How Ghost Ads Can Help Prove the Incremental Revenue of Campaign Spend

It’s not enough for an ad to simply trigger a sale: Today’s marketers want to know which ad it was, and on which platform and network it was featured. Ghost ads are emerging as one of the most popular measurement solutions to monitor ad efficacy and efficiency.

These types of ads, introduced by Google in 2015, are now receiving more recognition by the industry at large for their use on Google and Facebook. As Google notes, “It’s difficult to know whether that marketing expenditure is making an impact . . . The pressure to prove value is all the more important because marketers know the budget for one campaign can often determine the investment for their next one.”

But what are ghost ads? And can brands really benefit from their use? Let’s take a look at how this form of advertising is being used by publishers today.

The Potential to Reach More Consumers

Ghost ads are essentially conversion lift tests which, as Shane O’Neill, director of content at Nanigans, explains, monitor a control group of target consumers and flag when a brand’s ad would have been served to a user in that group. An advertiser can then bid on that impression using a ghost ad, which is so named because it is invisible to the user. If the bid wins, the ad is fed to the control group.

This data can be seen by the advertiser and ad platform and is used to ascertain whether there is a difference in consumer behavior among those who have seen the ad versus those who did not see it. O’Neill concludes that ghost ads “create apples-to-apples comparisons of users who were exposed to ads versus users who would have been exposed to ads.”

Another Means to Measure Incrementality

The use of these ads to test consumer behavior answers a huge question for the advertiser: Did my campaign spend result in incremental revenue? Incrementality is “measuring the lift that advertising spend provides to the conversion rate of the target population,” explains Rick Bruner, vice chair for the U.S. at i-com.

“We’re good at targeting the people who are already most inclined to buy a product,” he continues. “So for the advertiser, the question then is: What effect did my ads have on that target audience above the rate at which they would have bought anyway, absent the advertising?”

Using these ads to measure incrementality is easy. An ad campaign is fed out to a group of targeted consumers while at the same time, a control group of consumers who would have normally been chosen to see that ad is shown the exact same message. Each group’s reactions to the ads are then tracked.

The data from each group can then be compared to identify which ad creative, publisher, platform, user behaviors, piece of content, scheduling, etc., are more likely to generate a winning bid and eventual purchase. This not only answers the incrementality question, this comparison also gleans information for attribution measurement—a win-win.

How Valuable Are Ghost Ads, Really?

In understanding which ads facilitated purchasing and increased revenue, advertisers will be better prepared to optimize their future campaigns. For researchers, these ads eliminate selection bias in the testing. Algorithms often target consumers who are the already the most likely to buy; with ghost ads, publishers discover if the consumer would have bought anyway—regardless of whether they saw the ad.

For advertisers, these ads are a cost-effective measurement tool that’s easy to implement and tracks near real time, putting us one step closer to knowing exactly when a consumer made the decision to buy.

This article first appeared on the Videa blog.

Social Media Mapping: The Story Behind the Data

There are many tools available for ascertaining the Who, What and Where of social media, often in the form of a data dump. 

But according to Michael Lieberman, Founder, Multivariate Solutions, companies like Twitter offer data visualization maps that more clearly define all of the attributes of a campaign, hashtag, meme, site or other social phenomena and reveal the story behind the data structure.

Twitter, Google, Yahoo, WhatsApp and Facebook passively cull our personal information via a negative default – that is, requiring the user to opt-out rather than request them to opt-in. This leaves them with a treasure trove of our personal information.

“Social media maps help us visualize the patterns of connection that form when people follow, reply and mention one another in internet communication services like Twitter,” Lieberman explained. The web-like look of these maps show the nature of the conversation whether they are divided, unified, fragmented or clustered.

The use of mapping reveals four types of centrality measures – Degree (how popular), Betweenness (the amount of influence), Closeness (the ability to spread information efficiently) and Eigenvector (how importantly connected) that provide context to social media flow that shows key players and influencers.

According to Lieberman, there are six kinds of Twitter social network maps ——

1-Polarized Map – Using the Egyptian uprising as an example, this map shows two opposing opinions; Feminism and the Uprising (which may not support feminism). Polarized social media discussions have different, often opposing, hashtags.

2-In Groups/Community Map – For a hashtag like #cx (customer experience), these maps show conversations that all use the same or similar hashtags. This indicates the conversation around your hashtag and the penetration of its popularity based on the number of tweets.

3-Bazaar Map – Using @hillaryclinton as an example, the map displays several independent user clusters on many different topics which are not necessarily related to each other.

4-Brand Map – Showcases a public topic such as Disney cruises #disneycruises. These maps show a lot of different groups and hashtags that show what the conversation is around a subject.

5-Broadcast Map – Using @madmagazine or @CNN as an example, these maps show one large group that is essentially talking about same thing. It summarizes what is being shown in the Twitter feed, often stories of the day.

6- Support Maps – Are often corporate maps such as help@dell which has a single originating source that connects directly to many different individual participants.

“The use of a relatively easy analysis such as Social Network Analysis makes understanding the Twitter or Wikipedia conversation far easier than massive social listening platforms,” concluded Lieberman.

This article first appeared in Cynopsis.

Kodi Foster Warns of Living in a Data Echo Chamber

Kodi Foster, SVP Data Strategy, Viacom, is watching the media ecosystem carefully and is seeing some trends that could spell disaster for media companies. His talk at the recent PSFK Conference was a lesson in caution. 

For one thing, he believes that we are marketing to the wrong targets. Further, we may be coming to the wrong conclusions on the data we are capturing because the data itself is skewed, collected in echo chambers of like voices. Tread carefully, he warns, lest you be led astray.

The Internet is Dead

Foster’s company, Viacom, is focused on understanding the current themes of technology and one theme is clear; “The internet is dead,” he began. “But what I really mean by that is that the worldwide web is dead. It is essentially cloud storage and social media. When you really think about it, how many times are you on a dot com nowadays? Maybe Google, but everything else you are doing is on an app.” Or, he added, messaging which is poised to be even bigger than apps in the next year or so. He also believes that surfing the web is “not a thing” anymore and will probably not come back. The overall direction of the internet is decidedly evolving, as with everything else.

Technology Moves Our Cheese

If Foster believes that the tech landscape is changing as people use apps and messaging rather than sites, that leads us down a dark marketing path. According to Foster, “We don’t market to people anymore. We are marketing to the devices that are between us and people. We game theory the algorithms of these technologies so we can get our marketing and content in front of a human being.” Because of this, marketers are actually not trying to understand the human beings. They are trying to understand the biases of these technologies. Foster explained, if you are in marketing, understanding the biases of these technologies has become your job. “Your job is NOT to understand human beings,” he stated. This is leading us down a dangerous analytical path where the data we create every day “is shit” and “when you put shit in you get shit out.”  

Losing Focus in the Echo Chambers of Data

So when you use this bad data to make predictions and craft insights, there is the risk that the outcomes will be wrong. “Machines are not trying to connect people, they are usually promoting something,” Foster added. Promotion is curated. “Who is deciding that curation?” he asked. It becomes a self-fulfilling prophesy within ecosystems because the algorithms are only putting up only certain things it wants you to see and when you click and engage on it, it reaffirms the algorithm, even if it is not something you wanted to see in the first place.

“Socially connected people tend to be similar,” he posited, so there is the danger of creating “an echo chamber.” When you are fed information and you agree with it, the algorithm will calculate to feed you more of the same. “It’s giving you what it thinks you want because that is the only stuff you are seeing,” he added. This creates a social contagion, especially when it involves false information. This is only starting to come to light but, even now, we may not be fully aware of the ramifications. The data we are gathering and using from this echo chamber may be biased and suspect.

Amplifying Bad Results

The advent and growth of fake news is especially troubling. “We assume, because of these finely structured networks, that everyone within our network believes the same thing that we believe. We assume that the belief is bigger across the world than it actually is and that more people share that belief because the only people we are around believe the same thing,” he warned. That is the danger of social echo chambers.

We may not be taking this war on reality as seriously as we need to, despite the discovery of the privacy transgressions of Facebook and Cambridge Analytica. The reliance of the harvested data, collected in stagnant social pools and analyzed without context is leading us down a dark societal path. How can you have an accurate reflection of peoples’ behaviors if the data is not accurate? “What happens to human civilizations when the connected tissue on reality becomes more and more fragmented? How do you create a predictive algorithm for anything when there are tens of thousands of different versions of reality? Whose reality are you predicting?” Foster noted.

Knowing that this is happening and seeing the ramifications leads us to a stark juncture point. Should we just keep calm and continue to play the same marketing game? I believe complacency at this time would be a grave mistake. If we are being led astray now with human-created algorithms, wait until machine learning and artificial intelligence really ramp up in the next few years. Sleep tight, children.

This article first appeared in www.MediaVillage.com

IOVO – Data Power to the People?

The Facebook data scandal with Cambridge Analytica has resulted in an unexpected boost for privacy advocates. Not only is the public more aware of how their data may be used (or misused), there are efforts afoot to empower consumers to better protect their personal data. 

One effort is the recently announced IOVO (Internet of Value Omniledger) which is a DAG (digital acyclic graph- the next generation blockchain) that enables consumers to reclaim their data, secure it and better regulate its use. It is obviously an idea whose time has come, now with special urgency created by the Cambridge Analytical misuse. IOVO is based in Poland but is taking a global approach to data recapture-ment and security.

The panel at the recent IOVO introduction consisted of three very compelling participants; Jeffrey Wernick, an early investor in bitcoin and a self-described crypto currency entrepreneur, Brittany Kaiser, currently Co-Founder of the Digital Asset Trade Association and formerly Business Development Director, Cambridge Analytica and Krzysztof Gagacki, Founder and CEO, IOVO. 

The State of Personal Data

Kaiser noted that today, data companies are flourishing and will continue to do so in the future, despite the recent controversies. “We have given away our information in almost every action in our digital lives,” she noted. “Our data has been harvested, monetized, licensed and sold for advertiser targeting. We did not originally sign up for that. Our data should belong to us as individuals and individuals have been exploited,” she concluded. 

Unfortunately when we agree to terms and conditions on a site, we also agree to allow our data to be collected throughout our digital journey.  This has resulted in the creation of the world’s most powerful companies, such as Google and Facebook, to capture billions of dollars in value off our individual data by creating a new asset class and business model based on big data. However, if an individual does not agree to share their data, it can still be collected and shared through connections who have agreed to share their data.

“Data is not the most vital and valuable asset in the world. Data is the new oil and has even surpassed oil as value,” Kaiser stated. As companies continue to collect data on a massive scale and create algorithms in combination with other datasets, the business of data will continue to flourish in the future.’’

Changing the Business Model

The unfettered Wild West usage of personal data has got the change. The panelists agreed that it is time to create a permission based structure where individuals have control of their personal data and allow companies to use it on a carefully curated and controlled basis with some form of compensation. But where do we start? Is this a self-regulated effort by companies or will it require government regulations?

Wernick described his 4P data manifesto; Data is our Property, only used with Permission. It must be Portable and it represents our Personhood. There are currently “Data rapist platforms,” he exclaimed. Is it individual property or it is everybody's property? If it belongs to everybody then all should have all access to everyone’s data. “But we don't have access to others property. Platforms are curators,” he noted. Data should belong to the individual and require our consent to use. Unfortunately “data is frequently used in unknown ways and in ways that is not good for us. Often it is used against us. It is perverse.”

Power to the People

“I hope that as we expand the potential and possibilities in a blockchain world we can disrupt institution” business models that capture and sell personal data, Wernick noted and added that the concept of universal income may be facilitated in a world where individuals can sell their own data.  “Every human being has value and they are not getting compensated for it,” he concluded.

And the data collected is not just from the digital journey. There is also a plethora of offline data sets and big data aggregators that merge all of these disparate data sets together. Think Info group, Magellan, Equifax and Experian for example. “You can purchase offline ground truth data,” noted Kaiser. Where do you vacation, what are your credit scores, your behavioral patterns, your political data and what media you use. “Data scientists can figure out how you voted even if at ground truth is not available,” she added.

The solution is, according to Gagacki is to “Make data not just a legal right but also a core value of the internet economy.” An omniledger can provide infrastructure for all companies who want to benefit from an individual’s data ownership. Since the data adds value to a company he believes that “We are actual employees of these companies and are not being compensated. It is time to compensate all of the users.” He disagrees with the Facebook economic model. “We need platforms where you can share your value and be compensated for that. Imagine what that means,” he mused, “Everyone could feed themselves off the value of what they create every day.”

“Owning your data is your ultimate freedom,” stated Gagacki. “Large companies have their economic value based on users. We don't have rights in sharing our data and data is the most valuable asset in the digital economy. If we don't stop it now the future generation will have no rights.”

This article first appeared in www.Mediapost.com

The Media Crystal Ball

It is difficult enough to predict how the media landscape will look next year but it is arguably even more difficult projecting how things will look five years hence. A recent Mediapost event gathered some of the best industry minds together to share their thoughts about the state of the media state in 2023. Here is what they predicted:

TV Will Always Be Here

There was optimism regarding the robustness of the TV and the advertising business. According to Tom Goodwin, EVP, Head of Innovation, Zenith, “We tend to think that TV or advertising is dead but it is here to stay.” Further, there was a feeling that even old media has the ability to adapt and reinvent for the new digitized media environment. “The money keeps shifting to online and digital,” stated Brian Hughes, SVP, Audience Intelligence and Strategy, MAGNA. However, he believes that this momentum will slow that down and flatten as “old school media reinvents themselves.”

Delight the Consumer … or Else

Since consumers will have greater power to choose the ads they want to see, those companies that engage people in a negative way may experience blowback. Barry Lowenthal, President, The Media Kitchen, said that, “Facebook will be the big loser’” in the next five years because they fail at the “fundamental human truth” regarding shame about envy. “Facebook peddles in envy. Most feel bad about themselves after seeing Facebook,” he explained, “and you can't sustain a business on envy.”  He predicted that unless Facebook reinvents itself, they will lose money. “It is better to drive to gratitude,” he concluded.

There is also the thought that unless companies respect the consumer by providing them with relevant messaging (without getting creepily intrusive), consumers may decide to opt out and withhold their data and their attention. “Consumers will be in control of their data. GDPR will precipitate that,” noted Natalie Monboit, SVP, Futures for Samsung, Starcom USA, who added, “There will be a shift to drop data when in doubt. Consumers will have more self-sovereignty.”

Streamlined Processes Through Technology

Whether it’s the continuing increase in the amount and type of available data, the introduction of blockchain protocol into the media business, the use of artificial intelligence or the boundaries of privacy, all these issues will impact how we conduct our business five years from now. Monboit noted that, “Blockchain could be the one thing that disrupts today,” and has the potential to change the business.

Finally, it is important for media companies to continue to take risks. Sam Olstein, Global Director Innovation, GE Corporation, concluded that the biggest challenge today is the mindset to avoid more risks. Spending money on things that don’t pan out is “never wasted because it laid the bedrock for the next entrepreneur to benefit from previous mistakes.”

This article first appeared in www.Mediapost.com

Steps to Advancing Advanced Advertising

One of the most fascinating things about attending the B&C Multichannel Advanced Advertising conference each year for the past few years is observing the rate of meaningful change in the industry. This year, it seems that the pace is accelerating and that the advanced advertising ecosystem was changing in a profoundly meaningful way. Perhaps it was the realization that we need to work collaboratively to break down the walled gardens, set standards, solve for challenges like fraud and work towards universally accepted cross platform metrics.

Changing the Way We Do Business

Demand for advanced advertising is growing. According to Freewheel research, the percent of streaming ads on OTT increased 27% in 4^th quarter 2016 up from 8% in 2014. In tapping into this demand, many corporations are developing a set of protocols for measurement and also for developing talent to take the company into the future. Sean Moran, Head of Marketing and Partner Solutions for Viacom shared how his company developed new data-driven solutions and sought out-of-the-box talent.

With data, Moran explained, “We started working on our own capabilities and our own data with predictive analytics and our clients’ first party data. We built an engine that was optimizing every week and every quarter. We have 100% return business and are up to over 80 clients. We also work with over ten agencies. We have really scaled.” With hiring, ”We started to attract a lot of talent from Facebook and Google and get resumes from people who didn't work in TV.” The progress is palpable. “Last year we were talking about view-ability. Now we are talking about content and where it shows up,” he stated.

Breaking Down the Walled Gardens

A constant impediment in rolling out Advanced Advertising beyond custom buying are the walled gardens of data held by individual companies. But the walls are coming down. One of the biggest announcements in the press recently was the joint venture of Viacom / Turner / Fox Open AP which, as Moran explained, “is a buying system. We didn't get together to build the same capability tool. This just offers the ability to load in audiences across the industry. In the end it is third party measurement with third party accreditation.” Then he added tantalizingly, “I won't say what the third party is right now. We will announce that on April 7.”

The efforts to create this Open AP have taken a year, according to Moran, “to establish a single platform. I knew we were on to something when I got calls from agencies saying thank you. Trust is a huge topic right now and this is an independent measurement. Audience buying is the future and we needed to break down some walls.”

Ultimately, the best way to maximize the value of advanced TV inventory is to have an industry standard. This is no small feat when there are so many sellers –large and small - in the mix. Yet, Open AP is an important first step. “Our goal in building this was to include the entire industry,” noted Moran, who added, “All are welcomed to be a part of it. It is de-duplicated so it is a good footprint.”

Challenges Remain

Creative: One area of challenge is creative, the structure of which has not kept pace with the changing ways viewers can consume content. Moran posits, “We have taught writers how to write a show, in essentially three story arcs. Now we want to work to change that and overlay with data science.”

Stitching: Another challenge is stitching data across screens. Michael Bologna, President of Modi Media, noted that when it comes to data across platforms, “There is no one source that stitches it all together, but you can manually stitch it together. At the end of the day the advertiser doesn't care how we do it. They just want it seamlessly integrated into the ecosystem and applied.” James Rooke, GM, Publisher at FreeWheel added, “It is worth remembering how complex the problem is. Think about the fragmentation of viewing that is serving many different endpoints.”

Technology: A relatively new challenge is one of technology as each company is building its own better mouse trap. Bologna observed, “There is a lot of tech that can do different things. But everyone has made their own decisions using different technologies which don't work well together. If I want to buy you I have to handle the challenge of working across all of your technologies.”

Conclusion

As an industry we are making strides. Helen Katz, SVP Media Analytics and Insights, Publicis Media noted, “We are learning a lot as we go.”  Chris Wilson, EVP, National Television, comScore added, “Organizations help us get better. CIMM and the MRC help us guide the process.” The goal is to get to the essence of the issue. “If we can get to ground truth on what is working and not working for our clients, we can improve it,” stated Lindsay Leon-Atkins, CPG Measurement Manager, Facebook. “We can start to iterate. We engage with clients and partners to get to where we need to be,” she concluded.

This article first appeared in www.MediaVillage.com