“Data-Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.
Today’s column is written by Nico Neumann, assistant professor at Melbourne Business School.
Programmatic has for years faced a lot of heat for its lack of transparency and measurement, brand-safety and fraud issues. Over the last 13 months, we’ve seen shocking headlines related to nearly every major ad tech company.
In particular, fraud-related news seem to pop up almost every week. Most recently, the Financial Times discovered fake inventory many times larger than what is actually available, Uber sued its mobile agency and an independent study referred to Criteo as malware.
What has created this chaos?
Probably, the biggest issues are the novelty and complexity of automated buying technology and the continuously evolving ecosystem. New tools and solutions are launched constantly, which can be overwhelming for everyone. Most technologies also operate in a black box, leaving ad buyers in the dark about what they sign up for and the quality of placements they obtain.
The lack of transparency and ability to evaluate media quality properly from the demand side create a significant information asymmetry between sellers and buyers. Economists refer to such a phenomenon as a “market for lemons,” based on the seminal paper [PDF] of George Akerlof.
A Lemon-Market Example: Used Cars
The classic example for a lemon market refers to used automobiles: The seller knows about the quality of the offered car, but the buyer does not.
For example, the owner may have tampered with the odometer or hidden scratches and other accident signals. Hence, the buyer faces uncertainty about the quality of the sold car and bears the risk of being left with a lemon – slang for a defective car – as opposed to a peach, an analogy for high-quality products.
The quality knowledge gap between buyers and sellers has important consequences for market dynamics. Let’s say a good secondhand car would be worth $2,000 and a low-quality one only $1,000. If smart prospects can’t check whether a car will be a peach, they may try to avoid paying too much and only offer the average value of $1,500 to factor in the higher risk of ending up with a lemon.
In return, this means that selling a peach, whose true value is $2,000, would not make sense for a seller. The only way to make a profit in such a market is to actually sell a lemon ($1,500- $1,000=$500 profit). This is a vicious circle, and the more lemons are sold, the more buyers will adjust their expectations and lower their prices.
The Global Financial Crisis (GFC): A Lemon-Market Crash Warning
Lemon markets are dangerous. Without intervention, the amount of lemons can reach toxic levels, potentially to such a degree that a market collapses because only garbage is offered.
Consider the GFC about a decade ago. While there were many factors contributing to the crisis, the so-called subprime loans played a key role. Subprime loans were risky mortgages for people with low credit ratings, which is why they yielded higher interest rates and thus greater profits. Driven by short-term gains, many lenders accepted even borrowers who couldn’t repay their debt, creating mortgage “lemons.”
Another key problem was that these risky loans were sold to investment banks, which packaged them into bundles and traded them again, often to others who regrouped them into another package. Following repeated product bundling, it was difficult to judge who owned a quality mortgage and who had only lemons in their portfolio. Eventually, banks stopped trusting each other, trading halted and the market crashed.
The Open Programmatic Market and Hidden Lemons
Now let’s consider the open programmatic market. One of the most popular placement types is run of network (RON). MarketingTerms.com defines it this way: “Advertisers generally give up say over placement in return for low rates and broad reach. Ads may be placed randomly in unsold, less valuable portions of sites within an ad network.”
No need to say that RON already implies greater odds of receiving a lemon; peaches tend to be sold via direct channels rather than auctions.
However, it’s not only RON ads that bear a noteworthy risk. Many demand-side platforms and exchanges also offer media-placement packages, grouping ads they have available from various networks and partners. The result is again a high level of opaqueness: The long media supply chain and the bundling of ads into line items make it extremely challenging to assess ad quality and opens up so many avenues for fraud.
Turning A Lemon Market Into A Peach Market
To avoid a market crash for our ecosystem, it should be in the interest of every stakeholder to reduce the level of lemons and foster programmatic trading of peaches.
How can this be achieved?
The most obvious solution is to regulate a likely lemon market, as done for the car and subprime markets in the US and recently even for advertising in France. The goal of any such laws is to enforce transparency and legal warranties that diminish the risk of being stuck with lemons.
Where regulations are missing, it is important for advertisers to create legal frameworks that provide them with guarantees and means to examine ad quality. Brands should demand direct access to their data and costs from any media vendor, use third-party verification and make conflicts-of-interest disclosures mandatory.
Google has acted as a strong role model here, requiring participation in Ads.txt when using DoubleClick Bid Manager after October. Advertisers should follow suit and remember that all power lies with them: They can vote with their wallets and support transparent platforms and agencies.
Finally, we must improve education for advertisers and media traders: The more they know about programmatic ad trading, the lower the information asymmetry between buyers and sellers and, thus, the chance to become a victim of lemon-market mechanics.