The world of social media marketing is divided into 2 camps: Those drowning, with no hope of analyzing big data, and those who wish they had some data to drive strategy. And, both are drowning trying to garner insights to guide their social media marketing strategies.
First, let’s reach some agreement about what big data is. According to Lisa Arthur, at Forbes:
Big data is a collection of data from traditional and digital sources inside and outside your company that represents a source for ongoing discovery and analysis.
Lisa goes on to underscore the challenges of understanding or agreeing on big data:
Big data is new and “ginormous” and scary –very, very scary. No, wait. Big data is just another name for the same old data marketers have always used, and it’s not all that big, and it’s something we should be embracing, not fearing. No, hold on. That’s not it, either. What I meant to say is that big data is as powerful as a tsunami, but it’s a deluge that can be controlled . . . in a positive way, to provide business insights and value.
I think she did a great job of underscoring the problem with trying to get organizational traction on analyzing big data. And, most businesses are drowning.
The big data avalanche
Nearly half of all businesses using big data feel they’re drowning and can’t glean insights from the massive amounts of diverse data types flowing at them with alarming speed.
Recommended for YouWebcast: Turning Your Website Into a Lead Generation Machine
In some organizations, the amount of data not only defies analysis, it strains resources just to STORE all that information.
No big data to analyze
Other organizations struggle with antiquated ideas of WHO owns corporate data and some marketers find it difficult to get ANYTHING from their organization’s IT department. And, what they do get provides few insights because reports reflect what IT THINKS marketers want, not what they really need.
Analyzing big data: Building a plan
Marketers need a plan to analyze big data and, that plan might involve struggling with IT to share relevant data. But, what goes into building a plan to analyze big data — to corral the mess so it makes some sense ? I’ve actually shared my ideas about this in several recent posts. In one, I share my process to building a data analysis plan using goals and KPIs (Key Performance Indicators). I also shared my 4-factor model for assessing social media performance. I don’t want to rehash that content, so please visit these resources to learn more.
Today, I’d like to expand on what I covered in earlier posts with some salient issues for analyzing big data.
Focus on insights
Certainly, using KPIs to help determine what to analyze is valuable. But, other data points might also help optimize your social media marketing strategy. It’s important, however, to ensure you’re only collecting metrics that lead to operational insights — identifying things you CONTROL and which impact your market performance.
One of the biggest problems I see when firms analyze big data is they focus on developing and not using them. If knowing something doesn’t impact the way you manage your business, then why invest time or money on analyzing it. For instance, knowing sources of traffic to your website is really helpful in determining where to put your marketing efforts. But, knowing how many fans you have is really a nonsense number with NO implications for management.
Don’t limit your insights to metrics contained in standard reporting tools. Dig deeper for insights. For example, knowing that a particular Facebook status update did much better than average is nice. But, understanding WHY it got better lift is critical for improving your market performance.
To further investigate why a particular Facebook status update did better, record all the variables related to the post: time of day, day of week, topic, headline, image, links, etc. You’ll have to score qualitative factors, such as image so you can adequately express them. Now, TEST each of these variables (A/B testing is best) to determine which factor or factors contributed to the relative success of this update. Knowing which factors are salient gives you information needed to create MORE updates that exceed average ROI.
I really like building regression models using the testing results because it’s likely several factors contributed to lift. Regression provides insights into the factors that affected lift the MOST. Often it’s a combination of these factors that will drive the highest return.
Don’t stop with the numbers
We know most data available in social media is qualitative — unstructured — data, such as words and images. Scoring is a nice way to analyze this data (and is the fundamental behind how most SAAS products handle unstructured data), but you lose SO MUCH information when you reduce rich, qualitative data to a number — like in sentiment analysis.
Don’t stop with analyzing numeric data! Try to understand the beliefs, emotions, and norms expressed in the utterances of consumers on social platforms. Such utterances are guides for unmet needs, problems encountered with your product, who influences buying behavior, and a host of other insights.
For instance, I analyzed a Disney forum and discovered that multiple folks mentioned the advantage of staying near a tram stop at a Disney resort. Disney used this information to decrease the average distance from each room to the nearest tram stop. Without such analysis, would Disney have ever realized that tram stops were an important decision variable for resort visitors? Probably not.
Don’t feel that every number has some meaning
Just look at Google Analytics. Here’s just a partial list of the analytics available:
And, with each metric, there are multiple data points such as percentage of total. Plus, you can always add filters, sources, mediums, dimensions, etc. That’s an incredible amount of data and Google Analytics is but one of the myriad of tools available. Likely MANY of these numbers have little meaning and others are beyond your control. Trying to assign meaning to everything is like looking for a needle in a haystack — your vision is clouded by a bunch of stuff that doesn’t matter.
Again, KPI’s help guide your exploration within your data, but it’s dangerous to be too narrow in your analysis as you might overlook valuable insights. For instance, if you’re focused too closely on metrics that impact your market performance directly, you might miss factors with an indirect impact on performance, such as complaints.
I once taught a bunch of MBA students from big companies like P&G, GE, and Toyota. They’d been taught to pay close attention to data by their data-driven organizations. However, they only paid attention to behavioral data — how much, when, how often. They believed I was the stupidest PhD on the planet because I advocated for monitoring consumers’ emotion relative to the brand and understanding factors BEHIND the numbers — they why of consumer behavior.
Unfortunately, too many organizations — big and small — fall into the trap of over analyzing big data and missing the subtle insights contained in it. Don’t get so focused on analyzing the minutia that you overlook the big picture.