While the death of the cookie has once more been put on hold by Google, we know it's on the way out the door. Luckily, brands have already realized the power of Retail Media Networks to target and reach consumers across their paths to purchase using robust, first-party loyalty and membership data.
RMNs offer brands the chance to use that robust data to create truly personalized experiences. And that's pivotal; according to research from McKinsey & Company, 71% of consumers expect personalized experiences when shopping. When combined, retail media and the power of AI offer immense opportunities in this space for retailers, brands and shoppers.
We tapped James Taylor, CEO of Particular Audience, to explain how advancements in AI have enabled new targeting opportunities for brands and personalized experiences for shoppers via RMNs.
Threefold: Describe Particular Audience in three words.
Particular Audience: Private. Cookieless. Relevant.
Threefold: Let’s dig deeper. What is Particular Audience, and how does its technology support retailers with their Retail Media Networks?
Particular Audience: Particular Audience is an AI company that democratizes access to the technology that makes Amazon so effective at product discovery, conversion and sponsored products. Particular Audience is used by retailers globally from Target to Hamleys to grow basket size, convert more customers and remove the bottlenecks in retail media performance.
The on-site engines on most Retail Media Networks are still painfully manual in the same way that visual merchandising was manual 10 years ago. Particular Audience uses Artificial Intelligence to create a unique website, mobile, app and email experience for every customer.
Threefold: How can AI drive personalization in a cookieless era? Walk us through some of the possibilities and how retailers are already leveraging this technology.
Particular Audience: With data privacy regulations such as GDPR and CCPA, with Google phasing out third-party cookies (eventually?!), and iOS enabling opt-out of cross-app tracking, ad tech has been flipped on its head.
It is getting much harder to gather data and target ads.
Retailers need ways to better understand their customers without Personal Identifiable Information (PII), and for this reason, they’re looking to their richest source of first-party data: their products, their sales and the implicit engagement data they ingest in droves.
Product data is valuable because you have:
100% of your item data
Every single item interaction
Every item purchased
Every search
Every item image
All of your product description data
Why is this powerful? Demographic data, location data or information about what a customer bought six months ago can’t tell you anything useful about the in-moment intent of a customer. Product data and ‘wisdom of the crowd’ algorithms shortcut the path to items shoppers likely want at any given moment, making this privacy-preserving strategy a more effective way to understand your customers.
Search intent has also grown in importance because of changes in cookies and privacy because it is current, contextual and explicit with opt-in.
Threefold: Why is personalization in online media important for advertisers to get right today?
Particular Audience: People don’t click on irrelevant results. It’s a bad customer experience, it’s bad for sales and it’s a wasted opportunity for advertisers. At scale, it is also bad for the environment.
At Particular Audience, we believe the on-site engine of retail media should live within AI-powered search and personalization. This is how Amazon and Walmart have expanded inventory so successfully, and why that gap is why Amazon’s rivals in the US alone are estimated to be leaving $43 billion of ad revenue on the table each year.
You can’t scale inventory unless you have solved relevance.
Threefold: How does online search factor into the marketing funnel and why is getting this right important for retailers?
Particular Audience: Search is an exciting, explicit signal of intent, but keyword targeting is an archaic approach employed by the largest RMNs outside of Amazon and Walmart. Keyword search is a fundamentally broken technology requiring exact matches and a mountain of synonym rules and hardcode redirects to make search work for shoppers.
We are now living in the era of AI. Sentence transformers, large language models and AI that learn from real-life customer search behavior are combined to create intuitive search experiences capable of reducing zero-result search terms to near zero and automating the placement of sponsored products where relevant.
Consider Google’s ‘Broad Match’ capability and how much easier it makes running performance ads on Google. Advertisers don’t have time to linger in the longtail. RMNs should strive for the same AI-powered simplicity for their suppliers; the reduction in friction means more budgets will flow without human bottlenecks.
Threefold: How does Particular Audience prove the value it drives for retailers and customers? (measurement)
Particular Audience: Incrementality is a hotly debated topic in retail media today, but it shouldn’t be. Do customers buy more and more often? This North Star is good for the retailer, the advertiser and most importantly, the customer.
Particular Audience offers all the standard metrics from engagement to attribution, click-through and conversion. But attributed revenue does not necessarily mean incremental value.
Going deeper, PA can report on share of voice and share of basket, acknowledging the inevitable cannibalization of organic sales that would have happened anyway, versus the ginormous uplift sponsored product ads can have.
Retailers can test different algorithms against one another at a slot level, to optimize for engagement and/or conversion to benefit ROAS metrics. Soon, retailers will even be able to AB test search models within the PA platform, an industry first.
Threefold: Data protection is key to keeping consumers safe and happy on retailers’ digital platforms. How do you ensure your solution is unobtrusive, and privacy-compliant while delivering meaningful experiences to shoppers?
Particular Audience: I think this question is best answered by stepping back and considering some first principles on what data is required to make an experience world-class.
1. Personalization: A world-class experience is intuitive, automatic and personalized. Personalization is just good prediction. Of all the things you have to show someone, which one do you put first?
2. Data: Asserting to a customer their own interests or intent, based on some tenuous demographic data, is littered with pitfalls and simply does not work. A retailer has a tiny fraction of a percent of data from a customer. Customers are diverse and dynamic individuals with constantly changing needs, tastes and preferences. There are more customers than items, and the data is incredibly sparse.
Conversely, a retailer has 100% of their item interaction data, purchase data, descriptive metadata and image data. We can understand what items people compare with one another, what items should be cross-sold together and which items are descriptively or visually similar. You can accurately predict that if a customer engages with item #1, followed by item #2, you can sure as hell bet they’re likely to be interested in item #3.
Unlike customers, items are commoditized, there are many units of the same thing. The data is robust.
3. Context: Personalization needs to understand customer context in real-time, so it can predict what to recommend and how to rank it based on the probability of click and conversion.
There are two main approaches to doing this well, implicit and explicit signals.
Implicit signals are the items a customer engages with, what they add to cart and what they purchase. Explicit signals are things like searches, and navigating to a listing page.
These signals combine. For example, if I’ve been browsing Mens Clothing and I navigate to the Jeans Category, I should probably see Mens Jeans at the top.
Context is king.
4. Privacy: I think you saw this coming, but the beauty of all of this is you don’t need to know anything about who a customer is to predict and personalize experiences for them.
A retailer’s rich first-party product data, and real-time contextual signals are enough. We assign 100% random IDs to customers, which infer from implicit and explicit signals what these shoppers are interested in. We then use that to show them the most relevant product(s) to them.
This makes search and recommendation an intuitive, automatic and personalized experience.
That logic powers Particular Audience Sponsored Products Ads and is how PA achieves a 300% increase in ad fill rate, ad revenue and conversion.
It’s better for customers, retailers and advertisers and creates a materially more scalable and programmatic Sponsored Product Ads engine, requiring little to no manual intervention.
Threefold: What do advertisers need to consider when creating approaching retail media online to help them win?
Particular Audience: Customer Experience should be at the heart of any promotional decision; a strong customer experience solves every other metric. Advertisers should look for RMNs that share this ethos, and some things to consider are:
1. Are sponsored ads deeply intertwined within on-site search, merch and personalization? Is it one seamless system? Do you get the exact same sponsored placements in the same location when you open an incognito window or do they adapt to your unique journey?
2. There is a narrative in the market that the gold standard is to return relevant sponsored products in all search queries, as only really Amazon and Walmart achieve today. To really close the gap, I encourage a step further, returning relevant SPAs with every page load.
The hard fact is a retailer can only increase inventory once they have solved relevance. Until they do this, CPCs will remain inflated and dominated by a few large brands.
3. Are placements able to be explained, measured and optimized? Is relevancy a quantitative measure you (or the retailer) can test and configure? If there is a black box with no ability to explain why something might be happening, then there is no way to optimize results.
Rapid fire
The future of retail media is relevant, scalable and automatic
In 2024, I’m looking forward to the death of the cookie
The biggest misconception about retail media is that customer attributes are more important than behavior and context.
About Ella Rice
Ella Rice is a client manager in the Threefold North America team, where she supports CPGs in planning, building and measuring retailer campaigns. Prior to her role at Threefold, Ella spent three years on the Plan-Apps team within SMG, supporting CPGs with insights on campaign performance and building consultancies that leveraged Plan-Apps data. Ella previously worked as an account manager for a digital marketing agency. Want to chat about how Threefold can supercharge your retail media strategy? You can reach her at ella.rice@threefold.team.
About Threefold
Founded in 2008, Threefold is the world's leading Retail Media Network (RMN) specialist, headquartered in both New York and London. As part of the SMGÂ agency network, which employs over 380 retail media experts, Threefold's primary mission is to unlock incremental CPG budgets, curate media campaigns that supercharge sales and elevate its retail partners into top-tier omnichannel media owners, spanning in-store, off-site and online.
Threefold's services include consultancy, evaluation, and a white-label in-house solution that has seen the agency build, run and operate over 10 Retail Media Networks to date. In the UK, Threefold runs and operates Walgreens' Boots Media Group and Morrisons Media Group and additional live partners include major retailers like Asda, The Co-op and The Very Group.