Marketers Relationship with Big Data: It’s Complicated

Big Data Cartoon

Marketing is struggling to adapt to big data. Today, content is being created faster than it can be consumed and our attention spans are not getting any longer. The amount of content available on the Internet is increasing and shows no signs of stopping. This leads to a corresponding amount of data to be processed, stored, and analyzed. Every piece of content is packed with data that may or may not be valuable to a marketer, but the only way of finding out is to independently inspect it – a daunting proposition.

How do you determine what data is important? How should you measure big data?

Unfortunately, there isn’t a perfect answer to either of these questions. I’m going to focus my answer on making the most of social media and the Internet, but big data extends well beyond social media. Big data is disrupting all aspects of life. Think of all the data being created on a daily basis: companies track customer purchases, fitness gear tracks personal health and nutrition, and doctors and hospitals are tracking and storing medical information.

So back to big data and social media. To figure out what data is important, determine what you’re measuring. Are you just trying to measure volume on a topic or is sentiment more important? Maybe you’re only concerned with what a select audience of influencers is saying instead of the general population. When we think about big data we should be thinking about ways to cut it into smaller, more consumable pieces.

Let’s take a single tweet and think of all the different data points that might exist within those 140 characters.

  • Author- who is tweeting this?
  • Sentiment- is this tweet positive, negative, neutral, or mixed?
  • Location- where is this tweet coming from?
  • Timing- when was this tweet published?
  • Engagement- what sort of engagement did this tweet get?

I didn’t put these topics in any particular order, the marketer should decide which of these elements is most important. A single tweet is not enough to create insights or draw conclusions, but all of the data packed into a single tweet shows the complexity of the issue facing marketers today.

Author: the author of the tweet can show whether or not a campaign is reaching the correct audience. For example, a female tweeting about male razors might not be as important to marketers as a potential male customer tweeting about the campaign. Another consideration is whether or not the author is an “influencer.” You could create an entire blog trying to define what influence means but it’s certainly a factor to think about when examining content. For simplicity’s sake an indicator of influence might be a celebrity or person with a large social or real life following.

Sentiment: boiled down, sentiment asks if the person is speaking positively or negatively about a brand. The issue with sentiment on a single-tweet level is a lack of context. One tweet might be positive, but a previous tweet may have been negative or more positive, so the change in sentiment is where the significance lies. Additionally, when dealing with mixed sentiment it might be difficult for a person or machine to draw a distinctive conclusion or boil that down into a single number.

Location: the location of the tweet might be significant because if you’re a brick-and-mortar retailer and a person is tweeting about terrible customer service from a store it could have major implications. On the other hand, if I’m targeting a marketing campaign to Chicago and someone from New York City is tweeting about it, should I care?

Timing: similar to knowing the location, it’s important to understand when a tweet occurred. If someone is tweeting about an event six months after it occurred, should I care? When measuring the impact of content or examining a single tweet it’s important to understand if you’re reaching the target audience.

Engagement: engagement metrics such as replies, retweets, and favorites are good indicators of high quality content and potential impact. A tweet with 1,000 engagements might be more relevant to marketers than a tweet with 10 engagements.

In closing, there is no silver bullet for measuring and understanding big data and social media. Marketers should focus their analysis on what’s significant because in a world flooded with data it’s easy to get lost in a sea of numbers.