Who’s Your Best Agent? Here‘s How To Get The Data
- Contact Center
- BY Ryan
- No Comments
If you had 5 seconds to name your top agent, would you have a name? If yes, could you provide data on why you chose the top agent? Is it that they are professional? Or does he/she produce higher than normal customer satisfaction levels? If no, how come you don’t know your top agent? You
If you had 5 seconds to name your top agent, would you have a name?
If yes, could you provide data on why you chose the top agent?
Is it that they are professional? Or does he/she produce higher than normal customer satisfaction levels?
If no, how come you don’t know your top agent?
You do realize; they are your most valuable intangible asset, don’t you?
We understand managing a contact center is no easy task.
In between dealing with emails, firefighting and wading through tons of spreadsheet reports, there’s never time left to sit down with your team to see who’s doing their best or who needs a hand.
What’s worse, to date, most contact centers focus on collecting agent attribute data, which involve facts like age, gender, work history, and education level. Although such data is necessary, it is not sufficient to identify behavioral patterns like professionalism, which indicate top agent-customer interaction.
Fortunately, the raw materials for monitoring and understanding behavioral patterns that make a top agent already exist in your call center. It is the data created when your agents exchange emails with customers and answer calls.
For several years, Call Criteria has been developing a dashboard that captures agent-customer interaction data in real time across platforms. The dashboard is designed to help call center managers efficiently review agent performance and access the data they need to make good decisions. Here are some of the things we have learned in the process;
Passive Collection of Data Saves Time
While you can always have your supervisors and managers walk around observing the agents while they interact with customers, the problem with this method is that you can only hear what the agents are saying and not the full conversation. Side by side monitoring also poses the same challenge with the added risk that you cannot rewind the conversation for more in-depth analysis. Both take too much time.
A passive collection such as silent monitoring and recording provide the best method to save time and money. The Call Criteria Dashboard captures data across digital platforms like email, social media, chats and calls allowing QA analysts to listen and review calls for a performance review.
“As a supervisor, Call Criteria gives you an overview of how your agents are performing based on a scorecard. The scorecard has a series of questions which have a point value associated with them. So as agents make mistakes, we mark them down. The result is that the calls have a score from 1 to 100 and you can see your agents ranked based on this data.” Chad Tatton, COO and Co-Founder at Call Criteria.
There are A Lot of Smoke and Mirrors in The Voice Analytics Space
Although there is absolutely a case for voice analytics, especially in passively reviewing agent performance and carrying out minimal quality assurance, its important to note that voice-to-text analytics is still inaccurate compared to the human ear, only reaching an average of 77% accuracy. Therefore, claims about voice analytics QA software being 90% accurate at detecting events, identifying compliance violations and complaints, 85% speech-to-text and a self-driven QA solution are just that — claims.
Combine Human and Tech for Accuracy
“The best way to ensure the highest accuracy in call center analytics is by fusing human analysts and voice analytics to review agent-customer interactions. With the use of voice-to-text technology, companies are pulling the calls that have various keywords stated in the recording to identify the right calls to review with human QA analysts.” Ryan Stomel, CEO and Co-Founder at Call Criteria.
Amongst other unique speech analytics processes, Call Criteria also includes a layer of voice analytics in the form of a keyword spotting tool. This tool uses intelligence to flag out specific calls for a review by human QA based on the set parameters.
For example, while some of our clients may want all of their calls reviewed, some may wish to review calls that feature specific keywords or behaviors, e.g. Mention of company name or ‘thank you.’
When a client wants 100% of their calls reviewed but does not want to pay for the human review, we proceed to transcribe all their calls into text and work with the client to identify keywords which may indicate that something is wrong or calls where particular words were left out.
So each time an agent says a word when they shouldn’t, a flag comes up. It is this flag count that determines which call gets scored by a human analyst to ensure maximum accuracy.
We use voice analytics to sift through thousands of calls in split seconds so our human analysts can focus on reviewing the flagged calls against a custom scorecard. The result is an accurate and comprehensive performance report easily accessible by your team on a dashboard in real-time. Furthermore, we feed the machine information based on the human scored calls. As we evaluate the calls, we can continuously train the machine to recognize a model and become more accurate with every human minute reviewed.
“As our client base grows, we look forward to a future where our machines will become smarter and will be able to accurately identify only 5% of calls that need human review.” says Chad Tatton, COO and Co-Founder at Call Criteria.
Dashboard is Key
Last but not least, we have learned a platform that visually highlights analytic insights moves data closer to contact center managers. The Call Criteria dashboard has been designed to ensure call center managers and supervisors have the data to make good decisions about how to leverage agents strengths. And they can do it at when those decisions points happen.