For the first time, spending on digital advertising is greater than television. To experienced advertising professionals, the news didn’t come as a surprise. But its impact is significant for myriad reasons: in 2017, While industry watchers saw this coming for years, it still signifies a monumental shift in the ad business. And all of the players in the ecosystem – brands, agencies and ad tech companies – are scrambling to evolve to accommodate new requirements.
It was with these challenges in mind, and recognizing that digital advertising is now more than 20 years old – IAB/PWC tracks spending to 1996 – that my agency, Anagram, launched in 2015. Our guiding principles from day one have been to focus strictly on programmatic buying. We balance both creative and data inputs in our approach, and embrace transparency. We define requirements based on a campaign’s goals; select the best channels / exchanges / systems for the job; and then execute / monitor / optimize the media buys.
The most significant differences between programmatic and traditional media buying are the reliance on systems and software, amount of performance data available, and an advertiser’s ability to make adjustments on-the-fly (i.e., “mid-flight”). While these options present advertisers with varied new opportunities for performance improvement, they also increase their potential risk exposure by way of wayward spending, fraud, and visibility.
In addition to our guiding principles, we have defined three complementary axioms that continue to govern our company’s direction and behaviors, which we hope will help others better navigate programmatic.
1. ‘Digital-First’ Is No Longer Optional
Based strictly on costs, implementation speed, and flexibility, a ‘digital-first’ approach is a no-brainer. When comparing TV, print and radio options to digital in each of these areas, digital comes out on top.
But that’s only part of the story. A genuine ‘digital-first’ strategic mindset requires significant cultural and business process implications as well. As corporate brands have worked furiously to embrace this approach, the legacy advertising agency business model has lagged to adjust.
What’s worked for us is that we’ve taken the valuable components of the old ad agency structure – talent and creative chops – and fuse them with new agency requirements (technology, data management, speed-to-market) to establish a stronger foundation. This includes a new way to manage staffing, methodology, and partnerships.
2. ‘Integrated Marketing’ is Easy to Say and Difficult to Realize
Big data, for all of its benefits, has also introduced a litany of challenges into the marketer’s day-to-day activities. While the media landscape has fragmented exponentially over the past decade, it has also created an overflow of data (which has resulted in introducing terms like ‘data lakes’ into the marketer’s lexicon) that requires management and analysis. Additionally, the “big three” issues that have been front-and-center in the digital ad business for several years now – viewability, fraud, and transparency – are now on the radars of non-marketing executives. So, marketers must be prepared to involve CEOs and even CEOs in big data discussions.
One way to tackle these “omni-channel” (another popular buzzword) challenges is through integrated analytics dashboards. But these require the incorporation of data from “walled gardens” like Facebook, Amazon, and Google. While difficult to acquire and organize, once wrangled, provide a vital component to solutions that help create an orderly view of the world, and simplify analysis and marketing mix allocations.
3. If You’re Not Confused, You’re Not Paying Attention
While market consolidation in the marketing technology is inevitable – according to one estimate there are more than 5,000 companies in the category – it’s not going to happen in a meaningful way for several years, or perhaps even a decade. For every mammoth, deep-pocketed concern like Adobe, Amazon, IBM, Microsoft, Oracle, Salesforce, SAP, there are hundreds of pesky (and often well-funded) startups flooding the market. It is critical not only to rely on a trusted suite of proven technologies to deliver client solutions, but also to constantly evaluate new market entrants, because this market moves fast.
Long-Term Keys to Success: Focus & Adaptability
I’m proud of the progress that Anagram has made in the few short years since our inception. But I know there’s still a lot of work to be done. Staying committed to the core plan, while at the same time being flexible to accommodate evolving marketing conditions, will be an ongoing challenge. But we believe this is how all agencies will need to operate in order to thrive in the new marketing environment. Buckle up!
About the author
Adam Cahill is the President of Digilant US and CEO of Anagram, a Boston-based programmatic agency. A 20-year veteran of the digital industry, he went allin on programmatic in 2009, launching one of the first agency trading desks. Before joining Digilant, Adam founded Anagram, a digital media consultancy built for the modern marketing era. Previously, he served as Chief Digital Officer at Hill Holliday and SVP/General Manager of Carat in Boston. Adam has been named a Media All-Star by Adweek, a Media Maven by the Ad Club, and has led teams that have twice been named Media Agency of the Year.
Last week, the Senate Intelligence Committee held a hearing to examine the extent to which foreign actors manipulated digital platforms to meddle in the 2016 elections. To their credit, Committee Chairman Richard Burr (R-NC) and Ranking Member Mark Warner (D-VA) side-stepped the partisan wrangling and focused on how these bad actors attempted to undermine our democracy. And, for the first time, Senators Burr and Warner insisted the top executives from the big tech platforms testify in person. Given the dysfunctional state of the American political discourse, it’s especially remarkable that a group of Senators agreed to tackle an issue in a bipartisan fashion. Facebook agreed to send COO Sheryl Sandberg (Facebook CEO Mark Zuckerberg already testified in the spring). Twitter sent its CEO, Jack Dorsey. Remarkably Google declined to send Alphabet CEO Larry Page.
It’s anyone’s guess as to why Google didn’t feel it needed to participate. Its dominant role in the distribution and monetization of content, through its search and advertising platforms respectively, certainly would have been relevant to the Committee’s work. And that is before you even get to YouTube, which helped make Russia Today its #2 news property through much of 2017 and spread of misinformation in Chemnitz last week. Google certainly has enough lobbyists on retainer to fully prepare. Whatever the reason, the fact that “declining” was even an option demonstrates just how big Google has become.
Data Wars
One area not well understood in Washington is the relationship between data collection and the incentives that allow foreign actors to manipulate them. Damian Collins, the Tory MP of UK Parliament, has led a brilliant multi-party effort to document its findings in this area. Other policy experts have also begun to explore how data protection may help curb disinformation.
Again, this is where Google’s role as the company that collects enormous amounts of data on the public needs to be more clearly explored. A few weeks ago, we distributed research entitled Google Data Collection, by Vanderbilt Professor Douglas Schmidt which, for the first time, cataloged the expansiveness of Google’s data collection machine. Google owns the #1 browser (Chrome), the #1 mobile platform (Android), and the #1 search engine (Google Search). On top of that, Google’s analytics and advertising tools are impossible for consumers or publishers to avoid. All of these touchpoints allow Google to conduct a massive surveillance operation on consumers who have no way to avoid it.
Professor Schmidt’s research shows that a dormant, stationary Android phone (with the Chrome browser active in the background) communicated location information to Google 340 times during a 24-hour period. That’s an average of 14 data communications per hour – with location information constituting 35% of all the data samples sent to Google.
When an iOS device was running Safari in the background, but not Chrome, Google could not collect any appreciable data unless a user was interacting with the device. Conversely, an idle Android phone running the Chrome browser sends back to Google nearly 50 times as many data requests per hour as an idle iOS phone running Safari.
Perhaps the most striking revelation in Professor Schmidt’s research is that approximately two-thirds of the data that Google collects about consumers is collected in passive ways. That’s an incredible amount of information being gathered in the background, where consumers wouldn’t expect it.
Power Play
Google’s pervasive surveillance have enabled the company to rake in massive profits and squash competition in a variety of ways. However, there probably is no surer sign of their dominant position than their ability to decline – with no worry for the consequences – a bipartisan request to testify from the Senate Intelligence Committee.
At the same time they declined to participate in this hearing, rumors began to circulate that Google is creating a search engine just for the Chinese market that will allow Communist leaders to block access to content they don’t like. Google appears to be a willing participant in the suppression of over 1 billion people.
Is it time to start asking whether Google has outgrown democracy?
For more than a decade now, users of social media platforms and internet services have been providing personal data to companies like Facebook, Google, and Yahoo! in exchange for free services like online communities, email, and storage space. The value proposition is straightforward: We give the companies some personal information – name, birthdate, and where we live – and permission to track our activity when using said services, and they let us use their software. In turn, this data drives advertising models. It’s a bit more complicated than that, but not much.
However, several high profile events over the past couple of years – some directly related to web services, others on the periphery – have presented potential threats to this seemingly too-good-to-be true business model. Take your pick of which event might have raised your ire, as there’s something for everybody: Cambridge Analytica’s hacking of 50+ million Facebook accounts to support its efforts during the 2016 U.S. presidential elections (followed by the hacking of its entire user population); Yahoo’s admission in late 2017 of a massive breach of its 3 billion user base; or the Equifax compromise that affected 300+ million accounts, which the company is still trying to sort out, a year after its announcement. These followed other newsworthy problems in the mid-2010s involving household brands like eBay and Target, which themselves followed the TJX credit card breach, which involved about 45 million records.
Though, of course, 0we can’t predict the next meaningful breach, the past decade or so tells us that we’ll experience a least a couple each year for the foreseeable future. Hackers are an ingenuitive bunch, apparently. The motivation behind the efforts range from the acquisition of political power to greed and just plain hooliganism. For advertisers, though, the big question is: Do consumers care enough about these threats to change their behavior?
Is There a Shift In Consumer Behavior Underway?
According to new research from Pew Research Center, we could be seeing signs of shifting consumer behavior. According to its survey from June, more than half of respondents said they’d adjusted their Facebook privacy settings over the previous year, while 4-in-10 said they’d taken a break of “several weeks or more” from the app, and a not insignificant 25% said they’d deleted the app from their cellphones.
As a snapshot, these results are noteworthy, but not shocking. (And it’s too early to tell if they signal a larger trend.) Facebook seems to be approaching market saturation in the U.S., while the coveted teen demographic is using Instagram and Snapchat more frequently. It stands to reason that more experienced users are likely to tinker with privacy settings, as well as occasional bouts of abstinence.
(Weep not for Facebook or proclaim its impending demise. In the last reported quarter, it added an average of 22 million new users every day.)
It Probably Doesn’t Matter Much For Advertisers, At Least Not In The Short Term
While we may be seeing signs of Facebook user behavior evolution, most constraints on consumer data availability in the short term will be driven primarily by government regulations or self-policing actions by Facebook/Google/Yahoo/Verizon. The EU’s General Data Protection Regulation (GDPR) law enacted over the summer is the most obvious example of government involvement, which caused genuine angst among American companies, and instigated modifications and new ideas. It doesn’t seem like U.S. lawmakers will follow suit in the foreseeable future. This is due in equal parts to a lack of understanding of the issues, an administration that espouses deregulation, and attention to a litany of more pressing issues.
More practically, most companies outside of the largest 100 U.S. advertisers or so aren’t pressing for more consumer data. Indeed, they aren’t able to convert much of the existing available data to actionable marketing strategies. Weaponizing big data requires major investments in systems and resources that are beyond the reach of small and mid-sized companies. (And the scope of data currently being captured is genuinely mind-boggling.)
The Bottom Line
Given that brands have more consumer data available to them than most can practically deploy for advertising purposes, changes in consumer behavior in response to data breaches like the Facebook-Cambridge Analytica kerfuffle are unlikely to meaningfully impact marketing strategies for the foreseeable future. However, while consumers have historically demonstrated an extraordinary threshold for what they are willing to provide in exchange for free goods and services – and this premise is effectively the lifeblood of companies like Google and Facebook – the combination of high profile breaches, Congressional hearings, and product fatigue could prove to be problematic for the category over the longer term, and it’s certainly something we’ll be watching closely.
Publishers keep hearing about monetizing their data to drive revenue, but there are many ways to achieve this. Selling data, whether through a data exchange or a private marketplace, is perhaps the most obvious option to increase revenue. But if you’re already taking advantage or not interested in that option, selling your data isn’t the only way to make money with your customer data. In fact, there are many creative ways in which publishers can monetize their data. Here are six opportunities to monetize data – without selling it.
1. Sell Audiences, Not Site-Sections
When you’re collecting data on your readers, you can use that data to make their audiences more sellable to advertisers. By collecting and organizing audience data from across various sites and properties, you can sell audiences instead of pages or site sections.
Say, for instance, you have an advertiser who sells kitchenware. The first place you would think to put those ads would be in the recipes section of your site, since the people who visit that section most likely enjoy cooking. With a data management platform (DMP), however, you could create an audience of users who have a lot of interest in cooking and kitchenware. The you can display those kitchenware ads across your entire site aimed specifically at those users, rather than just in your recipes section.
Since your DMP allows you to access information about your audience’s interests, you could sell ad space to a kitchenware company even if your site has nothing to with cooking, appliances or home goods. A DMP gives you insights into your audience’s interests beyond your specific site, which provides you with the opportunity to work with advertisers you never could before.
2. Increase Average CPMs
If you’re focused on selling ads, increasing the price of each one will obviously help increase your revenue. But how can you do this? Offering advertisers the ability to target their ads to the exact audience they are trying to reach means they will be willing to pay more to reach that engaged audience. Better-targeted audiences are more sellable. When advertisers know they are reaching an engaged audience on your site, they’ll pay more to have their ads there.
3. Expand Your Overall Audience
If you’ve got large advertisers signed on for big targeted campaigns, you need to be able to fulfill those campaign targets. If you’re lacking enough uniques to make the campaign a success, 2nd or 3rd party data may help you expand your targets.
It might seem counterintuitive to pay for data to make money. However, if you’ve got gaps in the data available, you may need to spend more upfront to be able to attract more visitors to your site. More visitors means more impressions, which means more people who see the ads.
4. Get More Subscribers
For subscription-based publishers, reducing churn is of utmost concern. And the key to keeping subscribers engaged is serving the content they want. You can use your DMP to access insights into the characteristics of the people who subscribe to your site. You might find that your regular readers have various interests in common, beyond what you would expect based on the content of your publication. Perhaps, for instance, you run an online golf magazine, so you know your readers are interested in golf. Your data might show you that large portions of your audience are also interested in fine wine, bowling or apparel. By offering content that fits your readers’ interests, you can keep them engaged and coming back for more. Alternatively, you can essentially become a marketer yourself by promoting your content to people who are most likely to engage. From the previous golf example, this might include targeting golf fans on wine sites or pages to bring them back to your pages.
5. Reducing Remnant Inventory
Ideally, you’re already selling all of your ad space as premium inventory. However, that often doesn’t happen. It’s possible, though, to convert more of your remnant inventory to premium, which leads to a substantial increase in ad revenue. As the value of your ad space, and the value of your website as a whole, go up, your inventory will sell faster and for high prices. As a result, more of it will be considered premium.
Data is what drives this increase in value. With data, you can sell audiences rather than sections of your site and reach advertisers you wouldn’t otherwise sell to without audience data. This boosts the overall value of your site to advertisers and puts your ad space in higher demand, making it easier to sell ad space. As a result, you’re able to sell more space as premium rather than remnant.
6. Customizing Content
Your editorial team probably has a pretty good grasp on the type of content that performs best on your sites. And as the old saying goes: If it ain’t broke, don’t fix it. But if you’re looking to keep visitors on your site longer, you’ll need to offer them content recommendations that match their interests, so that they can view more content and see more ads.
Once again, the data can guide you by revealing what kinds of content are performing the best with your audience. Your data will tell what a particular reader’s interests are, allowing you to show them content that deals with those interests. If they read a lot about fashion, for instance, you would serve them content related to fashion.
Data is powerful. And, if used correctly, it can help you increase your revenue, even without selling it directly. They key is understanding who your readers are, so you can serve them appropriately. Data is a powerful tool. Ensure that you are using it in every way possible to strengthen your revenue streams and improve your site overall.
In just a few years, the number of publishers building out branded content has soared from 15 to more than 600—and counting. But it’s not just their ranks that have grown. Many publishers have expanded their capabilities, harnessing teams, talent and tech to help brands create sponsored content and capture audiences. The New York Times has gone one better, establishing T Brand, a complete content studio, to create unconventional immersive and artistic projects that go far beyond native advertising.
Through the strategic acquisitions of Hello Society, an influencer marketing agency, and Fake Love, a design-driven agency specialized in one-of-a-kind live experiences, T Brand has also nurtured new expertise in video, 360-degree filming, augmented reality, and virtual reality. Here, Peggy Anne Salz – mobile analyst and Content Marketing Strategist at MobileGroove – catches up with Graham McDonnell, International Creative Director for T Brand Studio, to discuss the company’s mission to create dynamic and innovative content with a decidedly human touch.
PAS: T Brand is leveraging its tech abilities and acquisitions, but it also benefits from audience trust in The New York Times. How do you make the most of both without blurring the lines between content you create for your brand partners and editorial content?
GM: There is quite a clear divide between the newsroom and advertising; it’s very much church and state. We obviously can’t use journalists from the newsroom. However, the studio is staffed with fully qualified journalists, many of whom have come from our competitors. So, we have the strength of a newsroom, in integrity, and talent, and we leverage this for our advertisers. We also maintain a high standard for our audience. After all, the audience coming to The New York Times expects New York Times journalism and content. So, it’s expected that our branded content should hit those heights as well.
It’s also very important to identify branded content as a piece of advertising content. As soon as an audience feels duped or tricked into reading branded content it will lose trust in the publisher and—ultimately—in the brand. Our most common branded content pieces are labeled “paid posts.” Traditionally, that post was a destination. So, it was a URL that readers could visits via our website or via our app—content that was paid, posted and lived indefinitely. The aim was to make that content accessible to audiences. We still make branded content a destination. But we also make it a journey because getting eyeballs on a page isn’t enough. We’ve started to add what we call “branded footers” to each piece of sponsored content we produce. This way, when a user is finished consuming a piece of content, we offer the opportunity to continue that conversation outside our paid-post environment.
PAS: Branded content is a major revenue driver for publishers, but not all publishers have been able to get brands to buy in for the longer term. How can publishers make that connection?
GM: Storytelling is what we as humans do and relate to—going back to cave paintings and hieroglyphics. Storytelling is an intrinsic part of our human nature and the best branded content taps into that. The best stories come when publishers partner with the client and dig deep into the story the brand wants to tell. You want to nail the narrative first, and you want to think about the execution second. It’s important to tell a human story, one that’s relatable for the audience, and not get ahead of yourself by trying to figure out if VR, for example, might be the best way to tell the story. This is a point many brand marketers forget and it’s up to the publisher to remind them. No one knows their business better than our clients, but no one knows our audience better than us. Put the two together, and that’s when you get the best—and most sustainable—results.
It’s all about the ability to marry the storytelling expertise of the in-house team with the advertiser’s goal to address its audience and express its point of view. So, brands come to us with their brand message and they look to us to position that message within a story that will resonate with our audience. Of course, brands will push hard to put their message first. It’s up to the publisher to push back and put the storytelling first.
PAS: Does it requires some “tough love” to strike a balance between the two?
GM: Yes, and that’s what will drive results. For example, some clients come to us and say they want AR or VR. They want all the bells and whistles and all the flashy toys. But we tell them it’s much more important to think about the story first and then how to tell it. Even in the word ‘storytelling,’ the story comes before the telling. This means taking a step back to think about the type of story your audience will want to engage with, not the tech you’ll need to tell it.
We were one of the first content studios. So, over time, we have found what works, and what doesn’t. Over time we’ve seen that we have much more time spent on our content than on the content given to us by brands. Sure, we get brands coming to us saying, you know, “We have got this video of our CEO. He’s very engaging. It’s 32 minutes long. We just want to put it and you don’t need to do anything with it. Just put it on and people will love it.” My response is, “No, that might not be the best idea.” You have to put the brand message in a story people will like – one that will take them on a journey not blast them with details and stats.
PAS:What is the best way to tell brand stories?
GM: There’s a simple sort of formula that all good stories follow called the narrative arc. The first thing you do is introduce an element, usually a character, that the audience likes and is emotionally invested in enough to care what happens to this person. Then you present a problem or a hurdle, some sort of challenge that must be overcome. Finally, you reveal the outcome, some kind of goal or reward. The brand message is an integral part of the story, but it shouldn’t be too obvious. Just like parents who convince their children to eat vegetables by hiding it the food kids are sure to like, branded content blends the brand story within the story—because these are the things you want your audience to digest. Making the brand message part of the storytelling makes it much more palatable. Once you’ve got the story, then you think about the execution.
The biggest trap content marketers fall into is trying to tick all the boxes. Sometimes they are pushed by their brand clients, and other times they are pushed by their own ambitions. They put together a package of four videos, three infographics and loads of cool stuff for impact across every channel. Without a strong narrative to link each piece to the next the outcome is a Frankenstein monster of fragmented content. We have found it’s more effective to limit yourself to telling very focused stories.
PAS:It’s clear that branded content has to be emotive, but it must also be effective. How do accomplish and measure this?
GM: The best results come when you have a deep partnership with the advertiser and a deep understanding of how audiences engage with content. Knowing the time of day people are consuming content on their devices is an important data point. It’s part of a larger, much more data-driven approach to know when to serve the right content at the right time. Targeting context increases engagement. If you have created a data-heavy infographic that is best consumed on the desktop or tablet, then there’s really no point in serving that during commuting times when most users will be viewing their mobile feeds and devices.
We’ve also found that around three-quarters of our programs have dwell times above the Moat benchmarks for audience attention. So, not only are we getting people to visit our content; we know they are staying to engage with it. Moreover, a vast majority of our programs surpass the Moat benchmark for scroll depth. This means they are scrolling down and exploring the content. We have succeeded in building a narrative arc that offers a reward well worth the audience’s time and attention. You’ve really done your job if you can bring rather drab content to life and a great example of this is when a client wanted us to help promote a white paper. We decided to do it in a quiz format that would draw the reader into the content. We had seven questions and after every question they asked, we gave them a little snippet of content, a statistic related to the answer they just gave. It was a reward scheme, and it worked—showing that it’s a very underused tactic but effective tactic to keep people engaged.
PAS: The international arm of the T Brand Studio is perhaps best known for its award-winning campaign for UBS, highlighting AI and what it takes to be human. The native advertising included a chat bot, a five-chapter article and a documentary-style video, surpassing target reach and engagement metrics. Another more recent campaign for Kia brought the Cadenza model to life in a series of live events. Should traditional creative advertising agencies feel nervous?
GM: As I said earlier, the best content comes from partnership. We don’t usually offer our services like an off-the-shelf product; we build a relationship with the client to tell a story in the way our New York Times audience expects it to be told. It’s not a case of jumping in on every brainstorm or every RFP; it’s a case of looking at what the client wants to achieve and answering the brief with a strong journalistic approach because that’s our strength. A lot of the time we’ll partner with other agencies when we create content for brands. We’ve also worked in tandem with other publishers on certain program to suit the needs of the client. The industry is moving fast and getting faster. Therefore, it’s important we all learn from each other, not fight against each other for first place. The bar for branded content is high—and so are audience expectations.
In May of 2017, Bill Simmons, former ESPN personality and creator of The Ringer, made an interesting decision to move his successful sports media empire to Vox Media. Simmons maintained editorial independence but relied on Vox for all the non-content creation requirements of a traditional publisher. This included all revenue and product development because as Simmons said “they are great at sales and technology.” The plan, it seems, was to offload the “publishing” part of the publishing business. This allowed Simmons to focus on content creation for his widely popular podcast and website.
This wasn’t a viable model 20 years ago when publishers had established oligopolies by controlling distribution in their respective geographic regions. As a result, the publishing needs of the business, HR, printing press, sales, and distribution existed to support editorial and content creation. Now that the internet has driven distribution costs to near $0, and oligopolies no longer exists, there is little need to combine traditional publishing sales, product and HR overhead with the journalism it is intended to support.
Even so, content creations like Simmons still need to make money. To do so, they rely on content management platforms like Chorus, the purpose-built platform Vox created to manage content and revenue. As of July, Chorus supported 350 media brands. The system includes “content management, data-informed multi-platform content distribution, integrated advertising, and a suite of publishing tools.” Media brands that strip away activities not directly related to content creation establish a sustainable model that allows them to create even more great content, a reinforcing cycle of success.
In his May 2017 post, Ben Thompson covered the changing business model while referring to platforms like Vox’s as the “Faceless Publisher.” The platform handles all the monetization, resources, and infrastructure, which allows journalists to focus on creating amazing content.
The term “Faceless Publisher” might be a misnomer in Vox’s case given they have established several great brands in the market including SB Nation, Vox and Eater. However, it is clear that this federated group of journalistic brands, supported by a single business platform, creates great efficiencies. This efficiency will deliver what readers and content creators need: a sustainable business model that delivers amazing content.
As of May 25, 2018, Google announced that DoubleClick users will be unable to rely on cookies or mobile device IDs to connect impressions, clicks and site activities from DoubleClick logs. Instead, they will be limited to Google’s own Ads Data Hub for those metrics.
For some, this means that they are satisfied to stay within the Google stack. But not every brand’s solution will be and should be limited to Google. If media buyers want to analyze their spend outside of Google’s platform and offer up any attribution, then just using Google won’t work.
“Some marketers who spend 75 percent or more of their budgets on Google will be fine just letting Google do the analytics,” saysAlice Sylvester of Sequent Partners.
Google wasn’t the only one to lock down its platform. In response to the combined pressure of GDPR and theCambridge Analytica scandals (related to handling of personal information), Facebook decided that it would shut down ad tools called “Partner Categories” powered by outside data brokers. Those tools let Facebook advertisers target ads at people based on third-party data such as their offline purchasing history.
This means advertisers will have access only to their own data, and data Facebook collects itself. If an advertiser wants to pull campaign-level insights to inform future campaigns, or use the data for the basis of an attribution model, then they are out of luck.
Introduction of Data Clean Rooms
Data clean rooms allow large inventory partners like Facebook and Google to share customer information with brands, while still maintaining strict controls. Data clean rooms were named for the completely airtight rooms where microchips and other sensitive materials get made. In this case, the rooms enable a shared environment between two or more companies that are completely secure from external access (no wifi), and where each company decides the level of visibility to their data. This eliminates – or severely restricts – the possibility of data leakage (which is what happened with Facebook and Cambridge Analytica).
“We and a partner combine a data set with very specific rules and controls around how each party can operate within the shared environment,” said Scott Shapiro, a product marketing director for measurement at Facebook, who noted thatFacebook didn’t invent the clean-room concept.
The driving force behind the concept is to create a safe space where data can be shared and manipulated without leaving the inventory partner’s environment. Specifically for Facebook, a brand can create anaudience based on first-party data – like a list of email addresses – and then push that list into Facebook, match it, and grab a copy which they can later combine with their data as the basis for attribution, measurement, and modeling.
How it happens in reality is that an advertiser will load a clean or wiped laptop or device that has never been connected to the Internet with that advertiser’s first party data, which in most cases is an email list. A second clean computer is loaded by Facebook or Google with impression-level and non-personally identifiable information (“PII”) campaign data.
Maybe, The Answer to Scaling The Walled Gardens?
For advertisers with reams of data and substantial programmatic advertising budgets, this is a great opportunity to scale the otherwise elusive walled gardens. Data clean rooms create a safe environment for data providers to share marketing information that brands need and crave to model future media buys and advertising strategies. If managed properly, with appropriate methods and standards, this technique would allow brands to really understand their walled-garden ad spends within the larger marketing ecosystem. For both advertisers and publishers alike, the stakes are high in the post-GDPR world of data governance, and there is no room for unintended data sharing because consequences are severe.
Marketers have been eager to get more insights out of Facebook and other walled gardens, but it remains to be seen how aggressively brands and agencies will use data clean rooms to make the most of the spending with the largest inventory providers (e.g., Google, Facebook).
There are two prevailing views for what the future holds:
Glass half-empty: These same inventory providers lack a compelling incentive to play well with others in clean rooms, beyond delivering another level of customer service in a marketplace they continue to dominate.
Glass half-full: It’s been a daunting year or more for the industry category, with virtually continuous coverage related to privacy violations, bad actors influencing politics, and fraud and transparency challenges. The ‘clean room’ concept may be a half-step that the duopoly can get behind, if only as a signal of good faith to the industry.
There is also the overriding issue of what kind of manpower (likely significant) would be involved to make the clean room option a viable reality. The usefulness of data in this kind of an environment may also be somewhat limited No matter, though: The data clean room concept is one that’s getting some attention. And, considering its appeal among the dearth of options out there that seem appealing to all the affected players – brands, agencies and inventory providers – it could be one that ends up getting traction.
About the author
Karen Moked is the Vice President of Marketing at Digilant, a programmatic media company in Boston. A veteran of the advertising and technology industries, she previously worked for Akamai and O’Reilly Media. Karen is a graduate of MBA-ESG in Paris and York University in Toronto. Connect with her on LinkedIn and Twitter.