"Any kind of algorithmically driven advertising is likely to show up in potentially non-brand-safe places."

Scale vs. Brand Safety: Is There a Programmatic Winner?

The world of programmatic advertising is in a tizzy. Apparently, some brands just discovered that there are seedy parts of the internet with which they don’t wish to be associated. From Chase contracting display advertising, to Kellogg (and others) boycotting Breitbart, to a YouTube shutdown for many advertisers, there is a surge of awareness around programmatic advertising and where it shows up. I don’t know about you, but this isn’t really news to me. Any kind of algorithmically driven advertising is likely to show up in potentially non-brand safe places. 300 hours of content is uploaded to YouTube every minute. The explosion of inventory increases at an amazing rate. There is no way to police that easily or completely. All solutions, from whitelists to automatic content screening will not be 100% perfect. What is a brand to do?

There are two approaches that brands can take with this issue:
1. Manually Review & Whitelist
2. Trust But Verify

What is Programmatic Advertising?
Programmatic advertising is a mix of content-aware advertising and user-targeted advertising that happens all over the internet. Rarely is there a direct relationship between the advertiser and the site on which the ad shows. For programmatic advertising, the site isn’t really important—it is the visitor to the site that is the target.

There is an amazing Rube Goldberg data machine that gets kicked into action when a page that shows programmatic advertising loads. There are user IDs, propensity to click, propensity to buy, frequency measurements, browsing data, and more factors get loaded up and a decision is made in under 20 milliseconds to enter the auction and show an ad. The ad is sent to the ad server within 50 milliseconds. It is amazingly complex, and happens in less time than it takes to blink your eye.

Two Approaches to Programmatic Brand Safety

Manually Review & Curate Whitelist

What This Means: Manually review website and video content for brand safety.

Pros: This is safe. You pre-select the content and sites that are part of your ad network. This works.

Cons: The internet is a very big place, and with literally millions of sites and videos being visited by your target audience every day, there is no way to effectively police that scale. It is unlikely that you have enough staff to wade through and cull out more than a handful of your top-sites. This limits your scale (you are making the internet a smaller, safer place for you and your brand) and potentially lets your competitors target your customers directly.

Trust But Verify

What This Means: Use all available ad-network safeguards. Review implementation of these safeguards with your agency or network owner. Then spot-check your ad placements from your reporting. (Check out a handful of sites/content that you don’t recognize, and evaluate.) Further, you can work with a 3rd party, like Integral Ad Sciences or DoubleVerify as an overlay on top of network exclusions and whitelists.

Pros: This maintains your scale and gives you the best opportunity to show an ad to your desired client, all while using standard tools to keep your brand safe.

Cons: This is the status quo. This is trusting the network and 3rd parties to police content, and augmenting their safeguards with a little elbow grease. This is how many brands approach the issue today.

Really, That’s It?

For most advertisers, making programmatic fully brand safe requires tradeoffs. If you manually restrict your network to pre-approved sites, you cap your reach and potentially forego showing your message to valuable existing and potential customers. Further, you have to invest in human staff to maintain and expand the white list. In so many ways, this defeats the intent of programmatic advertising, the main premise of which is reaching your desired audience wherever online they may be. If you trust in the network to do the whitelist and content vetting, you run the risk of your ads being shown in places where the chairman of the board (or your mom) would likely not approve.

Programmatic advertising has changed display advertising from its 2005-era remainder-bin status to a powerhouse of activation and retention.

So how does the smart brand decide between these two tradeoffs? That is a hard question to answer because programmatic advertising is, ostensibly, target-focused. The reason why your ad shows up somewhere is because the algorithm that decides targeting and bidding has said that this potential ad impression is worth showing. There is potential performance upside to this impression.

As advertisers and agencies, we have made the decision to trust in the algorithm. We trust the quality of our 1st party data, appended with the 3rd party data that the programmatic eco-system appends to it to create the model of success. In the name of efficiency and scale, we’ve outsourced some decision-making to mathematical models of probability. And, if we look at the performance over time, programmatic advertising has changed display advertising from its 2005-era remainder-bin status to a powerhouse of activation and retention. So, programmatic works.

But for some brands (and maybe all brands on some level), this efficiency potentially comes at a very high cost. For some brands, this is a true moral crisis. Can they willingly spend advertising dollars when they know that some of those dollars, regardless how few, are benefiting organizations and ideas that they find to be repugnant and reprehensible? For other brands, it is less about moral clarity, and more about preservation. Namely, would their desired demographic be offended or think less of their brand if they knew that this brand’s ad was shown alongside certain content? And finally, for some, this is about efficiency. Can they afford, from an opportunity or financial perspective, to not advertise in the most cost-effective way?

At its core, programmatic advertising is about the audience and not the content. From a strict advertiser’s point of view, the content exists to create an audience to whom you wish to serve ads. The advertiser is advertising to an audience of one – the singular website visitor—irrespective of the content. And, as a programmatic advertiser, that is your fundamental focus, identifying and advertising to a one-person audience. An advertiser could easily absolve themselves of any responsibility by understanding that the reason their ad is being shown against objectionable content is that their audience, regardless of size, has traveled to that content. The audience leads the advertiser to the content, not the other way around.

Programmatic Ad Placement Is Not A New Issue

Whitelists, brand safety, and advertiser alignment with content has been part of the advertising world since newspapers and magazines started selling remainder space. In some ways, part of the exchange for advertising efficiency through programmatic advertising is a loss of control. Advertisers are ceding advertising context to capture their desired audience. When Google introduced their content network in 2003, the issue of content and context vs. audience started to rear its ugly head. We have been fighting this battle for almost 15 years.

So why is it a hot-button issue today?

Much of it has to do with the mammoth explosion of content. There is so much content that is being produced at scale in order to further political agendas, to raise awareness for sensitive issues, to promote a world of disparate ideas that is both terrifying and awe-inspiring. We also live in an algorithmic, information-bubble world, where our consumer-behaviors are inextricably and irrevocably mixed with our personal exploration into the infinitely complex world of ideas, agendas and opinions that exist just a finger-tap or mouse click away.

Advertisers have chased this algorithmic world with ferocity because it fulfills the idea that we can deliver the right message to the right opportunity at the right time and create a mutually satisfying advertising cycle. And, truly, we can see that it works.

What Does The Future Hold for Programmatic Display Advertising?

There are solutions using AI and scoring methods, like Omnicom is building, while others like WPP use 3rd party solutions. There are other services that add a layer of brand safety to display, like Integral Ad Sciences, and earlier this year, DoubleVerify extended its solution to include YouTube. These kinds of solutions will start to bubble up new ideas and techniques that will be adopted and improved upon by ad networks and content aggregators (namely YouTube). So the technology-enabled content monitoring will continue to advance over time. We think that will happen rapidly. (When there is financial threat, namely the loss of advertising dollars, technology seems to rally quickly.) We are at an inflection point where the scale of content has likely outstripped the ways we currently monitor that content for safety. As a reaction, some brands may choose to pull back on programmatic and deploy on only curated sites. This will lead to a rise in traffic and impression costs on those curated sites, because many large advertisers will gravitate towards similar sites and content providers. This will push bigger dollars into a smaller advertising pool, potentially restricting the scale of reach and spend. Advertisers won’t revel in the smaller reach, and networks will chafe against a decline in ad revenue. Technology in the long run, will evolve to a point of stasis, where it may not entirely solve the problem, but it will likely keep it within acceptable error bounds.

So, What Should Your Brand Do?

Let’s bucket you into groups, ok?
1. Moral Imperative: For those of you for whom this is truly a moral issue, a curated whitelist is the only solution. That comes with tradeoffs in financial efficiencies and reach. But the moral imperative is clearly the guide.

2. Image-Conscious Brand: For those of you who are managing your brand image against programmatic efficiencies, it is likely that a combination of curated whitelist and trust but verify makes sense. Starting with an initial large cull of your placements, work with your networks, third parties (like DoubleVerify, etc.), and agencies to over-weight spend towards your best performing placements. You should mitigate brand risk, while maintaining the majority of programmatic efficiencies.

3. Opportunity & Financially Focused: This issue probably doesn’t enter into your thinking a lot, not because it isn’t important, but rather your true north is around efficiency and performance. For you, a trust but verify approach is the right one. Trust in the algorithms to drive your efficiency and don’t put too much subjectivity into the math.

There is no right answer. There is a series of tradeoffs that happen in scale and efficiency when you meddle too much with the algorithmic math that has driven programmatic advertising to dizzying heights. Technology will improve your brand safety over time, and your targeting will sharpen up so that your look-a-like audiences will become higher performing and you can stay targeted on quality content. Sadly, those improvements aren’t available today. But if waiting for networks and algorithms isn’t acceptable, you have to make the decision to put in a huge amount of human effort to curate the internet and suffer potential losses in scale and efficiency. How you make these decisions involves you looking carefully into the guiding principles of your brand, your staff, and your leadership team. That is where the answer lies, in your perception of your brand and your place in the world and how you honor that perception.

Want to learn more about programmatic? Our POV “Inside Programmatic: Agency vs. In-house Management” provides an insightful breakdown to further your understanding on the topic.

UPDATE: On April 6th, YouTube announced that it was restricting advertising on channels that have less than 10,000 aggregate video views. This is a big boon to advertisers who need the reach and efficiency of programmatic advertising, but are concerned about brand safety. Above, Manual Review & Trust But Verify were the two approaches that brands could take to help insure brand safety.

By excluding smaller channels from the YouTube Partner program, and having a review process before 10K+ channels can become part of the program, YouTube has given brands an added measure of safety. Effectively, YouTube has crowd-sourced their initial approval to the Partner Program, trusting that the wisdom of 10,000 aggregate views on a channel will ferret out the most egregious content. This is an enormous cull of YouTube channels, so both the Manual Review & Trust But Verify approaches received a big upgrade.

Tim Kilroy
About the Author

Tim Kilroy

Tim has been working in online marketing since the days of "buying a keyword" on AOL was a good strategy. Along the way, he has started and sold a couple of companies, been an IR Top 500 CMO, helped grow a home goods retailer from big to huge, and now happily hangs his hat at Elite SEM.

More Posts
Blog Categories
Search Engine OptimizationPaid SocialPerformance DisplayIndustry NewsFunPaid SearchShopping & FeedConversion Rate OptimizationConferenceCompany UpdatesInterviewPressUncategorizedEvents
Blog Archives
2017 2016 2015 2014 2013 2012 2011