By Phil Schraeder, President & COO at GumGum
Everybody loves programmatic for reasons so numerous and familiar we have no need to list them here. But let’s be honest, it can blindside brand managers, and leave them feeling like the mother who drops her kid off at school only to learn later that he’s been picked up by the police. How’d that happen?
Brand managers have a handful of tools to ensure brand safety: blacklists, whitelists, and even third-party verification to screen placements. So why the steady stream of epic ad misplacements?
Clearly the tools aren’t enough. Blacklists offer some protection from sites you know you don’t want to be on, but there are holes in their capabilities. Most notably, they don’t stop your ads from appearing on fake news sites, a scourge of our industry and democracy. Fake news sites can look and smell genuine to open ad exchanges because to date there is no automated way to test a site’s authenticity. Hopefully artificial intelligence will address this gap in the near future. In the meantime, even when you blacklist a site, with hundreds of campaigns running simultaneously, how do you ensure all of your partners adhere to it? You can’t, as Nordstrom learned when one of its ads appeared on Breitbart.
Blacklists offer some protection from sites you know you don’t want to be on, but there are holes in their capabilities.
Whitelists add more protection, but the list of unintended consequences is long. If brands relied solely on super-premium sites, they might not be reaching valuable audiences who do not visit those sites. Whitelists eliminate a marketer’s ability to capitalize on extremely valuable content – blogs, niche sites – that millennials rely on to research purchasing decisions. And that, in turn, means brands can’t reach their audiences at critical points in the purchasing journey. It also hobbles the best part of programmatic: the ability to optimize campaigns by selecting impressions from a huge pool of inventory. In brief, by focusing on a small collection of whitelisted websites, brands are missing out. This problem is only made worse when there are multiple brands in a single organization. Company-wide whitelists create headaches for brand managers, as each brand has different reach and targeting needs. Finally, if brands were to buy solely on premium websites, the cost of inventory on those sites would skyrocket, and ROI would plummet. And they could lead to a StubHub of inventory, with buyers scooping it up to resell at higher prices.
That leaves third-party verification tools, which are designed by some of the smartest data scientists in our industry. These providers design sophisticated algorithms to assess page quality, by examining the frequency of content updates, the number of ads on a page, and even the environment outside of the current page of an impression. They work closely with brands to understand all of their parameters for brand safety, and provide them with what’s essentially a robust whitelist. The problem? It’s not always easy to understand what’s happening on the other side of the screen.
None of these tools can help with ad misplacements on YouTube or other video-based properties, because they can’t interpret the contents of a video. A YouTube video channel can appear brand safe in every way: its name, its following, its content category. A channel all about puppies can be a great place to advertise, until one of those videos has a swastika in the background.
The Bigger Issue at Work
The industry can (and no doubt will) make improvements on these brand-safety checks and balances. We’ll deploy AI and machine learning to detect fake news sites, and computer vision to screen for offensive images. We’ll find ways to screen all impressions in real time in order to proactively block ad misplacements, rather than issuing a make-good and an apology later on. These are all great solutions, but they’re treating a symptom, not the disease.
Ad misplacements are a symptom of a macro-level problem. They say to a brand: You don’t have a strong understanding of where your messages appear, and limited transparency into your campaigns. They’re like the mother who trusts her kid is in school, but learns she has much bigger issues than missing math class.
In 2004, Stanford economist Paul Romer said, “A crisis is a terrible thing to waste.” Our industry has been responsible for a steady beat of egregious ad misplacements. Let’s recognize them for what they are: a continued lack of trust within the digital landscape altogether. They’re a warning bell that we need to be responsible and accountable. Heeding it will require us to break down walls, share data, and adopt newer technologies such as AI. That may make a lot of people uncomfortable in the short term, but it’s a necessary step in maturing the industry.