Big Data (for Marketing) is Real!

MachineLearningWe’ve been hearing about big data driving marketing for a long time, and to be honest, most is purely aspirational.

Using third-party data to target an ad in real time does deploy some back-end big-data architecture for sure. But the real promise of data-driven marketing has always been that computers, which can crunch more data than people and do it in real time, could find the golden needle of insight in the proverbial haystack of information.

This long-heralded capability is finally moving beyond the early adopters and starting to “cross the chasm” into early majority use among major global marketers and publishers. 

Leveraging Machine Learning For Segmentation 

Now that huge global marketers are embracing data management technology, they are finally able to start activating their carefully built offline audience personas in today’s multichannel world.

Big marketers were always good at segmentation. All kinds of consumer-facing companies already segment their customers along behavioral and psychographic dimensions. Big Beer Company knows how different a loyal, light-beer-drinking “fun lover” is from a trendsetting “craft lover” who likes new music and tries new foods frequently. The difference is that now they can find those people online, across all of their devices.

The magic of data management, however, is not just onboarding offline identities to the addressable media space. Think about how those segments were created. Basically, an army of consultants and marketers took loads of panel-based market data and gut instincts and divided their audience into a few dozen broad segments.

There’s nothing wrong with that. Marketers were working with the most, and best, data available. Those concepts around segmentation were taken to market, where loads of media dollars were applied to find those audiences. Performance data was collected and segments refined over time, based on the results.

In the linear world, those segments are applied to demographics, where loose approximations are made based on television and radio audiences. It’s crude, but the awesome reach power of broadcast media and friendly CPMs somewhat obviate the need for precision.

In digital, those segments find closer approximation with third-party data, similar to Nielsen Prizm segments and the like. These approximations are sharper, but in the online world, precision means more data expense and less reach, so the habit has been to translate offline segments into broader demographic and buckets, such as “men who like sports.”

What if, instead of guessing which online attributes approximated the ideal audience and creating segments from a little bit of data and lot of gut instinct, marketers could look at all of the data at once to see what the important attributes were?

No human being can take the entirety of a website’s audience, which probably shares more than 100,000 granular data attributes, and decide what really matters. Does gender matter for the “Mom site?”Obviously. Having kids? Certainly. Those attributes are evident, and they’re probably shared widely across a great portion of the audience of Popular Mom Site.

But what really defines the special “momness” of the site that only an algorithm can see? Maybe there are key clusters of attributes among the most loyal readers that are the things really driving the engagement. Until you deploy a machine to analyze the entirety of the data and find out which specific attributes cluster together, you really can’t claim a full understanding of your audience.

It’s all about correlations. Of course, it’s pretty easy to find a correlation between only two distinct attributes, such as age and income. But think about having to do a multivariable correlation on hundreds of different attributes. Humans can’t do it. It takes a machine-learning algorithm to parse the data and find the unique clusters that form among a huge audience.

Welcome to machine-discovered segmentation.

Machines can quickly look across the entirety of a specific audience and figure out how many people share the same attributes. Any time folks cluster together around more than five or six specific data attributes, you arguably have struck gold.

Say I’m a carmaker that learned that some of my sedan buyers were men who love NASCAR. But I also discovered that those NASCAR dads loved fitness and gaming, and I found a cluster of single guys who just graduated college and work in finance. Now, instead of guessing who is buying my car, I can let an algorithm create segments from the top 20 clusters, and I can start finding people predisposed to buy right away.

This trend is just starting to happen in both publishing and marketing, and it has been made available thanks to the wider adoption of real big-data technologies, such as Hadoop, Map Reduce and Spark.

This also opens up a larger conversation about data. If I can look at all of my data for segmentation, is there really anything off the table?

Using New Kinds Of Data To Drive Addressable Marketing 

That’s an interesting question. Take the company that’s manufacturing coffee machines for home use. Its loyal customer base buys a machine every five years or so and brews many pods every day.

The problem is that the manufacturer has no clue what the consumer is doing with the machine unless that machine is data-enabled. If a small chip enabled it to connect to the Internet and share data about what was brewed and when, the manufacturer would know everything their customers do with the machine.

Would it be helpful to know that a customer drank Folgers in the morning, Starbucks in the afternoon and Twinings Tea at night? I might want to send the family that brews 200 pods of coffee every month a brand-new machine after a few years for free and offer the lighter-category customers a discount on a new machine.

Moreover, now I can tell Folgers exactly who is brewing their coffee, who drinks how much and how often. I’m no longer blind to customers who buy pods at the supermarket – I actually have hugely valuable insights to share with manufacturers whose products create an ecosystem around my company. That’s possible with real big-data technology that collects and stores highly granular device data.

Marketers are embracing big-data technology, both for segmentation and to go beyond the cookie by using real-world data from the Internet of Things to build audiences.

It’s creating somewhat of a “cluster” for companies that are stuck in 2015.

Advertisements

How Political Campaigns are Putting DMPs to Work

 

Spacey

People like to speculate that social networks impacted the 2016 Presidental election, but Facebook isn’t that influential. I mean, I still don’t like cats very much, and that’s 90% of the content. 

We have all heard about the Democratic Party’s skill with data, and there is no doubt the Obama campaign’s masterful use of first-party registration data to drive online engagement, raise funds and influence political newbies helped put him over the line.

 

Four years later, the dynamics are mostly similar, but we have moved into a world where mobile is dominant, more young new voters are highly engaged and the standard segmentation – at least on the Republican side – might as well be thrown out the window.

In other words, everyone is getting influenced on their mobile phone, especially through news and social channels. There are a ton more mobile-first, new voters out there, and nobody is really sure which voters make up this weird new Trump segment.

To get a handle on this, political advertisers need to properly onboard and analyze their data to identify who they should target, where they live and what they like.

Understand Voter Identity

In politics, a strong “ground game” is key. That means real, old-school retail politics, such as knocking on doors and getting voters in specific precincts out on Election Day. All campaigns have the voter rolls and can do their fill of direct mail, robocalls and door knocking.

But how to influence voters well before Election Day who are tethered to their devices all day and night? It requires a digital strategy that can reach voters across the addressable channels they are on, including display, video, mobile and email. This strategy should leverage an identity graph to ensure the right messaging is hitting the same voter – at the right cadence.

Maybe “Joe the Firefighter,” a disaffected moderate Democrat who has had it with the Clintons, visited the Donald’s website and is ready to “Make America great again.” Before cross-device capabilities were strong, you could only retarget Joe the next time you saw his cookie online.

Today, Joe can get an equity message reinforced on display (“Make America great again!”), a mobile “nudge” to take action when we see Joe on his tablet at night (“Donate now!”) and follow up with an email a few days before the big rally (“Come see the Donald at the Civic Center!”).

Beyond this capability is the incredibly important task of laddering up individual identity into householding, so we can understand the composition of Joe’s family, since households often vote together and contain more than one registered voter.

Nail Geographic Targeting by County and District

Since “all politics is local,” it follows that all digital advertising should be locally targeted. This is table stakes for digital providers that work with campaigns, and targeting down to the ZIP+4 level has brought a level of precision to district-level outreach that approaches direct mail.

But direct mail (household targeting) is the crown jewel and digital is still trying to cross that divide, but is held back by a fragmented ecosystem of identity and, more importantly, privacy considerations.

This has always been a key consideration, given the fact that a small percentage of key districts can flip the presidency to one party or another.

Affiliation Modeling Through Behavior

Sometimes getting an understanding of someone’s party affiliation is super obvious, such as “liking” a specific candidate on social media. But, sometimes, a user’s affinity has to be derived through attributes derived through his or her behavior and the context of content consumed over time.

Data management platforms are bringing more precision to this type of modeling. Functionality, such as algorithmic segmentation, is helping digital analysts go beyond the basics. It’s fairly easy to correlate two or three attributes, such as income and gender, to estimate party affiliation. In this cycle, for example, we have seen a strong bias toward Trump from lower-income males with less than a college degree.

However, it’s hard for humans to correlate eight or more distinct attributes. Maybe those lower-education, low-income, rural males who love NASCAR actually lean toward Bernie Sanders in certain districts. Letting the machines crunch the numbers can give digital campaign managers an unseen advantage, and that capability has just now become available at scale.

“In 2016, relying on TV advertising to sway voters is no longer a solid campaign tactic,” JC Medici, Rocket Fuel’s national director of politics and advocacy, told me via email. “To secure the White House in November, candidates must now add a strong digital media strategy by utilizing best-in-class AI, correlated with strong voter and propensity data assets to ensure they are delivering ads to the right voter, on the right screen, at the right time.”

Social Affinity

One of the hot new areas for political campaign targeting is social affinity, the idea that there is a mutual affinity that can be measured between interests.

Yes, when someone “likes” Hillary, you have an obvious target. But, how about those folks who haven’t stated an obvious choice? Maybe 80% of Hillary fans also liked cat shelters, yellow dresses and Chris Rock.

When strong correlations between deterministic social behavior are shown, it becomes fairly easy to leverage that data for targeting – and make informed choices regarding media. People who liked Hillary also like certain TV shows, actors, causes and websites. Campaign managers can leverage data from Affinity Answers, Affinio and other companies to understand these relationships and exploit them to build support for candidates, while leveraging the ability to geotarget at very granular levels on Facebook.

The Free State Project, an organization committed to getting 20,000 “liberty-loving” people to move to New Hampshire and work toward limited government, just reached its goal – talk about a tough conversion. President Carla Gericke credits the use of data-driven targeting on Facebook for the achievement.

Speaking of social, it is also highly important to get the context right.

“Programmatic has introduced two new challenges: bots (who don’t vote) and brand safety,” Trust Metrics CRO Marc Goldberg told me. “In the age of immediate and shocking news, it has become more important that a political ad does not end up next to porn, hate or issues that are contradictory to the politician’s beliefs. One screen shot and bam, you are on Twitter.”

Onboarding And Offboarding 

Perhaps the most critical functionality for digital political campaigns continues to be the ability to “onboard” offline data, such as phone numbers, email addresses and party affiliation, and match it to an online ID for targeting purposes. This is essentially table stakes, considering the years of political investment in collecting offline records for phone banks and direct mail campaigns.

Previously, the onboarding of such data was limited to associating it with an active cookie for retargeting use. But with the emergence of real cross-channel device graphs, this data can now be tied to a universal consumer ID that is persistent and collects attributes over time.

Simply put, that onboarded email – now a UID – can be mapped to a number of identities, including Apple and Android mobile identifiers, third-party IDs from Experian and the like and device IDs from Roku and other OTT devices. In other words, the device graph enables that email to be associated with the voter’s omnichannel footprint, giving campaigns the ability to sequentially target messages, map creative to execution channels and truly understand attribution.

What’s even more exciting is the idea of offboarding some digital data back into the CRM. How valuable would it be to know that a potential voter watched an entire YouTube video on a candidate after being reached by the phone bank? Certain types of behavioral data, depending on compliance with privacy policies, can be brought back into the CRM to impact the effectiveness of offline voter outreach.

It is fair to say that 2016 is the most exciting campaign season we’ve had in a generation – and it’s only the primary season. As data-driven marketers, we will see campaigns push the limit in applying big marketing dollars to digital channels, trying to unlock new, mobile-first millennial voters, while persuading independents through more addressable advertising then ever.

It’s a great time to be a data-driven marketer.