Creating the Fabled 360 View of the Consumer

ImageDespite years of online targeting, the idea of having a complete, holistic “360 degree view” of the consumer has been somewhat of a unicorn. Today’s new DMP landscape and cross-device identification technologies are starting to come close, but they are missing a key piece of the puzzle: the ability to incorporate key social affinities.

In the nearby chart, you can see that online consumers tell us all about themselves in a number of ways:

Viewing Affinities: Where they go online and what they like to look at provides strong signals of what they are interested in. Nielsen, comScore, Arbitron and others have great viewership/listenership data that is strong on demographics, so we can get a great sense of the type of folks a certain website or show attracts. This is great, but brands still struggle to align demographic qualities perfectly with brand engagement. 34 year old men should like ESPN, but they could easily love Cooking.com more.

Buying Affinities: What about a person’s buying habits? Kantar Retail, OwnerIQ, and Claritas data all tell us in great detail what people shop for and own—but they lack information on why people buy the stuff they do. What gets folks staring at a shelf to “The Moment of Truth” (in P&G parlance) when they decide to make a purchase? The buying data alone cannot tell us.

Conversational Affinity: What about what people talk about online? Radian6 (Salesforce), Crimson Hexagon, and others really dig into social conversations and can provide tons of data that brands can use to get a general sense of sentiment. But this data, alone, lacks the lens of behavior to give it actionable context.

Social Behavioral Affinity: Finally, what about the actions people take in social environments? What if we could measure not just what people “like” or “follow” online, but what they actually do (like post a video, tweet a hashtag, or engage with a fan page)? That data not only covers multiple facets of consumer affinity, but also gives a more holistic view of what the consumer is engaged with.

Adding social affinity data to the mix to understand a consumer can be a powerful way to understand how brands relate to the many things people spend their time with (celebrities, teams, books, websites, musicians, etc.). Aligning this data with viewing, buying, and conversational data gets you as close as possible to that holistic view.

Let’s take an example of actionable social affinity in play. Say Whole Foods is looking for a new celebrity to use in television and online video ads. Conventional practice would be to engage with a research firm who would employ the “Q Score” model to measure which celebrity had the most consumer appeal and recognition. This attitudinal data is derived from surveys, some with large enough sample sizes to offer validity, but it is still “soft data.”

Looking through the lens of social data, you might also measure forward affinity: how many social fans of Whole Foods expressed a Facebook “like” for Beyonce, or followed her account on Twitter? This measurement has some value, but fails at delivering relevance because of the scale effect. In other words, I like Beyonce, so does my wife, and so does my daughter . . . along with many millions of other fans—so many that it’s hard to differentiate them. The more popular something is, the broader appeal and less targetability that attribute has.

So, how do you make social affinity data relevant to get a broader, more holistic, understanding of the consumer?

Obviously, both Q Score and forward affinity can be highly valuable. But when mixing viewing, buying, and listening with real social affinity data, much more becomes possible. The real power of this data comes out when you measure two things against one another. Sree Nagarajan, CEO of Affinity Answers, explained this mutual affinity concept to me recently:

“In order for the engagement to be truly effective, it needs to be measured from both sides (mutual engagement). The parallel is a real-world relationship. It’s not enough for me to like you, but you have to like me for us to have a relationship. Mapped to the brand affinity world, it’s not enough for Whole Foods fans to engage with Beyonce; enough Beyonce fans have to engage with Whole Foods (more than the population average on both sides) to make this relationship truly meaningful and thus actionable. When true engagement is married with such mutual engagement, the result is intelligence that filters out the noise in social networks to surface meaningful relationships.”

As an example, this approach was recently employed by Pepsi to choose Nicki Minaj as their spokesperson over several other well-known celebrities.

What else can social affinity data do?

  • Brands can use social affinity data to decide what content or sponsorships to produce for their users. Looking at their users’ mutual affinity between the brand and music, for example, might suggest which bands to sponsor and blog about.
  • A publisher’s ad sales team can use such data to understand the mutual affinity between itself and different brands. A highly correlated affinity between activated social visitors to GourmetAds’ Facebook page and those who post on Capital One’s Facebook page may suggest a previously unknown sales opportunity. The publisher can now prove that his audience has a positive predisposition towards the brand, which can yield higher conversions in an acquisition campaign.
  • What about media buying? Understanding the social affinity of fans for a television show can produce powerful actionable insights. As an example, understanding that fans of “Teen Wolf” spend more time on Twitter than Facebook will instruct the show’s marketing team to increase tweets—and post more questions that lead to increased retweets and replies. Conversely, an Adult Swim show may have more Facebook commenters, leading the marketer to amplify the effect of existing “likes” by purchasing sponsored posts.
  • Keyword buying is also interesting. Probing the mutual affinities between brands and celebrities, shows, music acts, and more can yield long tail suggested keyword targets for Google, Bing/Yahoo, and Facebook that are less expensive and provide more reach than those that are automatically suggested. As an example, when “Beavis and Butthead” re-launched on MTV, Google suggested keywords for an SEM campaign such as “Mike Judge” (the show’s creator) and “animated show.” Social affinity data suggested that socially activated Beavis fans also loved “Breaking Bad.” Guess what? Nobody else was bidding on that keyword, and that meant more reach, relevance, and results.

I believe that understanding social affinity data is the missing piece of the “360 degree view” puzzle. Adding this powerful data to online viewing, buying, and social listening data can open up new ways to understand consumer behavior. Ultimately, this type of data can be used to generate results (and measure them) in online branding campaigns that have thus far been elusive.

Want a full view of the people who are predisposed to love your brand? Understand what you both mutually care about through social affinities—and measure it.

[This post originally appeared in AdExchanger on 4.14.14]

 

Advertisements

Social Affinity

Is your media measurement as dated as this 1970s den?

Is your media measurement as dated as this 1970s den?

The New Panel-Based Audience Measurement for Brands

With the prevalence of social data, yesterday’s panel-based measurement for digital campaigns is starting to look like the wood paneling in your grandmother’s den: A bit out of fashion. Marketers have been trained to buy media based on demographics, and it is natural to want your ads to be where you think your customers are. For BMW’s new entry level sedan, that might mean finding the media that males aged 26-34, who are earning 75,000 or more a year, consume. That makes a lot of sense, but it also means that your paid media will always compete alongside ads for your competitors. That is a big win for websites with premium inventory that fits your demographic, because it means scarcity and high prices for marketers.

What if there was another way to measure what audiences are right for brands? And what if that data were based on a panel of a few hundred million people, rather than a few thousand? Well, thanks to Facebook and Twitter, we have just such a web-based panel of consumers, and they are always eager to share their opinions in the forms of “likes,” “follows,” and (more importantly) engagement. Social listening platforms have been able to tell brands what people think about them directionally, and measure how certain marketing efforts move the social needle. Listening is great, but how do you get to hear what to buy?

A company called Colligent has been going beyond listening, by measuring what people actually do on Twitter, Facebook, and other social sites. “Liking” is not enough (when me, my 10-year old daughter, and my mom “like” Lady Gaga, the audience I am a part of gets too broad to target against). What matters is when people express true affinity by sharing videos, tweeting, and commenting. When people who are nuts about a certain celebrity are also nuts about a certain brand—and that relationship over-indexes against normal affinity, you have struck real social gold: Data that can make a difference. Pepsi recently used such data to choose Nicki Minaj as a spokesperson over dozens of other choices.

What about other media? Nielsen defines television, Arbitron measures radio, and MRI defines magazine audiences by demographics. Now, for the first time, marketers can use social data—gathered from panels nearly as large as the buying population—to define audiences by their own brand and category terms. That’s a world in which Pepsi can purchase “Pepsi GRPs” across all media, rather than GRPs in a specific media.

This is the way brands will buy in the future.

[This post originally appeared on 2/21/13 in The CMO Site, a United Business Media publication]

Thoughts on Data-Driven Audience Measurement

A Conversation with Scott Portugal of PulsePoint

What are some best practices for the modern digital marketer? Cookie-based data makes knowing your audience easier  than ever. Developing accurate audience profiles, optimizing campaigns based on audience composition, and validating audience reach are all critical components for marketers doing targeted digital campaigns. I recently spoke with Scott Portugal, long time digital media veteran and currently VP of Business Development for PulsePoint, who has been working with PulsePoint’s Aperture audience measurement offering, what marketers should be thinking about when it comes to measurement.

Scott Portugal: First and foremost, marketers must really understand the goals of the campaign. “Branding” vs. “Performance” aren’t goals – they are notional indicators of goals. “Increase brand awareness amongst men passionate about health and fitness by 50%,” is a goal. The more specific, the better. It eliminates the guesswork that agencies have to do around media tactics, and most importantly, specificity in KPIs means everyone knows which data sets to use along the way.  Also, a modern marketer knows that buying digital media isn’t an on/off switch. Once the buy starts, the work starts. Prepare to optimize everything you can – look at performance across targets, media partners, creative (the most important and often least optimized variable), etc.. Good digital marketers are like good scientists: ask plenty of questions, account for all variables, and constantly test to find success.

What new tools are out there to assist in audience measurement, and supplement the standard offerings from Comscore, Nielsen?  

SP: Data is ubiquitous – some might say commoditized. But there are a few platforms out there that are taking novel approaches to audience measurement. Certainly our PulseAudience platform is among that group. We’re able to build audience profiles at the domain level, meaning at a very granular level we can infer the audience composition of a page even without a cookie. Another new player is Korrelate, founded by the guys who ran TACODA. Korrelate is in the business of helping marketers understand how different data sets perform across different platforms – essentially helping a buyer know what data segment to buy when and where. At a broader level, audience measurement platforms are starting to look cross-media, bringing together disparate data sets that show impact of a campaign on ALL digital activity, not just clicks.

What about social data? How are technologies like Facebook and Twitter enabling a more concise view of audiences, and helping marketers validate their choices?  

SP: If you think about Facebook and Twitter NOT as destinations, but as communication tools, then you can start to see where a more holistic audience view can be created. Social media is more than updates – it’s sharing news, communicating about brands, raising hands about interests, and more. Social data, when done right, is true first party data that goes above and beyond standard behavioral data. Marketers can understand not just when a user engages, but how, where, and how valuable that engagement actually was (likes, shares, tweets, etc.). it should validate a marketers choice around creative and placement, but only if the creative and placements actually include social elements. Social data is powerful, but it’s only powerful if it’s part and parcel to other data sets and targeting mechanisms used in conjunction with social media. Nothing happens in a vacuum, and nothing happens ONLY in one channel.

Your company owns Aperture. Can you provide some examples of how progressive media organizations are using audience measurement data? Is it about audience validation? Optimization? Upselling clients?  

SP: It’s about delivering value via insights up and down the funnel. It sounds like ad jargon but it’s what we strive to do with every single engagement. Cookie targeting works, but we believe that there is real value in modeling at other points of content interaction – insights that help guide and inform at all points of the campaign. Our RTB partners can leverage some of this data in real time; our non-programmatic partners work with our data and insights group to go even deeper via custom reporting and deeper dives on how to get consumers to engage. Data availability and normalization—what we do—is what makes the tide rise to lift all boats.

How can (the right) measurement data influence brand advertising? Is this the key to bringing more brand dollars online?

SP: Brands will feel safe moving dollars over from television to digital when they can do two things: ensure the environment is safe and ensure that they are reaching the right audience with minimal waste. Does television have massive amounts of waste in it? Of course – but as an industry we promised the world that we would eliminate much of that problem via targeting and optimization, so we have to lay in the bed we made. So measuring not just reach & frequency, but the impact of that reach & frequency is critical. Did search queries go up relative to their competitors? Did social commentary increase? Are there more tweets about campaigns in other platforms (did you create awareness that increases awareness in other channels as well)? Like I said earlier – understanding the specific goals of that branding campaign, and ensuring that the right creative is matched with the right tactics, will allow for the right measurement data to be used.

What’s next in measurement?  

SP: To me it comes down to cross-platform impact. Devices and screens aren’t truly linked yet, but the audience at the other end of that ad campaign is the same person. They tweet, they promote, they like, they friend, they blog, they comment, they shop….but they do it across multiple screens in the home, the office, and on the street. The best measurement companies are going to be those that can build an impact assessment across ALL platforms and show the points of interconnection. It’s a big task – but the ones who get it right will be the ones working directly with marketers and become embedded into everything they do. The next big push will be to show marketers that social, search, display, video, and mobile are all tactics inside the same strategy…and then show them how each tactic impacts the other.

This interview, among many others, appears in EConsultancy’s recently published Best Practices in Data Management by Chris O’Hara. Chris is an ad technology executive, the author of Best Practices in Digital Display Media,  a frequent contributor to a number of trade publications, and a blogger.