Consumer Data and Text Analytics (aka “OMG, I can’t actually read all this!”)

In Customer Experience, Data & Analytics, Loyalty & CRM, Social media by Juliette


Consumer feedback online: How you can spot it, sort it, and react quickly, using text analytics

Seth Redmore, VP of Product Management at Lexalytics, has kindly agreed to produce this guest blog post, to explain our readers what consumer analytics exactly is, and what it can do to help you gain consumer insight and improve consumer satisfaction.  We hope you enjoy this content and join the discussion by sharing your own experiences with consumer insight and online data!

Whether it be a review on your website about your company or a giant Reddit thread about your product, most consumer feedback comes in the form of text. Even a voice conversation ultimately comes down to the actual words that are being used (with hefty seasoning of tone and voice stress, and things like that which help determine your mood).

Consumers are having these conversations either directly with you or between themselves online, in forums and on Twitter and Facebook, leaving comments on blogs, etc. They’re talking about you and they’re talking about your competition, and they’re talking about your partners, and they LOVE talking online, where you don’t have to use your real name, and even if you do, you can just broadcast your opinion out there without having to see a sea of faces listening to your every word. Anonymity and hiding behind a blind really makes a difference in how open people are willing to be about how they feel.

So, now you’ve gathered all this content together and holy crap that’s a lot of feedback, what are you going to do with it all?  You certainly can’t read it all – I mean, seriously, you have a job and all to actually “do”.

Text analytics will sort the feedback into different (user-generated or default) categories, extract important keywords like entities, & generate sentiment scores for each review/comment as well as all sentiment bearing phrases within every review/comment. Then you can roll up those results into eye-pleasing reports your boss is sure to like.

Once a text analytics engine has all of this feedback structured (which takes a few seconds to about a minute), it becomes easy to make decisions based on this feedback. For example, I’ve got 10,000 reviews of a new product and I’m responsible to find out what people don’t like about it. By running a quick sentiment analysis on the reviews, I can sort the feedback by polarity (positive, negative or neutral) and simply read the negative ones. I can also use text analytics to give me a report of recurring negative themes and find out what the most hated thing is about this product.

One use case comes from a company that delivers frozen food in the United States. They administer open-ended surveys to their clients and use text analytics to crunch the surveys – without reading all the content themselves. Upon processing the surveys they find actionable insight like “delivery late” was mentioned 210 times, and “thawing” was mentioned 89 times – leading them to tighten up their delivery schedules and work on the refrigeration in their trucks.

Many other tips, tricks and use cases about text analytics will be presented at the Consumer Text Analytics Briefing where experienced speakers will be sharing their thoughts at the Prospero House, London Bridge (London UK).  See the full programme and speakers here