Big data and the “Big Lie”: the challenges facing big brand marketers

The challenge for companies of all sizes is how best to integrate the Don Drapers and the data scientists.

Previously, I wrote about the future of marketing being a fusion of the art of storytelling with the specificity of data and the objectivity of analytics.  Consumer attention is shifting from TV, print, and radio to digital, which has made tailored, real-time engagement with customers both possible and increasingly necessary. In a world of proliferating channels and constant competition for attention there’s a lot of pressure for brands to get it right.

There’s a sense among many in the tech community that marketing is an industry that’s locked in the past. Within the early-stage startup ecosystem, “marketing” has evolved into “growth hacking.” Growth hacking relies almost exclusively on data and analytics to develop strategies for optimizing customer acquisition, retention, pricing, and more. It’s iterative and rapid; learnings are incorporated into refining both the outreach and the product itself. While the goal is largely the same, the terminology reflects a differentiation between a traditional Don Draper “gut instinct” approach and the new data geek methodology.  The marketing industry is well aware of the need to adopt faster feedback loops and nimble campaign strategies; industry conferences feature panels on ramping up digital infrastructure, aligning CMO and CIO goals, and tying spend to outcome. Marketers do recognize the importance of data, but big brands care a great deal about emotional resonance, and the trend toward personalization makes targeted storytelling more complex.

In order for a brand to connect with a customer on a personal level, it’s necessary to understand who that customer really is. Appropriately leveraging data and analytics to target effectively, generate awareness, and drive traffic is challenging. For big brands, there’s often a lot of data, but that does not necessarily guarantee a lot of meaningful insights. I recently heard the CMO of L’Oreal speak at a conference; he described hundreds of thousands of posts per day about his company’s products, spread across a multitude of sites and apps, all over the world. That is data. But how much of it will lead to actionable insights? Is it worth it to gather, store, process, and analyze that information?

The proliferation of “social” data introduces additional complexity. Most of us have several active social media profiles. We engage in liking and sharing behavior on the web and on mobile, much of which is linked back to the big social sites. This seems like a wealth of mineable information for brands. However, the flip side of “social” is what’s come to be called The Big Lie: “the gap between social norm and private reality, between expressed opinions and inner motions.” We ensure that our Facebook and LinkedIn profiles present us in our best light. Our shared audio playlists highlight the artists we’re proud to call ourselves fans of — and conceal the mass-market pop that we actually listen to when we’re alone. We use Instagram to share our most gourmet dining experiences, not our Oreo habit. There’s an important distinction between user-generated data and user-volunteered data.  Targeting someone using data they generated but did not volunteer can put a brand squarely into the “creepy” zone.

Even when users do volunteer quite a bit of accurate info, it isn’t necessarily being captured properly. Go have a look at Acxiom’s AboutTheData* or Google’s Ad Preferences to see what marketers may know about you (RapLeaf used to give users access, but this seems to have changed). My Acxiom profile pegs me as a 32-year old woman (correct, but I gave them my birthday and gender on login), with a 15-year-old kid and no college education, who’s spent a total of $450 online over the past 24 months (not even close). This lack of accuracy presents challenges for brands actively trying to incorporate data into personalized targeting.

Independently of data validity issues, there is also the challenge of developing a meaningful analytics strategy. While startup marketing is often wholly digital, big brands must measure outcomes across many channels, and most aren’t yet adept at tracking metrics in the online world. At the recent Financial Times Future of Marketing conference, marketers lamented the difficulty of determining the ROI of a digital “social” strategy. Should it be measured as earnings outcomes tied to a specific series of interactions? Or should it be measured in terms of long-term brand equity and goodwill? When allocating spend, is it better to focus on a digital strategy targeting individuals who fit a data-generated profile or doing a broad traditional TV push that reaches thousands of potential customers, even if little is known about them on an individual level? It’s often difficult to determine the right KPIs for a campaign, which means that making a large campaign nimble is a huge challenge.

There will always be a need for creativity in marketing; the best targeting algorithm and most accurate data set in the world can’t make a boring campaign compelling. So, the challenge for companies of all sizes is how best to integrate the Don Drapers and the data scientists. If you’ve worked on either side of the industry (traditional marketing or growth hacking), I’d love to hear your thoughts on what approaches have worked for you.

*Warning – they will ask you for certain pieces of personal data, ostensibly to verify your identity, but likely also to harvest information. I didn’t correct any of their misconceptions.

Get the O’Reilly Data Newsletter

Stay informed. Receive weekly insight from industry insiders.

  • snagpal

    This article seems to provide social media site of story which is accurate …We have more then 70% who would never give the right identity for their accounts unless it’s to do with any transactions on web …Data collated seems to be much hyped unless very targeted effort is made by some organisation which already has data in there system ….relying solely on the social media data may be waste of time and energy :))
    Good thought on marketing :) Big Data needs to be used very focused for it to provide good results

  • Jud

    Couldnt agree more that social data is very “surfacy”. We present ourselves in a light that we think others either see us as, or how we want them to see us. Even more than that people tend to be compulsive, and we might not realize we want something until its right in front of us. Its like going to the grocery store for bread and coming out with a pizza.

  • Boris Y. Iyutin

    Not only data is biased, the data analysis tools and methods are skewed towards finding positive results. Given large dataset and list of variables long enough it is only a matter of technique to find some correlation. Even when researcher’s pay depends on the application of their results (HFT), most of the findings are throwaway. I would be very skeptical about any big data findings.

  • dusakabin

    canlı bedava ligtv, canli mac izle, canli yayin mac, d-smart izle, justin tv, justin tv izle, lig tv 2, ligtv, ligtv canli, ligtv izle, mac izle, p2p, trt spor izle, freedocast, eurosport izle bedava

    http://canlimacizlelinkler.blogspot.com/

  • Nyctele Nyctelecomm

    Ummm, big data, in the puter world refers to horizontally scalable databases (mongodb, nosql). Their queries do less tricks, but they do it really fast and you can plug in two and they will almost run doublly fast (hence, horizontally scalable) vice, a vertically scalable system like MySQL.
    All the rest you talk about is unrelated end user stuff.
    No, big data is not a big lie, it works really well if you configure it correctly. Couldn’t tell you about the rest though.