Make your customer feedback speak the right language

Take a quick look in your wallet for a receipt from a high street retailer. The chances are that it will invite you to email your comments about your recent experience with that brand.

Retailers such as McDonald’s are sent hundreds of thousands of pieces of feedback like this each month. In a few years’ time, you will be able to point your smartphone at a billboard in the street for it to download the product information and ask you to enhance the content and submit feedback.

So if social media networks are driving an increase in the volume of customer feedback and sharing, how do brand owners harness this vast amount of information and use it to drive more insight into the brand and the customer experience? And what do you do if you’re a global brand, like HSBC or Shell or Motorola, your customers communicate in over 50 different languages and you need to get the feedback out to your regional brand management at the beginning of next week?

Brand owners and researchers are finding verbatim feedback an increasingly powerful tool. Firms are experimenting with new ways of asking customers about the brand and customer experience, replacing repetitive and numbers driven studies with open ended questions asking people to articulate their own views and feelings about the product or service, and using new tools to analyse the information.

Often, software is now used to analyse the frequency and content of verbatim comments. It’s an approach that works really well if many of your answers have a highly predictive content. Customers of a train operator for example, will talk about punctuality and overcrowding on commuter routes, while long distance passengers tend to comment about the quality of the food service onboard and the cleanliness of the bathroom facilities.

That all sounds pretty straightforward, but there’s a debate going on as to whether automated coding programs can accurately capture comments that are broad and diverse. Or those that have a myriad of medical or financial terms.
Complicating matters further is how you compare feedback from several languages – the software algorithms that process the responses will be different. An esteemed Arabic survey respondent ’Enam Tresni’ turns out to be simply, ’insert name’.

One way to solve the problem is to analyse the feedback in the original language. This has the advantage of keeping the subtle local nuances close to the original source of the product or service for action by the regional brand management or customer services team. If the client is centralised, then you can back-translate a representative sample of the responses into English for analysis centrally.

Social media, verbatim comments, word of mouth, feedback – whichever name you apply – is starting to have a major transformative effect on how marketing campaigns are managed. Campaigns are executed very rapidly and brand managers no longer want to wait for lengthy analysis; they need information quickly in order to act on it. Look no further than Gap, who launched a new brand last week only to withdraw it five days later after a deluge of negative online customer comment.