Annalect’s Chief Marketing Officer Erin Matts talks with FleishmanHillard TRUE about how the explosion of social media channels has changed big data and what companies can anticipate moving forward about its potential.
TRUE: What has been the impact of new social channels on big data? Is it just a question of more of the same?
Matts: This idea of self-publishing — not just “I have my own blog,” but now I can like, comment, share and otherwise let my opinions be known on these platforms — has really been the engine behind the incredible amount of data that’s available from a consumer perspective. Social media is, in fact, what turned data from petite to big.
If you look at its proliferation and you track it over time, there’s a significant inflection point that happens around 2006 and 2007. That coincides with the emergence of YouTube, the emergence of personal mobile devices and, most notably, the emergence of Facebook. There was the famous quote by Google’s Eric Schmidt talking about how we’ve created more data since 2003 than was created in all the years before. But the exponentially expanding quantity, however, is only the tip of the iceberg.
We think about these social data points as signals from consumers. They tell us not only about their behavior, but also how they perceive things. How are they interacting with our products and services? So that’s not just a quantitative difference; that’s qualitative.
Today there’s socialization of nearly everything, and not just for those channels that are deemed specifically social channels. We can even be social when we’re doing our online banking. You’ve got features such as the “Chat with me now” that most credit card companies offer. We’ve got retail experiences where customers are promoting something they want to buy to help bring the price down in a Groupon model. You go into Kate Spade in Tokyo or Burberry in London and there are iPads and touch screens where you can log in and the shopping experience becomes more customized.
Traditional touch points are being converted into digital touch points — and all of them are becoming new sources of data and different types of data that we couldn’t collect before. Data is literally able to paint pictures of customers, and that makes engagement far more personalized.
TRUE: Are companies using big data to its fullest potential?
Matts: I’d say no, in many cases, but they are definitely starting to get what needs to go into big data to get the fullest picture of their customers. If I had to single out one honest mistake many marketers have made and many continue to make, it would be the failure to really expand their sources of data beyond social. It’s a rich and abundant bucket, yes, but it doesn’t always show you all the potential prospects you may have. Social can tell you a lot about what a consumer is thinking and feeling at a particular moment, but it may not be as predictive as you think about how that person may act. You don’t want to regard social as the gospel truth, but rather use that abundance and triangulate it with data on observed behavior and location-based information because what consumers say they want or plan to do isn’t always what they end up doing.
On the positive side, if there’s one trend I am seeing, it’s the evolution of most marketers to an understanding that they won’t get the success cross-platform or across a broader metric if they just depend on social. I think it’s the dawn of a second golden age of customer relationship management (CRM). We’re moving toward the unification of data sets, bringing the streams of data together and building a more holistic picture of customers. Technology has matured to such a degree that we can actually bring these disparate data sets together in something more approximating real time.
This can absolutely change the way we approach marketing decisions. We work with many clients who are doing media mix modeling, but you’re working with data that’s six months old because it takes six months to execute. It’s still valid, but companies need to see that with big data they can bring in real-time insights that, given how fast things move today, can help them keep up with consumer behavior and make better decisions.
When we talk to companies we often hear the same thing — “There’s so much data and we’re just not sure where to start.” It’s almost boiling the ocean for them. Their businesses are moving so fast, and most of them have data living in IT, in marketing, in customer service, in sales; e-commerce may have its own set of data. Those groups aren’t incentivized to share data and collaborate across those channels. Companies may not even have the right infrastructure yet to organize all of it, but there’s an intellectual understanding at least that unifying these streams needs to happen.
TRUE: What’s the most exciting development that will affect data moving forward?
Matts: I might bring up location-based technologies and the data these systems collect — like for instance when you head into Burberry in London. It really adds a new dimension, showing how consumers are actually behaving as opposed to what they say they like or think. With it, what we really want to do is triangulate as many disparate sources of data as we can to produce the richest picture possible of consumer behavior.
TRUE: I was actually thinking about the Internet of things when I asked that question. While it’s not fully realized, what will it mean for brands when products themselves interact with customers with messages and provide data?
Matts: Absolutely. It’s an exciting future to think about. The trick for marketers and their agency partners will be to understand how best to use that information. An interesting data point that came out of a recent advertising conference: 83 percent of consumers expect or want their favorite brands to know everything about them. That’s a fascinating desire balanced against the prevalent sentiment that following me around the Web showing me shoe ads simply because I showed interest in one pair may be an invasion of privacy. So there is a very delicate balance to be maintained, a very subtle dance about when it’s appropriate and helpful to suggest and lead and when it feels creepy. We haven’t entirely parsed that out yet.
TRUE: And how do you define favorite brand?
Matts: I think each brand — and it may change within verticals — may have a different definition of favorite brand. It brings up a very interesting topic. It’s easy when someone is a super fan, an evangelist for the brand on social platforms telling the world how great a brand is. But I consider myself a super fan of a number of products, but I may not crow about them on social platforms. So how does a company distinguish between customers who buy the product occasionally and super fans? That’s when triangulating a number of sources of data really begins to let that company know their customers enough to segment them.
Matts: We just issued a big study on this topic in October. First, there is the cumulative effect of so many and such big retail data breaches, but the public mindset on data has also been touched by the NSA leaks. Data security is now an issue in the mind of consumers. One in eight shoppers says they stopped shopping at the companies affected by the breach. You only have had to look at Target’s numbers to see the impact. And many consumers are changing their behavior in other ways: 39 percent of customers now check their credit card statements regularly; 23 percent of consumers have canceled credit cards affected by data breaches; and 18 percent have changed their online passwords on retailers’ websites. Those are good behaviors even though they’re in response to nefarious activity. In previous years, data breaches could be swept under the rug — and were. That’s no longer the case. Consumers are getting more and more savvy about how to control and protect their own identity and data. And at the very least from a corporate point of view, data protection now needs to be an element of brand protection.
That said, we have not seen a significant resistance to provide data online. But there’s a value exchange. I have to give you something for that data, and if the value of the something is right, chances are you are going to give me that data. Maybe not to one of those companies that had a breach, but in general yes.
Photo credit: Annalect