I almost never complete customer service surveys. Most of the responses go into a deep dark hole never to be looked at by anyone from the company.

At one of my previous employers, I implemented a customer survey system. It was an extremely simple value proposition. Complete a 10-minute survey and be entered for a chance to a great prize. We quickly ramped up to over 6,000 surveys a week from customers. The survey results confirmed everything we knew…WE WERE AWESOME!!!

Well… at least that’s what the survey said. We had 97% of respondents say they would recommend us to their family and friends based on the proven Net Promoter Score measurement.

Every week, the results were the same. I was hoping for some variability…some place to go improve, something to fix… something. The whole project felt like such a complete waste of time and money. I was thoroughly frustrated.

However, I later found all of the jewels in the survey were in the free text responses at the end. People would consistently rate us high on service, but all of the opportunities showed up in the open question section. It was valuable when I read through these, but there was zero probability of me keeping up with such a vast amount of data.

What I didn’t know at the time was the power of text analytics. I am not talking about a simple word count exercise you can do in excel; I am talking about true text analytics that can help give you real insights into your KOLs or team members.

You can read a lot about text analytics, natural language processing, machine learning, etc. It can be complicated. Let me break it down with a simple example of what is possible.

Mary has a dog and Mark had a cat. 

Today’s analytics solutions can break the sentence apart into all of the nouns, verbs, sentence parts, etc. The system can understand that Mary is the one with a dog and Mark used to own a cat. Additionally, it knows that Mary and Mark are most likely connected somehow. You could ask the system “How many women own canines?” The system can be intelligent enough to know Mary is a woman’s name and dogs are part of the canine group.

Text analytic systems can understand concepts and how they are connected to other concepts. You can even understand customer sentiments such as “Do people like our product?”  Imagine being able to analyze all of your free text from surveys each day looking for trends, issues, or opportunities.

Had I known the power of this technology five years ago, I am convinced I would have found incredibly valuable insight within our customer surveys to improve the service in our organization without having to read thousands of responses each week. Text analytics won’t get you to a 100% answer, but if you do it well, it will point you down the right path and increase your ability to analyze by 1000X.

Text Analytics in Medical Affairs 

These systems can be a great solution for helping medical affair leaders sift through a large number of insights submitted by the MSL teams. Additionally, if you apply text analytic principles and systems to your MIR submissions, you will be able to tap into another vast source of insights from KOLs and HCPs. We recently helped a client evaluate various text analytic options for their business. I would enjoy sharing what we learned along the way. Feel free to reach out with questions or if you would like help on your own text analytic project.