Tuesday, February 14, 2017

Speech Analytics Is Not Your Fairy Godmother

By: Diana Aviles




As we welcome the New Year we see an increase in people finally jumping on board and investing in Speech Analytics (SA) technology. Among many other things, my job is to be an advocate for SA in how it helps organizations see the big picture both in their call centers and with their customers.  People like me want to encourage organizations to invest in SA but at the same time we also find that it is our duty to be honest and forthcoming about the varying aspects of Speech Analytics.

This point brings me back to the title of this post - “Speech Analytics is not your Fairy Godmother”.  All too often I hear organizations say, “We want to have an SA program here,” as it sounds rather fancy and seems like it might be useful. Yes she is a very powerful ally who can offer you deep insights into the happenings of your business, but she does require proper investment to get the very best from her.  This certainly goes beyond simply dropping a lot of money on the software and expecting “magic” to happen.  For example, you would not just randomly search for things the same way you would Google directions to a particular Mexican restaurant and expect instant results.  The magic you seek comes from the individuals who know how to utilize the tools in the software as SA programs can actually be individually sustaining.

Automation is fantastic!  Personally I enjoy the quick convenience of running to Stop & Shop in my Chicago Bears pajama pants at 8am on a Saturday morning, rolling up to the self-checkout station and not dealing with any human being other than myself.  There are limitations to automation however.  There are times when gaining insight is not as easy as sending an automated email export to yourself.  Are there ways to set up some aspects of automation?  Certainly!  

Generally speaking most SA tools have options for you to set up a tag or a report that is run against queries for almost anything.  But queries can only tell you if they hit for something particular- on their own they cannot provide you with context.  A query run on “Email Issues” that occurred on 10/25 may very well return a high amount of hits.  If someone is not paying attention to trends within data and is only automating a report based on the query alone, she would not be getting the full benefits of Speech Analytics.  For example if a spike in volume for email related contacts is observed, that could be a huge sign to look into a situation to investigate what may have happened to cause that irregularity in the data.  From there, with the knowledge procured, based on let’s say an outage that took place with the email service on that date, I can empower an organization to take action on a situation.  SA tools are designed to do your heavy lifting and a good SA program with appropriate support can get you where you need to be if you allow them to do it.

The unrealistic expectation of automation may dampen organizations experiences with SA and as a result SA could potentially be viewed as fluff software when in reality it is simply misunderstood or undervalued.  Other instances find organizations getting overwhelmed by SA from ingestion/transcription related issues or systems administration.  With that being said, there are a few things you can do to prevent yourself or your organization from falling into the Fairy Godmother complex:

Have a set of goals you want to achieve with Speech Analytics – What do you and your organization want details about? When you’re first starting out it is perfectly normal to start small.  Rome wasn’t built in a day.

Don’t throw your eggs into one basket – Be open to having people trained in SA. When you have more than one person trained it greatly helps as it allows people to collaborate their experiences to provide better solutions.  In my tool I am always learning new tricks from others while simultaneously discovering new tricks of my own to share with others.

Play around – Tinker around with your SA tool.  Training is great but at the end of the day the only way you can eventually master something is by experimentation.  I have been exposed to Speech Analytics for 5 years now and those years of tinkering have helped immensely.  I can arrange metadata in more ways than one could count and even in my sleep (which is sometimes problematic because a normal human being shouldn’t be dreaming about excel spreadsheets.)

Trust your Speech Analytics software partners – Your Speech Analytics tools oftentimes come with support.  In many instances these people were a part of the software development or they come ripe with the experience of managing other SA clients.  Some organizations are distrustful of them as they may hold perceptions of a “secret salesman” mentality.  Trust them.  You are paying them to help your organization.  Let them consult and help you get you to where you want to be.

Thanks for reading- if you have any questions or comments please feel free to leave them down below!




Operations Manager, Speech Analytics

With more than 5 years of Quality Assurance experience in a call center environment, Diana's objective is to simultaneously promote and educate the world of Speech Analytics with a human touch; one which further emphasizes the importance of First Call Resolution and overall customer experience.

Follow Diana on LinkedIn.

No comments:

Post a Comment