The phrase ‘Big Data’ entered the public dialogue about four years ago according to Google Trends (which, interestingly enough, is a form of Big Data). Perhaps inspired by a decade and a half of books like Freakonomics, Moneyball and the Tipping Point, companies began mining petabytes of data from the cloud to better understand and predict the world around them. Perhaps that’s also when big data storage and analytics tools finally became accessible enough for large business to have and use.
Whatever the reason, we are now in an era where business value is tied to large quantities of data and numbers. Making decisions based on big data feels more reliable, and more honest. The trick is uncovering which data matters most and how to apply it in the best way.
Let’s take a look at the contact center. Organizations have massive volumes of data about their customers. With the tools and brain power to analyze it, there is significant opportunity to leverage this data to make the contact center more efficient and create a better customer experience. Unfortunately, some organizations choose to use these resources to focus on things like refining upsell opportunities in other channels instead of using it to improve their Interactive Voice Response (IVR) system to drive productivity and customer satisfaction.
The reason for ignoring the IVR in these cases seems almost like a Catch 22 – why waste big data resources on an old-school channel with low customer satisfaction (CSAT) numbers? However, that’s exactly why companies should think about investing those resources in the IVR to increase those CSAT numbers.
For good reason, it is now almost a mantra that big data can make your contact center stand out as a leader. For example, the ability to predict caller intent by looking at mobile behavior will put your contact center at the forefronts of innovation and customer experience. What can you do if a big data-driven IVR investment is not yet on the horizon? There is good news, your business can still create an amazing IVR with the “small data” already at your disposal.
By “small data,” I mean the data that is within the jurisdiction of your IVR team and any vendors charged with maintaining your IVR. Small data can include things like speech science analysis and call logs, both of which are comprised of quantitative data at a smaller scale than the millions of data points we see in big data sets. It can also include qualitative information like usability test results and even the knowledge inside the head of user interface designers who have worked on projects with many other businesses like yours.
To use a simple example, we recently helped tune a healthcare company’s IVR menu to improve routing accuracy. After collecting speech data over the course of one month and modifying the IVR accordingly (adding synonyms of pre-existing options, adding entirely new menu options and changing the “confidence” of the recognizer in various states) the company increased the number of callers able to seamlessly schedule an appointment through the IVR by about five percent, enabled nearly eight percent of callers to check on an appointment, and increased the number of calls able to successfully check their benefits by a whopping 47 percent.
As another example, Nuance recently worked with an investment company that wanted to simplify the task of selling shares. Funds are traditionally hard to deal with in the voice channel because there are so many ways to say them. Together we decided to analyze back end data to find out how many funds the typical customer had. It turned out more than 85% of their customers had only one or two funds in their account. This allowed the team to come up with a solution, letting 85% of those callers select from one of two funds.
“Which fund would you like to withdraw from? For example, you could say ‘sell my shares from [their first fund]’ or ‘sell 300 dollars from [second fund] and send me a check.’”
Other examples of small data that could drive a positive impact include the collective experience of staff and vendors or the branding research your marketing department has conducted. So while big data can floor your customers with its ability to predict their needs and learn from past interactions across multiple channels, small data can be just as impactful in driving higher levels of customer satisfaction.