© 2019 by Ganduxer Ltd

Using sentiment analysis to speed up innovation

Thursday, December 7, 2017

 

I often write about the process speeding things up. Probably a by-product of being an American who has lived in Europe since 2000 and being a generally impatient person that likes to get things done. Although I love my life here and have thoroughly enjoyed my time in Europe I always have a nagging feeling that it could be "better" if we just sped things up just a notch. If we apply this very general thought to innovation, how could this play out.

 

One of the critical steps to good innovation is the early steps around design thinking. Mapping the challenge and deeply understanding the core problem you are solving is critical to finding the right solution. Without this first step right, many a projects have gone on to create solutions that are in need of a problems to solve and fizzle out fast. I am certainly a fan of getting this step right but how could we speed this up using better tools?

 

There are new wave of very accessible, very powerful sentiment analysis tools on the market. These are normally focused squarely at the customer experience teams to understand customer sentiment from various surveys, complaints data and live feedback. These tools can mine through comments and disaggregate customer's actual comments. For example, the comment "the food was great but service was terrible", the system realises that this is not really a happy customer, but in fact 2 separate comments that need to be treated completely differently. They also use handy data visualisation tools (like word clouds) to articulate the key learnings quickly.

 

For those of us in innovation, numerical scores (like NPS) only provide a signpost to customers feelings, but the sentiment analysis can prove to be invaluable source of insight. Having this kind of tool at your fingertips, you can now sift through massive datasets and look for trends, outliers and drill into these for better understanding. What is also great is these tools can sift through all kind of unstructured data: Twitter, Facebook, Glassdoor, employee surveys and internal ideas portals to start to give a very joined up picture. What we are really looking for is where both employees AND customers see issues or solutions. We can then focus our design thinking efforts to get under the skin in these posts and start to refine/fix the customer journeys.

 

From this analysis we are also able to attach a WHO said this. We can then do some segment analysis to further inform our analysis before we even have that first detailed interview, focus group or sit on a cold street corner to get some primary research. 

 

I would say that this approach allows innovation teams to do very detailed desktop research in a fast and effective way before even looking at the problems in detail. Dare I guess, it might even speed up the learning process :)

Please reload

Recent Posts

Wednesday, September 25, 2019

Please reload

Search By Tags
Follow Us
  • LinkedIn App Icon
  • Twitter B&W