“Turning Text Analytics Findings into Working Solutions”, or “Microsoft Word Doesn’t Write Novels”

In the last 10 years in computing has acquired the ability to understand language. It’s produced text analytics software which is simply stunning. It can robustly and repeatably tell you every month what 100,000 of your customers were talking about in the freeform comments section at the bottom of your Customer Experience and Marketing surveys.

It can pick out the broad themes of their conversations.

It can even tell you when people are discussing a major topic indirectly – without them ever mentioning it by name.

But don’t get too excited. At least, not yet.

Grouping themes and determining topics of customer language tells you where the problems are. It doesn’t tell you what the underlying causes are (and it definitely doesn’t tell you how to solve them).

Just as Excel helps you group and analyse data but can’t interpret the intentions that produced it, text analytics (or text mining) software can’t understand what people meant when they said something.

So just as you need an experienced data analyst with Excel, the results of text analytics need an experienced linguist.

What precisely, for example, is making people unhappy when they say they’re unhappy about the topic they’re discussing?

On a recent project for a major retailer, our text analytics software determined that people were talking a lot about “managers”. And what they had to say wasn’t particularly charitable.

Using our experience in linguistics, we were able to shed light on why managers are viewed with such frustration and scepticism.

Originally, the title of “manager” came with the expectation that the holder of the title was entrusted to make decisions and exercise authority (or “manage”, as they say in Business). However, the tendency to call everyone a “manager” in the workplace (office and store) has confused consumers.

If everyone’s called a Manager, does that mean everyone has power or no one?

But when you encounter a problem and are being passed to a manager, are you about to get things fixed or merely bumped along the line.

So text analytics told us there was a problem with Managers. But it was only the linguistic insights that showed why people were having that problem.

And just as Microsoft Word never wrote a great novel, even linguistic interpretation of the Voice of the Customer doesn’t tell you how to fix things. 

There are two possibilities.

Route 1: Power struggles, widespread organisational upheaval. Much wailing and gnashing of teeth.

Route 2: Use language and naming to help customers navigate the store experience better. In particular, use creative writing skills to create job titles that tell customers what they can expect from that person.

Bet you can’t guess which one is easier, faster, and possibly more achievable.

To find out more about the latest developments in using linguistics, copywriting and Text Analytics to generate actionable insights for marketing, customer experience and Voice of the Customer, we’re always happy to share our opinion. Just like your customers. Drop us an email.