Operationalising text analysis as part of your NPS program
In an analysis of 1000s of customer comments for a healthy meal kit and delivery company we found the word ‘hair’ appeared in 0.5% of all comments. Naturally, it was predictably negative.

What’s an acceptable percentage of your total paying customers to receive food with human hair in it? In an analysis of 1000s of customer comments for a healthy meal kit and delivery company we found the word ‘Hair’ appeared in 0.5% of all comments.

Naturally, when the word ‘Hair’ appeared in customer comments it was predictably negative:

“Easy to follow program! Love the new lunches option. If I could give any negative feedback it would be the black HAIR found in my carrots with hummus”

“There was a long black HAIR in the lasagna. Enough to put me off Company Name for life. I expected a higher level of food handling and hygiene. Buyer beware.”

I have been a customer for a long time now, but was a bit annoyed about my pumpkin soup order this week. I was out for lunch on Monday and decided to have my lunch option tonight for dinner. A bit of a shock to find black HAIR floating in my soup (not mine as I have red HAIR). I have a photo as well if required. Very annoying and expecting a refund. Regards, First Name Last Name


While the above is certainly interesting, the true power in the analysis is how we can now operationalise these insights:

  • Find the Needle in the Haystack - Word clouds show the most prevalent words, but often the greatest insights come from hidden words like ‘Hair’. The most prevalent word/s are easily found but are rarely the most important.
  • See Themes not Words -  Word clouds produce ‘words’ not ‘themes’. Maybe the word 'Hair' should be categorised under a wider theme of 'Hygiene' or 'Food Quality', which cannot be done via a word cloud. Normally, groups of words ladder up to tell the true story within the vast unstructured customer comment data.
  • Operationalise the Insights - Word clouds are a static visualisation that don’t link up to your wider customer and operational data. Once we’ve identified words/themes identified, we can deep dive into the wider data to identify exactly how, when and where these themes are occurring at a product, process, or person level. We can isolate the issue and take corrective action. We also know exactly which customer reported the word/theme, so we can also take corrective action with the individual customer, to stop negative online comments that may dissuade others considering using the product or service, and have an opportunity to resuscitate the current customer.
  • Continuously Monitor It - Once identified as a ‘risk’ word or theme, we can set an ongoing alert to inform us immediately whenever specific words (e.g. ‘Hair’) appear in customer comments so we can take the required action at the organisation and customer level.

Your products, processes, people and competitors all change overtime.

Active Feedback

An active Customer Feedback Program that leverages operationally driven Text Analysis will help you see changes at the macro level, but also ensure you can see and act on changes at the touchpoint, location, employee and customer level, so you consistently deliver on and maintain your brand promise at all levels of your organisation.

For a limited time we’re offering a free analysis of your customer comment data. Get in touch if you would like to learn more.

Like to know more?

Book a time to discuss how we can help.