Category: Affiliate Marketing Tips

Where those concerns originate from is The Trade Desk’s efforts to consolidate its buying power into fewer but better paths to programmatic inventory. This is commonly referred to as the “GPID” initiative and it goes back 18 months or so. 

But those attitudes have hardened over the last 12 months as the ad tech vendor set out to build its own direct routes to programmatic inventory via OpenPath alongside GPID. 

Speculation is never far from moves like this — a company crossing the proverbial rubicon. Time and again the industry has seen how these moves become opportunities for disintermediation. To be fair, The Trade Desk hasn’t done this. In fact, it has gone to great lengths not to upset the apple cart. That said, what it has done leaves a lot to the imagination. 

Here’s why: Let’s say The Trade Desk identifies all supply paths that sell the same ad slot on a news site. It then decides that the fair market price for that ad slot is a dollar. And then it measures its chances of winning the chance to buy that ad at that price point in each supply path. 

So when it bids a dollar via exchange one it wins 20% of the time, when it does the same via its own paths to programmatic inventory it wins 22% of the time, and when it bids a dollar via exchange two it wins 1% of time. Something is clearly wrong with that last exchange. Maybe it has an astronomically high take rate that makes the dollar price per thousand impressions gross bid uncompetitive on a net basis. Or maybe there’s some sort of technical failure that makes the auction for that ad slot malfunction more often than not. 

Either way, The Trade Desk doesn’t care. All it cares about is being able to buy inventory at a fair price through that path. If it can’t then that path gets deprioritized. 

Rejection like this is hard to swallow at the best of times, let alone when the rejected can’t accept why.

“There’s not a lot of clarity on how a path gets prioritized but from what I can gather The Trade Desk’s bidding algorithms seem to favor those paths to ad inventory that have a low take rate including its own,” said one ad tech exec who traded anonymity for candor on what little clarity they’ve been able to get into how The Trade Desk spends its ad dollars. “This essentially means I may have to handicap my margin to compete, which will drive down my revenue and potentially challenge my business.”

That makes The Trade Desk’s intent (arguably) more ruthless than nefarious. 

After all, OpenPath has a low take rate and a team of execs who are working hard to optimize a direct path to publisher inventory. Of course, it’s going to buy more ads from this supply path — it’s one of its most efficient ways to buy ads. SSPs and ad exchanges may not like it because every bid not won by them is a fee not taken. But The Trade Desk is under no obligation to support all of them. There are few businesses that are expected to behave this way. 

“What’s interesting to me is the expectation that the DSP — in this case The Trade Desk, but it applies to any DSP — needs to explain ‘why’ they are using Path A versus Path B? Do SSPs explain why they choose to work with certain DSPs, but not others?,” said Mike O’Sullivan, co-founder of ad tech tracker Sincera. “Undifferentiated offerings suffering from price erosion isn’t a “Trade Desk” initiated trend. It’s the prevailing trend of programmatic.”

But is what The Trade Desk is doing fair? Some ad tech execs say no. 

“There’s a technical advantage The Trade Desk has over us now as well as a commercial one when it comes to giving advertisers a reason to buy more publisher ads from their own path rather than ours,” said an ad tech exec who would only talk to Digiday anonymously due to concerns of jeopardizing their relationship with The Trade Desk. “When you have influence over how ads aren’t just bought but also sold there will always be the cynics who say it has created the potential for bias.”

The bias being that The Trade Desk could eventually start to push more money toward its own paths to programmatic inventory over others even when it may not be in the best interests of its advertisers.

To be clear, this isn’t happening. On the contrary, The Trade Desk’s attempts to consolidate its spending around where it can get the best efficiencies haven’t cost publishers money they were already making from other ad tech vendors.

This is how The Trade Desk’s vp of inventory development at The Trade Desk Will Doherty put it to Digiday: “The money will transact to more and more premium publishers. But it will just do so through fewer paths. When that starts to happen the overall efficiency of the market rewards the edges, in this case buyers and publishers.”

What Doherty is saying is that The Trade Desk is indifferent to all supply paths because price in exchange for ad quality clears the market.

“It’s the purest form of SPO possible,” said Tom Triscari, an economist at consulting firm Lemonade Projects, who wrote an analysis on the systemic issues that have led The Trade Desk to this point. ”Sellers that sell worthless inventory for a living are worried and they should be. They have lived fat and happy for a long time now. Either get legit or die. That’s the result of indifferent pricing. It’s about time. Assume TTD intent is legit too.”

So why the heartburn from the naysayers? Additive as The Trade Desk’s consolidation efforts are now, they could also give it a lot of leverage down the road. 

It’s a bit like when ad execs used to give Google’s ad tech (DV360) for placing bids on programmatic inventory a hard time for running on its own ad exchange even though the media giant insisted the two technologies were agnostic.

And therein is the real issue for some ad execs. They’re grappling with some learned skepticism. 

“I don’t really care that The Trade Desk wants to go vertical and exert more control over the programmatic market because that’s the name of the game these days,” said an ad tech exec who spoke on condition of anonymity. “What they’re doing isn’t illegal. The reason I’m bothered is at times it reminds me of how Google has behaved.” 

It’s an interesting point given The Trade Desk sees itself as the antithesis to Google. But it’s arguably a bit of a stretch. There’s no technical advantage gained by The Trade Desk as a result of what it’s doing. It can’t fiddle with the auction mechanics of OpenPath to win more bids any more or less than when it bids via normal programmatic marketplaces.

That’s not to say all the furore over The Trade Desk is moot. Should it reach a point where publishers rely on the ad dollars from the advertisers that use it then it could use that influence to get them to do things they wouldn’t have otherwise. There’s nothing that’s happened to say this paranoia will manifest. 

The problem The Trade Desk has, however, is convincing the cynics there’s not a sting in the tail. It’s hard to change the narrative once it has established itself. There could be far bigger implications to this issue if it is not addressed. 

The ire of regulators, for starters. Increasingly, they want the largest ads businesses to be able to demonstrate that they’re able to responsibly manage conflicts of interests. Hence, the emergence of the Competition and Transparency in Digital Advertising (CTDA) Act, which specifically targets vertical integration in the ad industry. To the untrained eye, it could seem like The Trade Desk is doing just that. It may not be operating a traditional SSP, and it isn’t subverting their existing commercial deals with publishers, but it has hit them.

“My perspective on this topic is informed by my curiosity about CTDA Act which, if passed, would treat The Trade Desk like a digital advertising exchange and like a sell-side brokerage and would, therefore, bar the company from running OpenPath,” said Nikhil Lai, a senior analyst at Forrester. “Specifically, the Act’s rules about how companies must avoid conflicts of interest would limit the viability of OpenPath, which The Trade Desk can use to favor its integrations with publishers at the expense of SSPs.”

Editorial executives previously told Digiday in February they were encouraging employees in their newsrooms to familiarize themselves with generative AI and chatbots like ChatGPT to see if the technology could help them perform their jobs better.

Now, some companies have formalized teams to lead AI-driven experiments and projects:

BuzzFeed has a “brain trust” of about a dozen people that formed in December 2022, a spokesperson said. While it’s an “informal” and “fluid” group, the team consists of CEO Jonah Peretti, svp of editorial Jess Probus and founding editor and executive director of growth trends Peggy Wang. It also includes representatives from the tech and engineering departments, including a machine learning specialist.

BridgeTower Media has a seven-person team that was put together in mid-March by David Saabye, svp of digital product management at the B2B media company, to focus on AI experimentation and guidelines. 

Forbes also formed a seven-person group in March charged with discussing and establishing AI-related policies, a spokesperson said. The group includes representatives from the company’s corporate communications, human resources, legal, data privacy, editorial and cybersecurity teams. 

Ingenio hired director of growth, Geoff Skow, a year ago to oversee AI initiatives. Skow works with a team of five people from editorial, engineering and UX design to determine how to use generative AI to grow Ingenio’s audience, said president of media Josh Jaffe.

Trusted Media Brands put together a team of nine people a month ago, led by chief technology officer Nick Contardo and chief business officer Cameron Saless. The group was formed to “talk to other leaders at this company to hear what they think about [AI and machine learning] and how this could impact them and what we could do with it,” Contardo said.

TMB’s task force is discussing one or two “mid-size” AI projects the company can invest in over the next six months, Contardo told Digiday, but holistically has three purposes: sharing AI learnings between departments, determining which AI initiatives to test and deciding on company policy around the usage of generative AI and ML technology. The team is made up of managers from several teams, including business development, editorial and sales and marketing.

“Our approach isn’t going to be to take things like AI or ML and replace employees,” Contardo said. Instead, the focus is to streamline certain operations and see if advancements in the technology can improve the surfacing of content to readers, he added. 

AI experiments “have been bubbling up organically throughout” BridgeTower Media, Saabye said. The focus of the new task force is to “identify [the] use cases, put some structure behind them, and then at the same time build out the policy… essentially, the safety rails of how this can be used at the company,” he said.

Most of BridgeTower’s AI team are managers and come from its creative and production, SEO and web development, editorial, marketing, data businesses and legal council departments. “They represent the core parts of the business where we see an opportunity for some AI functions to have a material effect,” said Saabye.

The team is focused on three things when it comes to the use of AI: “cost savings, increased velocity and new products,” Saabye said. In other words, looking at how AI can replace third-party vendors and bring down the company’s operational costs, as well as using it to boost productivity by conducting tasks faster and better than before, such as highlighting key trends within a story to improve SEO. Lastly, the BridgeTower team is experimenting with AI tools to see if they can create products to analyze trends or large data sets.

The BridgeTower team is also tasked with determining the quality, safety and ethics around the use of generative AI, Saabye added. 

“We are looking at having the company’s stamp of approval on what you can do today and how you can use it. And then just like you do with data privacy, having that be a regular function of the organization to continually assess and evaluate your policy and your actions,” Saabye said. “That’s why we have legal represented on [the team] as well. It’s not that we’re constantly concerned about doing something illegal, so much as there are so many unknowns particularly in the areas of IP [and] government regulations in this area.”

But a number of publishers are still taking a more organic approach to testing AI internally. Noah Weissman, evp of content at Team Whistle, is leading AI efforts at the company, but not with a formalized team. Each manager is leading tests within their own departments, he said.  

“Our head of talent [has] a very different use case than our head of production. But it is an absolute innovation and strategy mandate to use it,” Weissman said. “It’s my job as one of the leaders on our team to make sure that people are adopting it so that we don’t get left behind in the past.”

Gannett doesn’t have a dedicated AI-focused group, but the company’s head of product, Renn Turiano, is overseeing AI efforts with managers across the company, a spokesperson said. Chris Lloyd, vp and gm at Gannett’s product review site Reviewed, is looking at AI opportunities for affiliate revenue, for example.

Not all media executives are convinced that generative AI technology will be the solution to the media industry’s ailments, however, even if their leadership is entirely onboard. 

“My CEO is fucking obsessed with AI… but I’m not totally convinced,” said one publishing executive on the condition of anonymity.