Is This The Best We Can Do?
The data available today can give us a picture, but it can easily turn into the wrong picture if you do not consider some important points.
Karin Robinson on Responsible Data Analysis
Like Karin Robinson, Associate Director for Social at OgilvyOne Worldwide, I too feel as giddy as a schoolgirl when I start thinking about the sheer volume of data available to me to help me analyse my followers, potential followers, engagers, and influencers on social media. I can lift the lid on my social platforms (to the extent they allow me) and start really understanding who follows me, whom I should follow, and why.
But also like Karin, I am as grumpy as an old man when I start thinking how agencies and organisations are presenting this analysis to their clients in order to make strategic decisions about social marketing.
My husband @skipfidura is a data guy. He likes to say, “I don’t believe any statistic that I didn’t make up myself.” That’s a really flippant comment, but it has an echo of truth in it which Karin painstakingly pointed out today in her #SMWLDN session titled, ‘Lies, Damn Lies, and and Social Statistics: Why Raw Data Can Tell the Wrong Story, And Why That Matters.’ The amount of data available today can give us a picture, but it can very easily turn into the wrong picture if you do not consider these important points:
Demographics: Age, Gender And Location. Like Dr. House says, ‘Everybody lies.’ Karin reminded us that people will lie about their age (are there really no Facebook users under the age of 13?), they can hide their gender, and they can change their location data. These 3 important factors can seriously skew your understanding of your followers. For example, Twitter analytics told Karin that 69% of her followers are male. Her own labour-intensive analysis proved the opposite: 33% are male, 39% are female, 17% are neutral (organisations, media entities, etc.) and 11% are unknown. Don’t believe the numbers unless you dig deeper!
Sentiment. Brands tend to focus on how many followers they have which gives them an overall value of their popularity. Karin rightly pointed out that people are more complicated than machine tools allow for; analyzing for positive, neutral or negative sentiment delivers an incredibly narrow view of your brand’s popularity. For example, these tools can’t track emotions such as confusion, optimism, defensiveness, sarcasm and annoyance. And they certainly can’t address that posters can feel more than one way within a single post. People are complex creatures; ipso facto, our posts are as well.
Visibility Trap. I loved that Karin has created her own phraseology to describe this phenomenon. She’s referring to the fact that there’s so much data available in one entity (Twitter) that you can easily feel as if you’ve gotten a perfectly clear picture of your followers; however, there are lots of other systems that do not provide you with this plethora of information. It’s the classic idea of putting all your eggs into one basket, and Karin wants us to not fall into this trap; she scarily commented that over 71% of all conversations aren’t visible (she calls this Dark Social). She also remarked that nine-tenths of social media users are not regularly posting on social media. Forgetting about these individuals is giving away an entire collection of people who may be genuinely interested in purchasing from you.
After ripping off our very comfortable blinders to the fact that we’re missing a huge chunk of our population, Karin handed us a virtual cuppa tea. She asked us to be more thoughtful and work harder to dig into the data for a real complete picture. Do your homework. Look at your customer data, your Google traffic and other information you have in house. Match all of it up to get a real understanding of your followers, prospects, fans, customers and more.
Personally, I don’t think she’s asking too much. Yes, we can do better, and we have a responsibility to ourselves, our clients, and the social media industry to do just that.
Read More About Social Media Strategy on Digital Doughnut