Building Brand Safety and Trust
Note: This post originally appeared on the Rubicon Project website.
Brand is the invaluable intangible. A positive brand experience can motivate impulse buys and generate fierce brand loyalty, while a negative one can jeopardize lifelong customer relationships.
It’s no wonder many marketers feel anxious about programmatic buys, considering they’re relying on soulless machines and sometimes-opaque algorithms to determine where their brand’s advertising will be showcased. One appearance next to the wrong content can bring a PR disaster, negative associations, and damage to the bottom line. Marketers have even told me they worry that placement on a site with a toxic reputation could cause them to lose their jobs, and put their children’s college education in jeopardy.
At Rubicon Project, marketplace quality and brand safety has always been part of our own brand promise. We’ve built trust with our brand and agency partners who rely on us to provide the highest quality inventory and strong brand protection capabilities. How do we do it?
High quality impressions, nothing less
We look at inventory quality holistically, with two major components. There’s a platform-wide component – standards that we apply to everything we allow into the marketplace – to protect against non-human traffic, unauthorized inventory, and generally bad experiences. And there’s a buyer-specific component, supporting the varying standards and expectations that different brands or agencies have around things like viewability and contextual brand safety. Our foundation of transparency is key to delivering brand safety on both fronts.
We have a dedicated IQ team with deep experience in applying those guidelines and making the tough decisions needed to protect our marketplace — backed by our Brand Protection & Security engineers and product managers that support these efforts.
All publishers are not created equal
Our publisher review process is central to our inventory quality and brand safety protections. Every new seller account considered for a seat on Rubicon Project goes through a thorough vetting process by our IQ team to make sure it’s appropriate for our marketplace. That includes human review of the individual sites and apps proposed for monetization.
After a seller is first onboarded, incremental sites, apps and new formats also get vetted for compliance with our IQ guidelines. Our processes are designed to validate that inventory is legitimate, authorized, declared properly, and doesn’t show signs of non-human traffic, with appropriate formats and ad density. We also check for basic brand safety guidelines: no porn, piracy, hate speech, or other questionable content. We use an internal system to tag all the inventory and then monitor it on an ongoing basis.
Stamping out ad fraud
Once we have a commercial relationship with a seller, our platform applies pre-impression filtering to all ad requests that enter the marketplace. In 2016 we acquired a specialty anti-ad fraud vendor called Audience Forensics that helps identify and investigate potential anomalies. Using a combination of proprietary and third-party data and tech, we filter out traffic that looks like bots, or looks like it’s being spoofed. The traffic we filter out never goes to auction.
We also use post-impression monitoring, firing a third-party measurement sensor on a high proportion of our inventory to check how well the pre-impression filtering is working, and to keep overall non-human traffic levels low. (Even the most reputable publishers have some amount of non-human traffic, and no pre-impression filtering system is perfect.) We identify hot spots that might need a closer look from our team, and feed those findings back upstream so that our blocking and filtering process keeps improving.
In cases where a buyer identifies traffic believed to be invalid, we’re committed to a serious investigation of that claim. It’s our firm belief that buyers shouldn’t pay for bad traffic, sellers shouldn’t benefit from it, and we as an exchange shouldn’t take fees on it. Only by aligning economic incentives across the ecosystem can we drive out the bad actors.
Ads.txt: on the front line of brand safety
As a premium exchange, we only want to work with authorized traffic. That’s why we’re very excited about Ads.txt: it provides a lightweight, industry-standard way for sellers to declare exactly who is authorized to sell their inventory.
Third-party analysis shows the Rubicon Project is tied for #2 industry-wide in terms of ads.txt coverage. We’ve already seen great adoption by our sellers, and we are continuing to inform and educate sellers about how Ads.txt protects their ability to monetize.
Many DSPs are already blocking unauthorized inventory, and we’re working to block it automatically on our exchange – while we’re also planning for a time when all buyers will decline buying on undeclared inventory (that is, sites that lack an ads.txt file).
Contextual brand safety and viewability
So far, we’ve talked about our platform-wide efforts to protect brand safety, but individual agencies and brands have their own expectations in certain areas — particularly when it comes to contextual brand safety (such as the specific content on an individual page) and viewability.
Because standards for measurement vendors and thresholds vary, our approach is to work with individual buyers to demonstrate how our inventory can meet their specific needs, including the creation of curated private marketplaces when that’s appropriate. We’re even doing this for in-app inventory, where the measurement challenges today are far greater than on the web.
With our long history and team committed to marketplace quality and brand safety, the Rubicon Project exchange gives buyers the ability to transact confidently on high-quality, high-reach inventory. If you’ve got questions about how we’re protecting your brand, don’t hesitate to reach out.