Out with Ad Networks, In with Curated Marketplaces

In a business that loves its buzzwords, the conversation surrounding “curation” has practically gone viral — in more ways than one. Not only is everyone talking about the issue, but it’s making some people feel ill. Call it PANSS (Post Ad Network Stress Syndrome).

I get it. People hear about curation, and aren’t even sure if it’s a new concept. It’s giving some people a very unpleasant feeling of deja vu. Haven’t we seen this movie before — and didn’t it end badly, as evidenced by the number of ad networks that disappeared in the programmatic age, not to mention the scars many buyers and sellers have.

Well, I’m here to tell you that I understand your concerns but the concept of curation is most definitely not the ad network model simply dressed up in new packaging.

Allow me to bring some clarity to the situation or at least better articulate how I see things.

Ad Networks Are Gone for Good Reason

I speak from experience, having helped drive revenue at Advertising.com more than two decades ago. In the past, ad networks were primarily designed to offer efficiency to brands and agencies, compared to the more time-consuming approach of working directly with publishers. They were often characterized by low CPMs on a handful of recognizable media properties, with a glut of longtail publishers lurking below the surface, offering cheap reach and strong click-based engagement at scale. Ad networks optimized for themselves first and foremost, and advertisers and publishers a distant second, and third. Curation is both more transparent and clear on that proposition.   

The advancements of ad technologies — the ad exchange, DMPs and DSPs — subsumed these value propositions, with dollars quickly flowing from ad networks into open-auction programmatic. The panacea for efficiency, scale, and performance delivered a new business model to agencies. 

A lot has changed since then: 

• Demand-side platforms: Buyers entrust DSPs to decide in real-time whether they should bid on an auctioned impression and how much they should bid, based on buyer established parameters, a step change from the ad network’s insertion order (IO) model.
• Third-party cookies: Programmatic’s fabric still largely functions on cookie matching. Yet only one in 10 browsers will support third-party cookies, per eMarketer. Audience targeting in DSPs is broken.
• Performance: More accountability is attached to media investment. Legacy performance metrics, such as unique reach and  CTR, served as primary KPIs in the ad network era and can easily be gamed by publishers (see the “MFA” playbook). They have been replaced by actual business “outcomes,” measuring ad influence on consumer attention, brand lift, purchase intent, and sales.
• Optimization: Ad networks used data-driven optimization for real-time decisioning, however the algorithms most times prioritized revenue or RPMs (revenue per thousand) over buyer objectives. Today’s algorithms are smarter thanks to AI, more relevant and use more signals, but importantly, they are focused on buyer goals.
• Quality: While viewability and fraud detection are minimum requirements for most brands, suitable content adjacency is also table stakes — going beyond site lists and keyword blocking. 
• Sustainability: Up to 20% of global internet infrastructure emissions are attributed to online ads, with programmatic as the primary offender, per the ANA, while 2% of publishers are responsible for 50% of emissions. Brands are paring down supply paths to performant, low-carbon inventory pools.  

There is an ongoing movement toward helping brands navigate the open web in a manner that is more palatable and predictable, guiding advertising dollars back to publishers, the bedrock of the internet.

Ad networks never could, and never will be able to check all these boxes.

Ok, So What Is Programmatic Curation?

Ari Paparo of Marketecture did a great job explaining some of this in his October 21 newsletter using our chart here, but I want to offer my definition.

When I refer to “programmatic curation,” what I’m talking about is a mechanism to purchase select inventory and custom-built audiences using data. This data is a combination of contextual clues, analytics, and behavioral insights, optimized for clients’ real business results, not proxies or guesses. This isn’t cherry-picking audiences or sites per se. It goes beyond what most DSPs offer and is well past what networks have ever promised. It’s about enabling brands and buyers to reach more of the right users in the right contexts without identity limitations, while avoiding sites they want nothing to do with (like MFAs), and optimizing based on individual campaign performance in addition to platform algorithms. 

Data is the differentiator in this new wave of curation, and not one to be taken lightly. This iteration of curation is aimed at delivering precision and scale, even amid massive signal loss. Buyers will still be able to reach their most valuable customers even without legacy identifiers, and they can do so in a far more transparent and carbon-efficient manner.

In other words, advertisers should get more of the inventory they want while crafting far more data-driven and effective media plans…while ad sellers of premium content should command more revenue for their high-value audiences. All sides of the industry should benefit.

Which is something that ad networks could never pull off, let alone promise.

This article was written by Andy Monfried, Founder & CEO at Lotame, and originally published here.