When will politicians wake up to text analytics?

During the 2008 US Presidential elections, Nate Silver used statistical data to correctly predict the winner of 49/50 states, and all 35 victors in the senate races. He analysed numbers to look into the future. Now we can analyse words to change it.

Mining for gold

Nate Silver showed that big data works even in the fickle world of politics. However, Silver’s power only extended so far; he was an observer, not an active participant.

Polls can indicate which topics are important to people, but people don’t think in numbers.

We think in words.

For the first time in the UK election, a small addition to the survey could give parties the info they need to fine tune policy and communications in real time.

All you have to do is ask people what they want to talk about….and make sure you’ve got some text analytics capability ready to process their answers.


Big Ears

There are alternatives to text analytics

Supermarkets have floating voters too

We regularly process and interpret large volumes of free form text replies for brands.

In one instance, working with a supermarket, we looked at the comments in the surveys completed by people who had shopped with our client but normally shopped with a competitor.

The floating shoppers talked mostly about the pleasant surroundings of our client’s stores and the service.

And they talked about these much more than the relatively higher prices.

This was a valued insight that had never been reached by the usual quant data research.

Our work offered them a new direction for their comms.

For us, text analytics is a combination of computer software and human intelligence. (You can read more about text analytics here and a case study here).

In the same way that Mr Silver can read statistical data more fluently than most people, so we use our linguistic expertise to interpret the data provided by text analytics software.

All data needs translating when transferring it to human beings.

Real opinions in real time

Text analytics will be as powerful a tool for political parties as it is for brands.

As well as polls, there is a phenomenal amount of text data generated spontaneously by political events. [1]

Tweets, newspaper columns, forums and Facebook, these channels are equivalent of the space at the bottom of the survey – ‘what do you want to talk about?’

It is real data, in real time, from real people.

The question is, who will be bold enough to use it first?

Just as Obama’s campaign team were the first to harness the power of social media in 2008, so the side that embraces text analytics in the next UK election will have an enormous advantage.

Click here to download our white paper on Text Analytics. And if you want to listen to what people think of you, email Chris.


BACK TO POST For example, during the first Scottish Independence TV debate, people posted 250,000 tweets about it