Imagine a customer leaves this comment for a hotel.
“The hotel really messed up on this stay, they put me in a mediocre room and then charged me for someone else’s meal, but to their credit they picked up a couple rounds of drinks to make up for it and promised to do better next time. They better be perfect next time.”
There is an obvious coolness factor to software that can understand that this comment is bad for a hotel chain, but not catastrophic because the customer will be back. Beyond “coolness”, one must ask if there is sufficient business value in an engine that can mine sentiment and determine intention, to warrant the investment of time, resources and money to bring a system like our Salience engine into an organization. The best way to answer that question is to see an example of how the software is leveraged and what sort of ROI it brings the organizations that leverage it.
Let’s look at a Voice of the Customer example to see how a tool like Salience can pay for itself.
A client of ours is a VOC provider that has lots of customers in the fast food industry. They used Salience on the feedback data of a fast food retailer in the Midwest that was seeing a drop in sales in Western Indiana. Using their VOC platform powered by Salience, they were able to examine customer feedback data and see that the drop in sales was being driven almost exclusively by a significant falloff at two of their stores. Then, looking at the sentiment around these particular outlets confirmed that customers weren’t happy, but by digging deeper they noticed something strange: the sentiment wasn’t universally poor at these outlets, it was up and down. Further digging led them to realize that 2 of their managers had basically checked out, and were just punching the clock and not worrying about the operation.
OK, nice example, but if the consumer sentiment is that bad at those properties wouldn’t they have figured it out anyway by a random sample review of their customer feedback? The answer is, of course they would. Software like Salience isn’t smarter than a human, in fact it’s not as smart as a human, but it is a whole lot faster and more consistent. In the above example, you can use a VOC system to spot an issue early on, and in many cases stop the loss of revenue that would have dragged on for a while.
The previous example was around spotting sentiment, but the mining of text goes way beyond sentiment, as do the business use cases. Salience can also be used to spot user intentions, things like intention or desire to purchase, or switch providers, or even do harm. Imagine the value to a marketer if they could generate a list of folks they knew were looking to purchase a flat screen TV in the next couple of months, that’s a pretty high value target list. The point I’m trying to make is that ROI is everywhere because text is everywhere. Every business uses text (emails, forums, twitter, Facebook, inter-company documents) to make business decisions every day.
An engine like Salience doesn’t do anything that we can’t already do by ourselves, but that is in fact what makes it so valuable. It can act as your surrogate and amplify your ability to cover this sea of information that flows over all of us every single day. Don’t think of Salience as a magic box of functionality, think of it as a lot more eyes that can alert you to things you need to take a look at.