Why are writers such an oral lot? (Bad) text analytics on @26characters’ followers
I honestly believe that in the next couple of year, text analytics will blow marketing departments apart.
Briefly, here’s why:
When you see the amount of data consumers are offering about their attitudes towards brands (on social media, customer satisfaction surveys, emails to the company and in call transactions) you start to crave some way of processing it, to make sense of it. At least, the left-side of my brain does.
In the last 20 years, two things have come together: a greater rigour in the science of linguistics and more powerful (i.e. cheaper) computer processing.
Now there are a dozen software packages out there that will run language-processing on huge amounts of text, from your desktop.
In the most basic form, they provide sentiment analysis. Who, and how many people, liked it/hated it/were neutral.
In the more complex forms, they can give marketers an early warning on what their consumers think about their brands (or even better, what they hate about their competitors’ brands).
I’ve written more on this here, if you have five minutes: “What did text analytics ever do for marketing?”
In between these two extremes, Word Clouds give us a little nudge in the right direction. They let us see words that are commonly used, and give weight in the visual representation to how often they are used. Of course, context is important: people could be saying they hate that particular word.
“26” is the UK’s wonderful non-profit organisation dedicated to inspiring an appreciation of words in business and life. Yesterday, I looked down the list of their beautiful and talented Followers on Twitter and something about their bio’s caught my eye. There was a lot of love for writing, of course, but an equal amount of love for cake and tea as well, I thought. Could this be right?
I wanted to do some quick text analytics to make sure it was true, rather than the fact that I was looking at the list at 4pm.
I right-clicked and copied the bio’s from Twitter. I pasted the bio’s into Word. I used CTRL+F to merge all instances of ‘copywriter’ and a few other synonyms with ‘writer’. I also took out a couple of other words that would be necessary in anyone’s bio, (including the @ sign). I could have done better. But I was hungry.
I copied the resulting document into Wordle’s beautiful word cloud generator. I played with colours and horizontal/vertical alignments. And this is what I found:
This shows me a couple of things: writers love design almost as much as they love words. (Until now, I’d thought it was just me.)
Yes, they also love cake.
But then it shows me that the ways we describe ourselves are often very similar.
Perhaps Twitter bios are subconsciously a way for an individual to show which tribe they belong to.
However, for any writer who wants their bio to stand out on Twitter, the word cloud shows the need for us to find our own ways of describing something familar.
Which, in the end, is the real job of any writer.
I’d love to hear from you on what you think of this abuse of text analytics.
And if you know a psychologist who can explain why writers who love words also love cake, and why we’re such an orally-fixated bunch, I’d love to hear more.
I might even buy you a cup of tea next time we meet.