Aggregation is the word…

It is 45 years since the film ‘Grease’ came out, but its timeless message that ‘to be truly popular you need to change how you look’ has aged impeccably. And in the world of CPG advertising, ‘Shopper Media’ - the Sandra Dee of the lower funnel - has taken this to heart.

After years of being awkwardly mocked at sleepovers, Retail Media Networks (RMNs) have finally emerged at the fairground (in tight leather trousers and smoking) and suddenly everyone is interested. While it wasn’t the first retail media estate, Amazon heralded the arrival of Retail Media Networks in their modern form. The emphasis on tailored experience through the use of customer data and integrating on-site and off-site advertising only really came about with AWS. And so for the last few years, while the wider retail media landscape has certainly expanded, the majority of the growth has come from Amazon but - with retail media capturing the zeitgeist and ‘experts’ born daily - the market is fragmenting rapidly.

For possibly the first time, retail media is no longer the weird cousin of above the line (ATL) media, but is receiving interest like never before. This fragmentation is coming from a few different factors, but is driving increased sophistication as competitor retailers push to lead the emerging sector. In the last two years in the UK alone, we have seen the launch of loyalty-card-powered programmatic advertising across Tesco, Boots, Sainsbury’s, and now Waitrose, with others looking to follow. Even more recently we’ve seen the introduction of this same loyalty-card data across on-site ecommerce, and into OTT with ITV and Channel 4. 

To anyone in retail, the benefits of this are obvious - the ability to drive penetration and frequency by addressing actual shoppers who have purchased your brand before. And to the ATL agencies who are newly arrived in this space, loyalty-card data provides a fantastic opportunity for quality data with the deprecation of third-party cookies on the horizon. RMNs also offer the closed-loop attribution and commercial return which has historically been the bane of the upper funnel. And so naturally more players are entering this space, from both retailer and agency sides.

This increased engagement is leading to pressure externally and internally for more sophisticated buying mechanisms. Externally, there’s an increased desire for consistency and efficiency from the retailers with wider media buying. Internally, the sudden increase in interest presents scale challenges for the retailers. Each member of staff within these retail media agencies is another head which needs justifying, and the more technical these managed campaigns are, the more specialist heads are required.     

This sets the scene for what’s likely to be the next big phase for retail media: MetaDSPs and aggregators. In short, a MetaDSP is a platform that sits on top of multiple Demand-Side Platforms to allow the buying of media across all of these DSPs from one central location. Alongside these, and slightly differently, you have aggregators which provide a unified front-end for a collection of ad inventory providers, from which you can view and optimise campaigns. The broad theme here is one of consolidation. With the arrival of more players in the market, we are in a gold rush for retail media and the emphasis is currently on which of these players can develop the most persuasive proposition, or corner the market by securing enough long-term commitments that make them too big to ignore. 

There are a couple of key reasons these MetaDSPs and aggregators are likely to play so prominently in the future: resource, scale, and efficiency.The resource question is an important one. Brands, particularly larger brands with in-house teams or sizable media agency commitments don’t always want managed services. More commonly, they want to be able to compare like-for-like with their wider buy, and they want to be able to reduce the manpower required by collating their spend into one place.

Conversely, for the smaller brands, it’s impossible to keep abreast of the developments of these new technologies, or to manage them all themselves, so individual self-serve capabilities in each retailer is not necessarily a viable solution for them. For a Brand Manager at a small to mid-sized FMCG, that would involve working with approximately six retailer agencies, at least four separate UIs for self-serve sponsored products, and potentially three (and climbing) separate self-serve programmatic off-site propositions. Aggregators and Meta-DSPs are likely to offer a solution here. By pooling the spend, minimising friction between individual retailer campaigns, and allowing the optimisation of media through a central access point, they offer a benefit for brands of all-sizes.

The scale challenge is a tricky one. For the vast majority of the market - especially within FMCG - no one outside the very biggest US markets has enough audience scale to provide a meaningful line on a media buyer’s budget. This issue is magnified the moment audience targeting or category splits are applied to this audience, and it makes it very difficult to deliver a six or seven figure campaign budget without an extremely high frequency, very broad targeting, or a long campaign window.  MetaDSPs will aim to bundle spend across retailers to create meaningfully-sized audiences, without requiring multiple log-ins and optimising across different platforms. This should help to drive scale and overcome the challenges presented by siloes. But there are likely to be challenges. A truly cross-retail centralised aggregator is unlikely any time soon, because a true MetaDSP or aggregator for retail would likely make it very difficult to avoid driving sales in competitors, or even pushing existing customers into the arms of another.

The retailers also want to avoid fully commodifying assets before they’ve learnt their true worth as other targeting methods are being deprecated.This means that it’s more likely the aggregators will continue to push to accumulate clients to the extent that they become a de facto cross-retail one-stop shop. The challenge here for brands is that most spend in these areas sits outside of spend commitments with the retailer, which starts to carve out the shinier channels for incremental investment rather than allowing them to strengthen joint plans. 

There are also inconsistencies between the aggregators and DSPs in how they measure and attribute sales, and crucially incrementality remains very difficult to prove through the methodologies being used. When exposed audiences aren’t controlled for, it’s tricky to prove existing shoppers weren’t already going to buy your product. And not all of the players are doing this. This risks encouraging brands to optimise campaigns towards ROAS measured on existing sales, which longer-term, suffocates growth.   

Perhaps most significantly, the focus of these aggregators and DSPs - so far - is entirely on the programmatic and ecommerce sides of the business. While more channels within the retail media estates are moving into auction models, the vast majority is still bought directly. These channels were historically ignored by media agencies, and to this day the majority of media buyers would struggle to identify an aisle fin in a police line-up. This becomes an issue when brands use the arrival of MetaDSPs and aggregators to drive efficiencies more aggressively and consolidate their spend, reducing their agency roster to a single big player and buying through aggregators and MetaDSPs instead. That seems like an easy win, but the network agencies have not historically had the inclination to work across retail media. The expertise is niche, barkers aren’t that sexy, and profitability comes through buying efficiency. This naturally incentivises neglecting the in-store and ecommerce channels which are not available through programmatic means, and ignoring the 70-80% of retail sales which are still going through bricks and mortar. 

This would risk us heading back to the dark days of retail media. Where Joint Media Plans and in-store media are left for a stretched brand manager to coordinate, while the more ‘exciting’ channels are managed by a large media agency. Where JMPs and in-store media are treated as the ‘cost of doing business’, with low expectations of compliance and even lower delivery, a bad experience for the consumer, and a worse experience for the brand. 

The antidote for this is for aggregators and MetaDSPs to think on working vertically (deeper into the retailers’ media estates), and not just horizontally (across different retailers within the market). There will naturally be challenges with this, especially with setting up the back-end and revising processes around booking and managing of heritage media channels, and needing to revise commercial partnerships to take into account the set-up of these media estates and the areas which are accessible with committed spend. However, the benefits on offer make this a challenge worth tackling. We are seeing early signs that some of the aggregators are flirting with this approach, but for this vertical approach to be effective, retail media agencies will have to go a lot deeper.

The gold standard here would be for a brand to be able to access a full retail media estate through a central MetaDSP, which sits on top of multiple buying functions within the retailer - both the traditional in-store channels, and the programmatically bought retailer offsite. Initially this would likely be a combination of a white-labelled DSP for programmatic ads sat alongside a simply skinned form allowing the submission of approved media plans to the retail media agency working in the back-end to collaborate with the relevant internal teams. With investment behind it, and the right stack, there’s no reason these couldn’t be gradually moved into allocation tools that are pegged to demand, and monitored by a Yield Manager to protect the most valuable inventory, allowing these resources to be bought programmatically.

For the true value of retailer first-party data (1PD) to come to fruition, more of the media estate needs to integrate this data, and this can only really happen at scale through a programmatic buying set-up. This sort of collation of spend would give genuinely integrated plans; feed retailer 1PD through the full media estate; deliver the scale and sales sought after by the retailers; and give efficiencies on resource for brands with more end-to-end campaign planning - all coming to a head in more effectively integrated experiences for consumers.

The scale of this opportunity should put vertical integration at the top of the agenda. Happily, it’s also more immediately deliverable through the existing retail media partnerships. The onus sits with retail media agencies, and the retailers who contract them, to drive the step change here. And with the increased attention the sector is receiving, there’s never been more appetite to drive sophistication within these estates. Whilst it won’t be simple to translate 20 years worth of heritage media channels into programmatically bought aggregators, this seems like the quicker win than the horizontal opportunity, of aggregating cross-retail opportunities with complexities around ownership of 1PD and the driving of competitor sales.

The move horizontally should come as the market matures and more of retail media is available through these aggregators and Meta-DSPs. All this to say, it’s exciting to see the growth in interest in retail media, but the question for the aggregators and MetaDSPs is this: will they use the metaphorical leather trousers and perm of this convergence of attention as a means to drive genuine step-change in the industry, providing truly end-to-end campaigns from top of funnel through to purchase? Or just the shiny new toy of retail programmatic and ecommerce?