Making Big Data Smart

It should come as no surprise that the digital analytics and marketing industries are talking about big data. After all, big data is a very exciting and popular topic, with many industry experts writing and speaking about big data, and why it’s important. All of this content and popularity is good for the big data movement; however, the conversation could still be improved.

By its definition, the collection and analysis of big data is supposed to be large, complex, and difficult to work with. From the outset, this already seems like a conversation about to get extremely technical, costly, and time consuming (I can see the “non-data geek” executives in the room checking out right about now). In my experience, none of these attributes correlate well with successful projects. This is why we should shift the conversation away from big data, and talk more about smart data.

So What Is Smart Data?

Smart data is a subset of big data and should be a part of every big data project. The goal of smart data is to quickly show a return on investment and lay the foundation for ongoing big data projects. More specifically, smart data has three attributes:

  1. Smart data requires at least two data sources. Just like big data, smart data will combine at least two previously disconnected data sets.
  2. Smart data is efficient. It looks for the largest potential impact, with the least amount of resources.
  3. Smart data is actionable. The analysis and insights that come out of your smart data project need to have a clear objective, and one that can be acted upon in a fast, efficient manner.

How About an Example?

A great smart data project to start with is to incorporate marketing campaign data from a digital analytics tool, such as Google Analytics, within a CRM system, such as

For our example, let’s assume I run a B2B company that uses Google Analytics to track my marketing campaign performance. Because I don’t sell directly on my website, I’m unable to tie sales back to specific marketing campaigns. The goal of my example is to connect marketing campaign data with my CRM so I can compare the value of the leads I get from different campaigns. Once this information is available, I can make better data-driven decisions on where I should invest more, and where I should better focus my optimization efforts. The process I will follow is:

  1. Capture the data. To start you off on the right foot, I’ve included some example code that will access the __utmz Google Analytics cookie and store the campaign value of a visitor. With a little time and research from a good developer, this can be updated to capture a number of different items, such as traffic source (Google, Bing, etc.), traffic medium (PPC, email, etc.), or keyword.
        <    script type="text/javascript">
      function _uGC(l,n,s) {
      if (!l || l=="" || !n || n=="" || !s || s=="") return "-";
      var i,i2,i3,c="-";
      if (i > -1) {
      i2=l.indexOf(s,i); if (i2     <     0) { i2=l.length; }
      return c;
    var z = _uGC(document.cookie, '__utmz=', ';');
    var campaign = _uGC(z, 'utmccn=', '|'); 
      if (campaign == null) {
      campaign = '';
          <    /script>

    (Don’t worry if you don’t understand the above code. Find a developer and make that person your new best friend. Seriously. A good developer can make magic happen and make your efforts a lot more effective. Also, I didn’t write the above script and make no claims to such. I can’t remember where I got the script from, but I have used and modified it several times, so I can verify it works.)

  2. Pass the data. Now that you have the campaign data stored in a variable, your developer should create a hidden field in your contact forms and pass the variable into the hidden field for when someone submits a form.
  3. Collect and analyze the data. After you’ve confirmed campaign data is populating in your CRM records, make sure to give enough time for the results to be an accurate reflection of typical prospects. For example, if your average sales cycle is 30 days, don’t try to analyze any of your new data until you have enough converted leads or sales to accurately gauge the effectiveness of certain campaigns.

Project Wrap-Up

For about 10 hours of effort, you should be analyzing valuable data you previously didn’t have, and be poised to make serious optimizations to your campaigns. But don’t stop there; take your smart data project further by building on what you learn. Do you get a large amount of qualified leads from paid search, but little sales? Maybe your sales staff needs some training on the differences between paid search prospects and cold calling prospects. Maybe certain sales reps respond to certain campaigns or traffic mediums better than others. What else can you learn from the data?

Once you’re done with one smart data project, reassess where you can make the biggest impact with the least resources and start again. Chances are your organization will have a big data culture before you know it.

Big Data image on home page via Shutterstock.

Related reading

Flat business devices communication with cloud services isolated on the light blue background.