Is the whole greater than the sum of its parts in data analysis?

Posted on August 18, 2011 by Rebecca Denison

When we talk about social media analysis or monitoring, most of us think of something along the same lines. Analyzing mentions in social media to try to find insights, sales leads and the like. The focus is on finding mentions of a brand or its products in channels like Facebook, Twitter, blogs and forums.

There a few, however, already thinking up ways to use social media data for a very different purpose. Walmart is planning to use your social media habits to provide better and more relevant product recommendations. If I buy the latest album from Bob Schneider from Walmart.com, and you Google “top new music releases,” Walmart wants to show you an ad recommending the album your friend (me) just bought. Cool (and creepy), right?

The next big wave in social media data analysis is coming, and it has little to do with brand mentions. It has everything to do with you, the consumer. It would seem the next step is to move away from brand-focused search to customer-centric data aggregation.

There would be far more value in a brand mention if you also knew what brands the mention-er liked on Facebook or who their friend’s recommend to them. While we can derive a great deal of insight from pouring over singular mentions of our brand or our product, there is so much more to be learned from understanding the whole consumer, not the whole conversation.

Think about it this way: if I say, “I have to hand it to Duke, they’ve got a great squad this year!” does that mean more to you than if a Duke alum says it? (For the record, I would never say such a thing. Go Heels!)

If I’m known for disliking a specific brand or loving a specific brand, mentioning a competitor should have much more weight and meaning than a person who has never been part of the conversation before. In today’s world, it’s hard to connect those dots.

I’m sure this is someone out there developing a great new (creepy) tool that will let me not only see who is talking about my brand but follow those folks every move. And I cannot wait for the day that I will be able to use those kinds of insights in my work.

We are whole people. We are more than just singular mentions, and I can’t wait to learn more about the folks who are talking about brands and companies I care about. Imagine the kind of knowledge you will be able to learn.

Do you focus on mentions or do you focus on people? Have you considered mining and aggregating data this way?

About Rebecca Denison

Passionate UNC graduate (and basketball fan) interested in social media and measurement. As a biochemist-turned-communications professional, I spend my days as a senior social media analyst at Digitas in Chicago. Through my work, I have been able to establish social media monitoring and measurement best practices. I’m excited to explore more aspects of online measurement like traditional Web analytics, search metrics and integrated data models as I continue to learn and grow.
This entry was posted in Life as a senior analyst, Measurement and tagged , , , , , , . Bookmark the permalink.
  • Petula Neale

    I believe that some sites have started to do this in a more rudimentary way by asking the user to create a detailed profile of what they like. One example is the profile creation process on Google Boutiques where you make a series of selections for your view of the site,  and then continue to refine by tagging items as ‘Love’ or ‘Hate’.  Other Boutique users can follow you, and filter out users with a different style/body type/point of view. The Boutique approach is very onerous and manual. You do see more relevant items the more you refine, but it feels like you as the user is doing most of the heavy lifting. I would think the logical next step would be for the Boutiques site to aggregate this persona data and possibly re-sell the sort logic to the various retailer sites that get referral traffic. Those sites could then  push more relevant products to you based on the closest matching persona as you browse based on the item(s) you express an interest in based on time spent on the page. One challenge for this kind of approach would be when you are shopping for someone else – I don’t want to see apparel that would suit my size 2 friend and her tiny bottom if I happen to be shopping for her birthday on the same site where I shop for myself.

  • http://rebeccaadenison.com Rebecca Denison

    I know you’re right about that! There’s also Hunch which asks you all sorts of questions about what you prefer and like trying to make better recommendations based on what you tell it.

    I’d be even more excited to see the data integrated across channels. If I’m talking about looking for a car on Twitter and then jump to a random website, why not serve up a relevant car ad to me there? Understanding me as a whole person, across channels, would be wonderful and much more relevant.

  • http://twitter.com/JGoldsborough JGoldsborough

    I haven’t thought much about the “suggestions based on your network” integration into branded e-commerce sites, but I can definitely see the value in that. But what I think is really intriguing about what you describe is the ability to see what else people who are mentioning my brand are doing. And here’s why.

    Just saw a presentation from Spike Jones at FH (used to work at Brains on Fire) about ambassador programs and generating passionate WOM from true fans, not influencers. In most of the case studies Spike presented, the connections and relationships between fans of a brand came when the company was able to ID what its fans were passionate about and how their product can be a part of their lives. This type of insight can be expensive, but would be a lot more possible if we had the customer information you’re describing here. Cheers!

  • http://rebeccaadenison.com Rebecca Denison

    That’s a good example. Definitely along the lines of what I was thinking.

    If I’m friends with a bunch of fans of a competitor, that should affect how you look at my brand mentions vs someone who maybe is connected to a bunch of your own fans. I’m already digging into what other brands and things fans of ours like on Facebook to try to get a whole picture of the consumer.

    But what I really wish is I could be able to see that Justin G from Kansas City has spoken positively about us in the past and like baseball and dogs. I just want all the context I can have without being creepy. If that’s possible… :)