data

Data: the oil which lubricates the network

In Hardt and Negri's seminal writing, Postmodernization, the authors described the digital area as a paradigm in which "providing services and manipulating information are at the heart of economic production." We can see this play out when thinking of the most successful and dominant players in today's market - Google, Facebook, Apple, Amazon, Netflix, Spotify, etc. Each of these firms provides services (and sometimes product goods) which connect nods (that would normally be disparate) to create more customized experiences. Things connected to other things. Things connected to people. People connected people. Digital facilitates the network which connects nods that otherwise would be disconnected and data is the oil which lubricates these networks. Today's dominating technology extracts the data we passively shed in our day-to-day actions -- research, communications, movement, consumption, commerce -- to optimize our day-to-day lives, helping us make better decisions through the benefit of collective intelligence. Make no mistake, this is not without its tradeoffs, of course. Just like anything else, there are both positive and negative consequences to these technologies and their ability to "provide services and manipulate information." That said, here is one of the coolest examples I've seen lately where data was used to lubricate the network and provide utility to the people.

Ever visited a city and wondered "where should we eat?" In most cases, you try to avoid the chain restaurants because the Olive Garden in NYC is likely the same as the Olive Garden back home. Instead, we venture to fully experience the culture of that city through the tapestry of its cuisine. You know, eat like the locals. We consult Yelp and curated lists from blogs to find the best places to dine in the city. But what if we could use the data people shed as a proxy to find the best spots based on where the people actually go? Apparently, the folks at Crimson Hexagon -- an AI-powered, social listening technology company -- wondered the same thing. To illuminate this curiosity, Crimson used its platform to identify the most popular foods and drinks in New York City on Instagram according to the use of associated hashtags. They gathered the posts containing those tags in the five boroughs over a period of time and visually represented where the true 'hot spots' are in the city. 

Full transparency, Crimson Hexagon is both a close colleague and friend of mine. That said, it does not bias how cool this active truly is. Check it:

Created by Crimson Hexagon

Potbelly: Feed Your Smile

Growing up, there was no part of the school day more coveted than lunch hour. It was the moment between 8am and 3:15pm where socialization among peers was uninhibited by classroom rules. We would use this time to recount the latest episode from previous nights’ most popular show, catch up on the latest grade school gossip, and simply enjoy the recess from scholastic pressure. We experienced this slice of utopia from elementary school all the way through college. Who knew it would come to a screeching halt with the onset of contemporary adulthood?

Let’s face it, lunch hour just ain’t what it used to be. Those happy days, my friend, are gone. The cultural changes in lunch time behavior have seen fewer and fewer Americans leaving their desks for their midday meal and a slice of happiness. This created both a challenge and opportunity for our client, Potbelly, a 40-year-old sandwich shop whose food and quirks garnered a deep love for the brand. For QSRs like Potbelly, lunch is the busiest time for their restaurants, and a softening of traffic during this day-part has a tremendous impact on their business. To curb this, Potbelly would have to create new catalysts to drive store traffic.

This, however, would be no easy feat. Potbelly does not spend a lot of dollars on media and communications. With only 450 shops around the country, Potbelly was being outspent by their competitors in the market — in most cases by two orders of magnitude when compared to the Subway’s of the world. Combined with the rapid growth of the QSR category, due to more and more sandwich shop competitors seeking out would-be lunch goers, Potbelly was finding it increasingly difficult to stand out from the crowd. Of the 1.4M sandwich brand conversations in a 12-month period on social, only 2% of those conversations were about the Potbelly brand. This meant that if Potbelly was going to break through and establish meaningful connections with lunch goers, the brand has to give people something to talk about. Easier said than done. To help Potbelly overcome these hurdles, we figured we’d start with its soul, and use the brand’s conviction to play a role in people’s lives beyond the products it sells. The conviction was pretty simple — Potbelly exists to make people really happy. However, no one outside the company knew what the brand stood for because it hardly communicates it, and certainly didn’t demonstrate it overtly. We saw this as a great opportunity for the brand to not only “zag to the zig” of the crowded QSR category but also connect with humanity. And if the brand is committed to ‘making people really happy’ then it ought to do just that — make people really happy. So we sought out moments of “unhappy” and made it our mission to remove them in ways that were uniquely Potbelly.

We started by associating the email addresses and corresponding Twitter handles of Potbelly’s most ravenous customers. The thought was that if we could find what these people had in common beyond their love for Potbelly sandwiches, then we could possibly identify a cultural driver for the brand to connect with them beyond “the bread.” Of the 360K email addresses, we were able to match 60K Twitter handles. We then used Crimson Hexagon, an AI-powered consumer insight platform, to analyze these individuals’ social conversations to uncover similarities. Of the 60K Twitter handles analyzed, we found that these people disproportionately shared Soundcloud files, and at an unusually high rate. On the surface, that doesn’t sound so discriminating. Who doesn’t love music, right? Considering the size of Soundcloud, relative to the Spotify's and YouTube's of the world, we felt this was quite selective. Not to mention, this proved to be fertile territory for the brand, if for no other reason then the fact that Potbelly enlists live musicians to play in their shops during lunch hour. This gave the brand permission to use music as a means to connect with their customers in a way in which their competitors could not.

At this point, marketers would typically generate creative ideas around this territory, present them to the client, and then test the creative in focus groups to see how it made people feel. However, the behavioral sciences have proven time and time again that self-reported data can be biased, if not simply flawed. Potbelly is not a big spender on marketing and media, so we didn’t have the luxury of potentially be wrong. We had to reduce risk wherever possible. Since we couldn’t outspend the competition, we’d have to outsmart them. To do this, we partnered with Michigan State University’s Media and Advertising Psychology Lab to see if we could really make people happy. Instead of asking respondents how they feel about the content recommendations we prepared for Potbelly, we asked their bodies, observing their psychophysiological responses to the content stimuli. With our academic partners, we exposed respondents to the proposed Potbelly content in a simulated Facebook environment and tracked their cardio, respiratory, sweat gland, facial expression, and eye-tracking responses. This allowed us to not only better understand how people might truly respond to the proposed content but also how they may feel when we consider the associated psychological drivers that led to their physiological responses.

With the psychophysiological study to inform our creative efforts, we were all set to “make people really happy” and feed some smiles. To start, we set our sights on arguably the most unhappy place ever — Twitter. Whether it be complaints, laments, rants, or outright anger, Twitter can be full of bad vibes. Using social listening tools like Crimson Hexagon, we identified 100 unhappy tweets and responded to them with100 original songs performed by improve musicians, in real-time, on one day. The results were tremendous. We generated 10x more brand engagement than the brand had ever experienced on one-day, produced 20x more Twitter followers than the brand normally acquires on a given day, and reached 12 million people during the one-day activation.

But most importantly, we made people really, really happy.

To scale the happiness, we then developed an algorithm that synthesized weather and traffic data, results from major sporting events, and social media happenings across key DMAs to find the most unhappy cities that could use some good vibes. Subsequently, we delivered a little happiness in the form of smile-inducing content every morning throughout the month of June to the people who needed it most. Massive traffic jam on the Dan Ryan in Chicago, torrential downpour in DC, or a big loss for the Rangers would trigger the algorithm — what we called the “Smile Scale” — and we’d identify the right kind of happy to deliver to the right people in those regions programmatically using Facebook and Instagram’s sophisticated targeting tools. As a result, we generated a huge lift in store traffic leading to a 1.4% sales increase in the first sixty days of launch. After the entirety of the campaign, we reached 16 million people and stimulated 800K relevant social engagements. That’s a whole lot of happy for roughly 45K in media spend.

For a competitor brand with challenges at every turn, we helped Potbelly leverage the possibilities of today’s hyper-connected world and the insights of behavioral science to understand people better and create more powerful consumer connections. All of which enabled Potbelly to increase brand engagements by demonstrating its commitment to making people really happy — by actually making people really happy.

Check it!

Potbelly_Cannes_5_3_18_Final_CMYK_7063X5008.jpg

Unlocking Networks: Want to truly understand people and make accurate predictions? Look at their networks.

It’s been said that good marketers see consumers as complete human beings with all the dimensions real people have. But do we marketers really understand people?

For decades we used demographics to identify and segment groups of people in an effort to create better products, serve relevant messages, and forecast more accurate predictions. This is the holy grail of marketing.

But demographics don’t describe “real people.” While gender, race, age, household income, and other demography-based inputs are “truths,” they are static facts and do not accurately describe who people truly are. This, of course, is why savvy marketers focus their segmentation efforts (to whom they target their messages) on psychographics — people’s interests, preferences, and attitudes — because they paint a more vivid picture of "real people.”

Now we’re getting somewhere, but not close enough because psychographics are merely byproducts of our networks. And networks are much better indicators of who people are, and what they are likely to do. 

Let’s unpack this further.

By “networks,” I mean the groups of people with whom we exchange information, experiences, and behaviors: friends, family, classmates, co-workers, teammates, congregates... our people.

And our people give insight to who we are and how we see the world. Within each of our networks are shared beliefs, unwritten rules, rituals, and social norms that guide the behaviors of the people in the network. As Aristotle said, “Man is by nature a social animal,” and these dynamics are the glue that keep our people connected. 

Much of our daily life is governed by norms — unwritten rules we follow to remain community members in good standing. As such, our interests, proclivities, and actions tend to follow the way of our networks and spread in a predictable and contagious fashion.

Our networks inform our psychographics. Therefore, not only are our networks more powerful descriptors of who we truly are than typical demographics, but they are also more holistic representations of ourselves than psychographics alone.

Unfortunately, traditional marketing segmentation misses the mark. Common practice identifies groups of people based on demographics (with a bit of psychographic seasoning) and buckets them into target audiences — a group of passive people waiting for marketing messages to wash over them.

But people aren’t passive, and audiences aren’t real, so this approach often leads to broad generalizations and trite overtures. Peek into most creative briefs, and chances are you’ll see brands targeting “millennials,” as if everyone between the ages of 18-34 are the same because they were born within the same generation. It just isn’t so. As a result, marketers make blanket generalizations about a cohort of dynamic people, and the subsequent work often falls flat.

What a waste.

Networks, on the other hand, are dynamic, human, and innately social. And people use their networks to describe themselves. Take me, for example. I’m a Collins, I’m a Michigan Wolverine, and I’m a non-denomination Christian. I subscribe to these networks and take on their respective characteristics to stay in good standing with my people — as we all do with our own unique networks.

Understanding the dynamics of these networks is the gateway to consumer intimacy and relationship development because these groups of people are, in short, real. Marketers would benefit greatly by shifting their focus from talking at passive “target audiences” to engaging with active “target networks.”

Even more interesting, networks are also more accurate indicators of what we’re likely to do. This is heavily supported by behavioral science research. Humans are naturally inclined to take on the actions of the people around us, so much so that our behavior can be predicted from exposure to the example behavior of others. And we are most influenced when we observe the behavior of people most like ourselves — our networks. That means if brands can understand the dynamics of my network, then not only will they better understand me, they’ll also be able to predict my behavior with a high degree of probability.

Now that’s powerful. 

These predictions are driven by the natural propensity that people have to rely on one another. We’ve built trust within our networks and rely on their expertise and experiences to help inform our decisions. In fact, research shows that we trust the recommendations of our people more than any form of advertisement or media.

The collective intelligence of our networks help us decide where we go, what products we consume, who we vote for, and which brands we choose. As a result, our consumption patterns naturally follow that of our networks. Want to predict what people will do next?

Watch the behavior of their networks.

Contrary to conventional wisdom, we are not independent agents in this world, where our decisions are driven by our preferences and IQ. Rather, we live in complex systems — networks of people — where members therein help shape each other’s affects, cognitions, perceptions, and beliefs. We rely on our ability to learn from the behaviors of our people, and they set the example for how members of the network should also behave. These networks move forward on the basis of a simple, subconscious, question: “Do people like me do something like this?”

If the answer is "yes," then we follow suit; if "no," then we don't take action.

We don't inquire.

We don't share.

We don't buy.

It’s that simple. And it all starts with people — real people — and the influence of their networks. This sets the stage for a more actionable approach where brands can deliver ideas, products, and communications in an effort to influence consumer behavior.

Considering the ubiquity of social media in today’s connected world, marketers can now apply network thinking to the use of these tools in a way that promotes social pass-along and enables more accurate predictions.