Big data - or just data in general - is a topic that marketers love to talk about. Working with data is also something easier said than done. Especially when you have a lot of it.

We do a lot of social media listening at DeMoss for several clients. When you are in the business of reputation management like we are, having a pulse of what is being said online (and where) about all of our clients is clutch. Not only has this information helped us with crisis management situations, but has also helped our clients better refine their own social media strategies and prioritize channels where they already have highly-engaged audiences. Being better informed about social media behaviors of people online has helped us be more strategic in providing digital counsel.

Despite the fact that the tool we use (Brandwatch) helps us organize massive amounts of data into easy-to-read dashboards, the information overload can still be overwhelming. We have the ability see any public post ever created about a client - which for some brands is quite a bit of data. When I'm creating a report for a client, picking out which data points to include can be daunting. Out of all the information available, which information could lead to actionable next steps?

I try to tell a story

Think of it in a real life storytelling scenario. Last year a corporate jet crashed behind my house. When I told that story here, I include the details that drove the story home. I included the loud jet noise, feeling the explosion and the fire near our back deck. There were a lot of other things going on simultaneously during that time, details that were directly related to that story. I could have mentioned the temperature outside, the reaction of my next door neighbors, the news trucks and emergency vehicles blocking roads in our neighborhood, or even all of the smoke Megan had to drive through to get to our house after the fact. However, those details weren't important for the main story I was telling - the social media impact of that event happening near us.

I had a plethora of information but chose a few details to tell a more concise and relevant story.

Journalists do the same thing (for better or worse). They may be covering an event and have a lot going on around them. Even though they could report on every detail they see, the only details that appear in the news are the ones that actively contribute the story they are trying to tell.

Data can work the same way

Nobody should ever get into the business of data collection for the sake of data collection. Everyone wants to say "oh yeah, we work with big data" but it means nothing if there's no value coming out of it. Through social listening or even something like lead capturing, you can proactively try to learn things like:

• What is actually being said on my brand?
• Where are people talking the most about my brand? Is it where we proactively spend time?
• How did people react to this news story?
• How did our most recent campaign affect our overall brand lift?
• Through opt-ins, which channels do people voluntarily want to hear from us (SMS, email, Twitter, mobile app push notifications etc.)?

Plan what you want to learn before you try and learn it.

Sure, you may find a few surprise bits of information that are interesting and worth investigating along the way. However, without a focus or purpose, telling that story will feel impossible.

Once you know what you are trying to learn, do everything you can to gather information on it. Afterwards, tell the story of what you learned and how you went about doing it.

When it comes to data, marketing and PR should have a "scientific method" approach:

• Ask a question
• Do some preliminary research to draw a hypothesis
• Perform an experiment (do the data collection via campaign, online listening query, etc.)
• Analyze your data (what happened and why)
• Draw a conclusion - tell the story of what you learned

What do you think? Is there value for brands to collect data without a clear intent or purpose of how they'll use it?

How else could you incorporate storytelling into big data?