Can I Do Retail Media?

Ads are a wonderful and tempting business to be in. At some point, many founders realize that they have a critical mass of user attention that they could turn into money. This is both true and false. Yes, if you have enough users on your platform, it is possible to build an ad business. However, delivering on this potential is far harder than most people assume. In my experience leading Walmart’s retail media advertising business and getting Instacart’s retailer ad business off the ground, I’ve seen what separates the winners and the losers from nailing this strategy. 

‍

The good news is this: Vertical SaaS is strongly complementary to building out a retail media advertising arm. In this essay, I’ll lay out the prerequisites and potholes for the founder on this journey.

‍

Is my business suited for “retail media?” 

The simple definition of retail media is when a company that sells something has an additional business of selling ads to the companies that make that something. But it gets tricky when you talk about retail media in the context of a market.

‍

• For a retailer, the “something” is a physical good, and the companies that make that physical good are manufacturers. Walmart sells a lot of Cheerios, probably more cereal boxes than any company in the world, so General Mills (the maker of Cheerios) buys ads on Walmart.
• The picture is slightly more abstract for a goods marketplace. Marketplaces like Etsy, Faire, or eBay don’t own or stock any physical goods; they facilitate digital listings from third-party sellers, and it’s those sellers who buy the retail media ads. This is also the dominant ad model of the retail business on Amazon and Alibaba.
• For a services marketplace, the “something” is a digital listing of a service, such as a hotel room, a flight, a handyman project, or the availability of “Buy Now Pay Later”  financing. The companies that might advertise those services in the marketplace can be the service providers like the hotel chain or airlines who buy ads on Expedia, the handyman on Thumbtack, or Pottery Barn on Afterpay. 

‍

The common theme here is owning the transaction. If you have a base of transactions happening on your software platform, then you have a straight-ahead opportunity to add retail media advertising as an adjacent business. It’s the transaction data that creates ad value.

‍

• The newest expansion of retail media is services platforms. Shipping and ride-sharing companies, hotel chains and airlines, banks and credit cards—they are all exploring adjacent media businesses. This is where some Tidemark Vertical SaaS companies will play. In this space, the platform sees transactions that define valuable audiences, and the advertisers can be anyone who wants to reach those audiences.

‍

What are you selling? Who are you selling it to?

You can power rank the advertising options from best to worst:

‍

1. You're selling visibility for the products or services that you sell (i.e., the retailer and marketplace options above)
2. You are selling access to the audience of customers who bought something (i.e., the services platforms option)
3. You are selling whitespace on the website and apps that you own

‍

Those first two options are healthy additions to your core business. At eBay Advertising, we took the transaction data from annual sales of 100M cellphones to create advertising value for T-Mobile and AT&T (option #2) because what a new phone owner needs is a cell service plan. We then built ad products for the 3P sellers listing the phones (option #1). Ditto for cars and car insurance. 

‍

eBay also did some of option #3—sell random ads in the white space—and I can tell you from experience that this tactic can too easily lead to the “enshittification” of your product, where the service becomes worse while also more profitable. Don’t do this if you have better options.

‍

So who should you be selling ads to? Look for some set of potential advertisers that are relevant to your SaaS platform's core business—either companies whose products/services your platform transacts with, or related companies that want access to your audience.

‍

At this point, you still need to be introspective. Ask yourself: do you have an audience that advertisers really care about? Even if you do, the juice might not be worth the squeeze for pre-scale startups or niche audiences. As a rule of thumb, you can’t do advertising with less than 1M customers. (You could do advertising’s bare-knuckle cousin, lead-gen, but that’s an entirely different blog post.)

‍

The underlying question is, “Why would they buy ads from you?” and the answer is usually attention. Advertising is about paying for attention, so in a commerce environment, manufacturers pay for higher visibility by being ranked better on the first page of search results. Even if you aren’t doing commerce, you are selling the attention of your hard-won audience to a company that can’t find it somewhere else. 

‍

Uber is an example of both of these approaches: In their UberEats app, they sell attention via a search listing to restaurants that you can order delivery from. That’s a $500M run-rate business today, integrated into the back-end and front-end of the core platform. In the traditional Uber app and on screens inside some cars, Uber is selling the attention of a traveler en route—often a business traveler going to a specific destination. That attention is useful to local restaurants, entertainment businesses, or any company trying to reach business travelers. This ads business has a run-rate in the hundreds of millions. (Full disclosure: My company, Criteo, works with Uber Eats on retail media for their grocery-delivery business, and that run-rate I will not disclose.)

‍

The most important differentiating component is that retail media can “close the loop.” That means the platform can prove the performance of an ad in terms of driving a transaction. So in a retailer or a marketplace, a place where transactions are occurring, you can easily demonstrate that the person who saw the ad or clicked on the ad bought the product that was advertised. That's the business of closing the loop. The easier time you have of measuring results on your platform for an advertiser, the easier it’ll be to spin up this business unit. 

‍

It sounds simple, but that's the engine driving $140B of annual investment globally, with the North American market on the path to $100B. Ads tied to transactions with data that closes the loop are the fastest-growing part of all advertising. 

‍

When you combine these questions, you can see the path to monetization for your Vertical SaaS business. Ask yourself: What transaction data does my platform see? What valuable audience does it attract? Who would buy attention from that audience or for the things I sell? Is that attention worth it in terms of value and scale? Can my platform “close the loop” to prove that the ad had a commercial impact? And lastly, is the scale on my platform big enough to make going after the ads opportunity worth it?

How to set up an ad business

If you think you have necessary components––the data, audience, scale, and ambition––to launch an ad business adjacent to your core operations, the next question is how to execute. This implementation centers around three primary considerations: Sales, Technology, and Culture.

‍

1. Sales team

‍

Ad sales differ fundamentally from SaaS sales—the same customer needs to be resold every quarter, and annual contracts are rare. While starting with an evangelist/lone wolf seller can work well, you’ll still need strong performance management and clear goals. Additionally, you’ll need at least one back-office person from the start. This individual should be focused on billing advertisers monthly, reporting on ad performance results, handling creative assets, etc. Do the math carefully. If you can’t afford two full-time positions against your projected revenue, don’t start selling ads.

‍

A common misconception among first-timers is thinking that a well-designed self-service platform or user flow will remove the need for salespeople. The reality is that “self-serve” is never truly self-serve. While these systems can significantly reduce the back-office staffing needed for campaign setup and management, they don't bring in customers - that still requires dedicated salespeople.

‍

2. Technology

‍

You can either build advertising capabilities into your core product or rent an ad platform from someone else. (Repeat disclosure: I work for a company that rents a popular retail media ad platform.)

‍

‍Option 1: Build‍

‍

This approach applies only to monetization cases where you will sell ads to the same companies that pay for your core platform’s service today. Typically, this works best in vertical ecommerce marketplaces, where sellers pay for increased visibility (e.g., restaurants on Uber Eats or hotels on Expedia). The main advantage is that the ad purchasing can be integrated into your listing flow, and you already have systems for measuring outcomes.

‍

However, this approach requires significant engineering resources and will complicate your user flow. It’s OK to MVP this, but scale your revenue expectations accordingly. Important warning: If the companies buying the ads are different from the companies buying the rest of your services, then do not build. Lastly, be wary of engineering overconfidence, which will tell you it’s not that difficult to build an ad server. 

‍

The fine folks at Tidemark asked me to back up this assertion, so consider these fundamental challenges of in-house ad serving: How will you filter out unbillable bot traffic? How will you forecast specific ad capacity? How will you balance monetization and relevancy - and measure success? How will you control budget pacing as user activity fluctuates? These are just the basics. My advice: "Don't get into a land war in Asia, and don't build your own ad server."

‍

Option 2: Rent

‍

This route suits companies unwilling to spend engineering resources on building commodity ad capabilities. Even with a rented ad stack, your tech team will need to handle integration work, especially for pushing custom audiences into the ad server. Expect rental costs between 10-25% of revenue, particularly at smaller scales. These costs are separate from internal sales team expenses, though some ad stacks include sales services in their fees. 

‍

So, do the math: You probably hope for $1M annually in gross ad revenue to make the retail media exercise worth it. You might pay $250K for software and $500K for a starter ads team. The final $250K could be consumed with engineering time for setup and integration. Yup, the result is that you’d break even on a million bucks in Year 1, and you have a 25% gross margin business from then on. But, if you grow capacity to $2M in gross ad revenue, then you have a 50%+ gross margin—which is why people do ads. 

‍

Many companies rent initially, then build once they scale - a logical approach for Vertical SaaS companies with strong tech talent. The crossover calculation is simple: When does annual rental cost exceed ongoing engineering team expense? For most companies, this point lies well above $100M in gross revenue. Quick math: At $100M gross ad revenue, a 10% ad stack cost equals $10M annually. This could fund roughly 25 full-time tech staff. If you can build what you need with 15 people, you'd increase profit by $4M yearly. But this raises a crucial question: "Is building an ad stack the best use of 15 headcount?" This leads us to our next issue...

‍

3. Culture 

‍

The cultural impact of retail media is perhaps the most underrated consideration, yet it often determines success or failure. Product-led organizations frequently view ads as a distraction or, worse, a corruption of their clean user interface. Since retail media typically serves as a secondary, adjacent revenue stream, organizations risk dismissing it as 'the dirty thing we do for money.' No startup can afford this kind of cultural mismatch.

‍

If you do ads, leadership must embrace it as a legitimate part of the overall strategy and be able to clearly articulate what new ad revenue will enable. For instance, could adjacent ad revenue allow you to offer your core product at a lower cost, accelerating customer acquisition? This might enable a freemium model or - as with Netflix - a lower-priced subscription tier to drive growth. Or perhaps ads help your core customers scale rapidly, demonstrating your platform's value faster (a la restaurants on Uber Eats or retailers on Afterpay). Your entire team - not just leadership - needs to understand and articulate these strategic benefits.

‍

Everyone has to know the “why” of ads, or culture will hamstring the new business. In e-commerce-based retail media, it is typical for the team controlling the front end to oppose efforts to insert advertising units. Purely financial arguments about how much revenue justifies hurting the user experience are a sign that your ad strategy is losing to culture. 

‍

Will ads be meaningful?

Ad businesses require a path to millions of users, with things becoming particularly compelling at 10M users. This scale presents a challenge for B2B vertical SaaS companies, though it becomes more achievable when a Vertical SaaS Vendor (“VSV”) extends to the consumer. B2B2C SaaS platforms can touch millions of consumers and tap into scaled transaction data, which gives them a special shot at creating valuable audiences. Sometimes the consumer touchpoint is not obvious: a SaaS company powering a retail point-of-sale system could use ads to influence a lot of real-world shoppers (looking at you, Dutchie).

‍

If you have line-of-sight to 1M users, then you can size the ad's potential based on ARPU (average revenue per user). While best-in-class businesses can generate upwards of $30 ARPU, most achieve closer to $3 per user. This 10X difference stems from multiple factors: demand versus supply, audience reach exclusivity, ad sector maturity, sales activity quality, and a host of other factors. If your customer reach multiplied by a $3 ARPU reaches single-digit millions, then ad monetization merits consideration. If your SaaS platform is the only way to reach a difficult-to-find audience, like homeowners in the middle of moving, or your marketplace participants have a 30% margin to spend on growth, like Air Jordan speculators on StockX, then you can aspire to double-digit ARPU sooner. In this way, your success in dominating a valuable vertical directly feeds into your retail media success. 

‍

That is the overall point: Ad businesses typically succeed or fail based on their vertical’s value and scale, with retail media businesses adding a tantalizing ability to close the loop on transactions. Anyone in Vertical SaaS reading this is already striving for vertical excellence and scale, so it’s wise to think about where ads might fit in—or accelerate—your future success.