your competitors. By losing control of probabilistic auctions and ML-based efficiency, RMNs risk commoditizing their ad inventory.
  
  
• Demand without performance leads to low budget utilization and fill rates: Advertisers want to spend! But if the ads don’t perform, budgets go underutilized, leading to low fill rates. Worse, poor performance frustrates advertisers, causing them to either reduce spend or churn entirely.
  
  
• Over-indexing strategy for the wrong advertising partners: The majority of RMN spend often comes from endemic advertisers. For instance, if a large portion of revenue is driven by endemic-sponsored products — say 70% — with 30% from display and a small share (e.g., 10%) from non-endemic sources, RMNs may find themselves shaping strategies around a smaller subset of demand. This could limit their ability to explore new growth opportunities.

The limitations of retail media SSPs are clear — they fail to improve inventory performance, shift the risk onto advertisers, and lead to underutilized budgets. To truly scale, RMNs must go beyond simply sourcing demand. Instead, they need to harness their unique strengths, like first-party data and real-time user behavior, to drive relevance, performance, and growth.

This is where machine learning and purpose-built ad tech solutions come into play.

Retail media networks hold the cards 

It’s important to understand, SSPs solved a problem in digital advertising at a much larger scale than retail media by enabling real-time bidding across hundreds of thousands of publishers. The dynamics in retail media are unique. For most advertisers, only 10 to 20 RMNs are likely to be of real significance.

Retail media networks aren’t just another publisher — in fact, they are customers of endemic advertisers. These networks should be seen as mini walled gardens, more aligned with the models of Amazon, Meta, and Google than with open internet publishers. RMNs have distinct advantages: inventory close to the point of transaction, the best first-party data in advertising, and built-in relationships with brands already selling through their platforms.

To fully capitalize on these strengths, RMNs should follow the playbook used by Amazon, Meta, and Google, who dominate 65% of digital marketing by focusing on:

• Using the best tech and machine learning to drive relevance and performance
• Scaling demand with ease of use and de-risking media investment
• Expanding ad inventory without sacrificing organic metrics

A retail media SSP alone won’t bridge these gaps.  What RMNs truly need is a purpose-built, machine learning-driven partner. like Moloco that solves all three. By maximizing the predictive power of first-party data, Moloco empowers retailers to take control of their ad ecosystem and unlock sustainable growth.