Good Morning Everyone!
Has anyone replaced their agent's headsets/microphones since using CallMiner? If so, which ones work best for your company?
We switched soon after our go-live with CM. We currently use Plantronics Super Plus Headsets (HW261n). Our audio quality has improved which has greatly increased our accuracy within CM.
Let me know if you have any other questions.
Was your improvement in accuracy measurable? Eg. You were able to gauge the impact using data and drive a ROI with it?
We saw a ~10% reduction in category misses after the change. It closed the gap on a few of our scores, but I can attribute any specific ROI to just this change. We also had other agent level initiatives underway that could have helped the category accuracy, but this was the largest wholesale change we made during this period.
Thank you for the feedback! T
Can you share some of the agent level initiatives that you implemented to improve category accuracy?
We are currently working on demonstrating some test calls that will help drive home the idea that the agent needs to speak clearly to be understood properly by the customer and eventually CallMiner.
Kyle Carter, any best practices you can share?
Absolutely. Sorry for the uber late response.
We actually have a call center in Tijuana that we score using CM. I did the exact same rollout that we did internally (in Cincinnati), but did it via video conference. Our rollout meetings consist of a high level discussion on how CM works (transcription), how we build searches/cats and then calibrate into scores, what scores look like, etc. Pretty standard, and it was very successful internally.
All seemed well until a recent site visit, where management advised me that agent adoption was really behind what was expected. This was also around the time that CM scores were added to the agent scorecard, which affected their monthly incentives. Previously agents were held only to standard call center metrics (AHT, ACW, etc).
The agents were concerned that due to their accents, they were at an unfair disadvantage in comparison to US based agents. I listened to some calls on site and developed a training class/PPT presentation (15min) and Q&A session (30min) for the agents based on my findings.
I pulled back the curtain (slightly) by showing agents how our SA admin (Allyson Keown) works behind the scenes to add aliases to our syntax. I did this by showing them the syntax for our "Recording Disclosure" category (which is painfully long). I then played calls from one of their highest scoring agents, and let the agents watch the transcript as the calls played. Earlier in the day, this agent had stated to me that he has a very strong accent and had no idea how he was scoring so well. The secret to his success was that he articulates key words and phrases clearly, whether he knew it or not.
By showing the agents how someone with an admittedly strong accent was able to score well overall, not just compared to Tijuana agents, our adoption has gone thru the roof. We ask agents to change behaviors regularly and try to measure those behaviors with CM. Once they saw that calibration was an ongoing process on the back end as well, they completely bought in. CM went from being big brother to a collaborative effort.
Scores have increased 3-6 points in less than a month in all groups. We are incredibly proud of the improvement in this group and hope to continue the positive trend.
Let me know if you have any questions, and again, sorry about the late response.
No, but I'm **jealous**. We have old ones, but we have worked extensively on call clarity, mic placement, and background noise. It has helped.
Do any of you want to sell left over headsets to me at a discount, i need, like 1500. Oh they need to be bonded in India and Costa Rica, if you don't mind.
Richard Britt - Let me buy some Beats stock real quick and then I might have a recommendation on where you could get some headsets.
Wow!! What a visual! Next time agents complain about headsets, this picture will be a great reminder for a change of perspective!
Retrieving data ...