5 Examples of Information Asymmetry in Digital Advertising

by Jordan Bentley
100 US dollar banknotes

Information asymmetry occurs when one party has some key knowledge that the other party doesn't have. 

A simple example might be when your friend sets you up on a blind date with their weird co-worker.  In this case, your friend has some key knowledge that you didn't have, namely that their co-worker is weird.  A classic, more financially motivated, example is that of a bad used car often referred to as a lemon.  In this case, the car dealer has some material information such as the engine being defective which they do not share with the prospective buyer.

It turns out that in digital advertising, information asymmetry is a big problem.  Here are five common examples:

  1. Media Arbitrage. Media Arbitrage is when a vendor, platform, or agency charges a client for media at a higher rate than what they bought it for.  This is one of the most popular business models and is completely ethical as long as the advertiser is aware of it.  We see this same business model in the physical world with retailers like Walmart.  Walmart buys products like milk at a wholesale rate and then sells it to consumers at a higher rate than what it was bought for.  However, there are a couple key distinctions to point out.  First, when a consumer buys milk, they know exactly what they are getting.  Milk is regulated by the FDA which protects consumers from buying milk that is really just water.  Second, consumers can compare the price of milk.  This limits Walmart's ability to say charge $20 for a gallon of milk because no one would pay that price.  In digital advertising, advertisers do not benefit from either of these protection.  Moreover, as is common with media arbitrage, reporting and transparency are often limited.  As a result, the vendor, platform, or agency buying the media tends to prioritize buying cheap low quality media in order to increase the amount of money they make (buy low, sell high).
  2. White Labeled Solutions.  With the rise of digital advertising platforms like Google's Ad Words, Facebook, and Criteo, running digital campaigns has never been easier.  Moreover with the popularity of APIs, vendors can easily create what appears to be their own platforms that in reality are powered by other 3rd parties.  As a result, the number of companies reselling advertising solutions has never been higher with some estimates over 100,000.  The challenge with having so many white labeled solutions is that they rarely disclose how much cost they are adding into the transaction.  It is not uncommon for as many as five white labeled solutions to be involved in a single media buy.  In this case, five different companies have to each add their cost on top of the original raw media cost which is ultimately paid for my the advertiser.  With so many costs being added into the transaction, it is nearly impossible for that advertiser to generate a positive return on investment.  Of course, due to information asymmetry the advertiser has no idea that their campaign is doomed from the start.
  3. Pay Per Click.  Pay per click is the most common form of performance based advertising.  The model is quite simple, you only pay the platform or vendor when someone clicks on your ad.  It turns out, unsurprisingly so, that in order for the platform or vendor to make money they have to be pretty good at predicting when someone is likely to click on an ad.  If they are not, then they will spend too much money on running ads such that the rate the advertiser is paying is either unprofitable or simply too high.  To control for this, these platforms and vendors tend to show ads to either (1) people who tend to click on a lot of ads or (2) places that tend to get a lot of clicks like apps.  Most real world consumers do not meet either of these two scenarios.  So as an advertiser, how to you know if the clicks that you are paying for are the result of real interested potential buyers or someone who just likes to click on ads?  This is where information asymmetry comes in because you do not really know.  Advertising platforms and vendors that run pay per click campaigns do not disclose what tactics they use to produce a low cost per click.  Moreover, as an advertiser you do not really know if the person that clicked on the ad was a business owner looking to buy your product or a 14 year old kid that clicked on the ad to get some free points to the game he's playing on his phone.
  4. Lookalike Audiences. Lookalike audiences rose to popularity in the last five years as the popularity of advertising on Facebook grew.  Of course, targeting lookalike audiences is not unique to only Facebook as many platforms offer this option.  One of the main reasons marketers use lookalike audiences today is to help increase the audience they are targeting as only targeting the potential customers in your CRM (customer relationship management) may result in a very small audience.  Lookalike audiences by definition may have similar characteristics or attributes to your current or potential customers but are people you likely aren't currently targeting.  As a result, targeting a lookalike audience can obviously be appealing however there is just one key question.  How similar is this lookalike audience?  Again, unfortunately due to information asymmetry advertisers do not really know.
  5. Campaign Performance Reporting. The purpose of advertising at the end of the day is to ultimately increase revenue for an organization.  Advertisers naturally want to spend their ad budgets on the campaigns that are producing the highest return on investment.  Similarly vendors, platforms, and agencies want to retain and grow the advertising budgets that are being spend with them and thus include some type of campaign performance reporting as to increase these odds. As a result, the goal for these 3rd parties is to identify a way in which they can claim as much credit for the success of the advertiser as possible while technically not completely lying.  To accomplish this, 3rd parties will evaluate various attribution models to see which shows the greatest impact and then justify the use of said attribution model.  In addition to this, many 3rd parties will also include the use of some cross device model as this allows them to attribute credit to an ad that may have been shown on a desktop when a conversion occurred on a mobile device.  All of this creates a clear conflict of interest for these 3rd parties as they are effectively grading their own homework.  This becomes a key issue of information asymmetry as these 3rd parties rarely share the cross device model and assumptions that they used or the details regarding the attribution model that they used.  In the end, advertisers are left comparing apples to bananas with the various reporting that they receive from the different 3rd parties that they work with.

At Spill we are passionate about about helping marketers solve the problem of information asymmetry by providing complete transparency around campaigns executed by 3rd parties.  If you have any questions regarding any of the tactics mentioned here, please feel free to reach out to us and we would love to answer any questions that you have.