Big Data Blah Blah Blah (Part 1)

BjA3fwpCQAIiRMnFriends, I have something very serious to discuss with you today. As you may have heard, there is a huge problem greatly impacting all of us. In fact, should we not do something about it, I would go so far as to predict that the end is nigh. It may be too late to save some of you, and for that I’m very sorry. You see, you have all been affected by a “big data” problem.

You didn’t know you had a big data problem? How could you not? It seems like every time you turn around, someone is telling you about how you need a solution for your big data problem. Just doing a quick scan of the recent news, I can see some compelling articles:

Screen Shot 2014-03-26 at 10.33.48 AM

Screen Shot 2014-03-26 at 10.34.11 AM

I’m not a big fan of hype and I can tell you that big data has reached a hype level that makes it undeserving of our admiration. The next time someone tells you that you need a “big data something or other,” you should send them to the corner where they can be openly mocked by their peers. Check out the Gartner Hype Cycle, which shows where Big Data is currently positioned – at the top:

Screen Shot 2014-03-26 at 10.36.27 AM

I’ve been somewhat of a technology geek my entire life and I’ve gotten pretty good at spotting these type of inflated concepts. I can understand why people use the term because it covers a wide array of very general data and computing challenges. This is great for those who would like to be able to describe to their family what they do for a living without actually having to explain anything, but alas, when something means everything, it must mean nothing.

Ambiguity is the biggest problem with the term big data. I’ve seen so many products and startups that describe fairly ordinary technical challenges as big data problems in order to make their solutions sound more interesting. This type of glomming onto a term mocks the very essence of the word “big.”

So do we marketers have a “big data” problem? In most cases I’d say that’s not the main problem facing marketing organizations. With some exceptions, most marketers have a fragmented data problem. Marketing organizations are generating a reasonable amount of data, but these days, most data processing technologies wouldn’t break a sweat dealing with the volume. The bigger challenge is that every marketing application servicing every channel produces its own data, with different syntax, different ways of identifying visitors or customers, different frequency of delivery, differing levels of detail, and differing data volume to actionable value ratios.

Has all of this focus on the data impeded the progress towards better marketing interactions and customer service?

If you spend a lot of time building your Hadoop cluster just to answer the question, “What makes my customers want to do more business with me?”, then I think the answer is obvious.

How much time have you actually spent thinking about the questions you need to answer versus building a system that can answer every question? To paraphrase Darth Vader (a great marketer by the way), don’t be too proud of this technological terror you’ve constructed. The power to process any data is insignificant to the power of customer centricity.

How can we get back on track?

Okay, this is the part where I plug the benefits of Tealium — it’s my soapbox after all and if I still have your attention, then I must be making some sense. Of course it’s not just Tealium that solves these problems, it’s most of the forward-thinking tag management providers. Tags are a great way to solve some of the fundamental data fragmentation challenges faced by marketers.

Fundamentally, tags can unify visitor event data, solidify visitor identification, and provide a common dataset for analysis and action to be taken. Contrast this to standard “big data” approaches to the same problem, which require you to pull everything back together with different formats, data quality, latency, and metric definitions. If you go down that path, you don’t have a “big data” problem you have a big (pause for effect) data problem.

In my next installment, I’ll spend some more time talking about the strategies for dealing with data fragmentation, and the benefits of unifying your data. Stay tuned!

Recommended Posts