Home > Columns > Executive Interviews

Zenarate Executive Interview

 Click on the image below to view the ebook

CrmXchange's Managing Partner, Sheri Greenhaus and Casey Denby, Senior Account Executive, Zenarate, discuss Zenarate's human centered AI Technology that augments human performance through their AI coach.  You can visit Zenarate in booth #522 at CCW 2022.  

Tell me about Zenarate.

Zenarate is a human centered AI Technology that augments human performance through our AI coach. AI coach is akin to a flight simulator for customer engagement. Whether it's on the sales or service side, inbound or outbound, or any industry, agents can practice to mastery with any scenario that they could possibly imagine.

We can simulate any type of voice or chat contact engagement through Zenarate AI Coach. In a simulated training environment agents have access to an augmented and immersive role-play before they take their first live call or chat on the production floor. Zenarate’s AI Coach is intelligent, patient, unbiased, and working 24/7 to take the stress and the inconsistency of the agent training equation.

What are the use cases?

There are 2 primary use cases. The 1st is new hire in training. The 2nd is ongoing coaching or upskilling of any tenured agent.  These agents are not in a production environment when training with AI Coach

Agents are evaluated through a quality or an analytics platform. When an issue is identified; maybe the agent that is not good at empathy, or does not get the compliance right. During downtime, they go off the production floor into our simulator and practice. They are assigned what we call an AI Coach story; they log in and practice. This training helps close skill gaps or helps with the launch of new products and services.

Typically, call center training is a knowledge article that they are reading while they're on the phone with the customer. Today, most agent training is done by pulling the agent off the phone.  This method of training is not efficient or cost-effective, especially if you have 5,000 agents across 17 countries and 35 languages.

With a simulated call, agents can go onto the platform and practice the call before they go into a production environment. They practice call flows as many times as they need to master best and required practices, so contact center leaders can be confident that agents know just how to handle any customer call that they would get.

Does the company look at the analytics to determine which agents need more training?

Our platform has embedded simulation training analytics and trainers can listen to simulations to score whatever is practiced in the simulator for an added human element. Zenarate AI Coach is not tied to production performance. Rather, contact center leaders will use production analytics to pinpoint skill gaps, like if and agent does not score well on empathy, training leaders will recommend they practice empathy on the AI Coach simulator to get better. Remember, you don't get better without practice.

We're seeing that we're really the only ones in the space offering voice and chat simulation that is almost an exact replica of what you would encounter in a production environment.  It’s the practice of the chat and calls over and over that brings agents to mastery. We perfect the role-play experience for front-line associates. You can think of Zenarate as a behavior reinforcement solution.  

Is the tool used for screening during the hiring process?

Think of a BPO hiring hundreds or thousands of agents a month for a communication job. These communication roles are essentially the face of your company. Ninety percent of customers that call in say that a contact center interaction has a large bearing on their loyalty to a company. Why would you put somebody who has no communication skills on the phone?

Our platform has a guided version and an unguided version of the simulation. The guided version is akin to training wheels while riding a bike or bumpers when you’re bowling. It literally gives you real-time next-step guidance for what you're supposed to do and say on the call.

The unguided version is just like having a conversation. There's no guidance of you telling the agent exactly what to say. AI Coach uses Natural Language Processing allowing agents to use their own words without scripting, and Natural Language Understanding generating real-time customer responses simulating a customer on any topic or persona.  The level of difficulty can be changed for each call type as well.  

New hire screening is the last step. Literally all they do is get in, put a headset on, press the play button and communicate with their AI Coach, which sounds like a human voice … literally telling them what to do and say next.  It’s a great way to find out who is the best fit for this type of job.

Do agents get a score for hiring? Does it flag or grade them?

Zenarate AI coach has robust reporting and analytics capabilities. Our analytics provides a simulation scorecard along with a variety of other reports. The simulation scorecard is for chat or voice. It tracks the number of practices, the AHT, the back and forth, and then it tracks each step of the call flow. Did the agent pass that step or did they not? If an agent fails at any point, the system gives them in-the-moment AI coaching, giving agents the ability to adjust on the fly. From a chat perspective, we have even more visibility!   

How many languages do you support?

Our simulator is available in 13 languages for voice and concurrent chats.

What is best practice for the total number of chats an agent should handle at one time?

Most people are going to do 3 to 7.  You’ve got to be able to handle a minimum of 3 to make chat a worthwhile channel for the organization. I would say best practice is you should be able to handle 5 conversations at the same time.

Do you measure the average time to respond in chat?

We track average time to first response, fastest response time and slowest response time.  You can see it visually via a heat map. If your acceptable limit is 2 minutes between chats, and the agent is at 2:33, it's going to get a deeper red on the heat map. It shows each individual chat, which ones you're doing well on and which ones you're not, and then it goes down into the number of interruptions, as well as which step of the call you were interrupted on.

What reports do you provide?

We provide a various number of standard reports and we can also customize them. This allows each individual client to tell us what they need.  We can add or change according to what each client requires.  We can build reporting APIs if they prefer to do reporting on the raw data themselves in Tableau or Salesforce analytics, or whichever platform they prefer to use. Reporting comes standard.

For customers who are using your chat solution, what are they seeing?

To start, they are seeing if the person is a right fit to be a chat agent. There is also less time between each individual chat and chats are more accurate. I think the biggest hurdle for many companies is that they think all voice agents can be great chat agents and vice versa.  With our system, companies are able to assess if this person will be good for chat. We call simulations stories.  Each story can have a chat or a voice simulation, so agents can practice it either way. 

Our collected customer data  shows on average +56% speed to proficiency, +31% call quality scores, +22%, higher conversion rates whether that’s collections or sales, +33%,  -32% in attrition, and this all leads to a 23% net reduction operating expenses.

You reinforce learning through behavior reinforcements, allowing agents to learn by doing. It is proven science that 70% of how we learn is through doing. If you introduce a topic, then practice, the retention on that subject is 75%. If you don’t and you go from topic to topic, the retention is between 5 and 10%. And as you continue on in topics, it continues to decrease to 0 if you don't reinforce the learning. We actually encourage people to review earlier topics as they're continuing on their learning journey. And then, as you add new topics, the retention is really high. That topic mastery is what they are taking to the production floor.

How do you know if agents are production ready?

The agent logs into what we call a master story which has many scenarios. The agent clicks play and they have no idea which story or call they're going to receive. It could be any one 30 to 50 stories created during the AI Coach pilot implementation. The agent needs to prove that they can meet best and required practices for all of the stories.  We then look at the report to see if the agent is ready for production.

Companies build in an acceptable number of interruptions and acceptable number of duration (such as handle time, etc.) that is allowed during simulation to certify agents.  If I'm a BPO, I want to be able to tell my senior leadership and my customer that we're certifying top performing agents and that they are certified with what is needed for the job.

We work with some of the top telecoms and they’re trying to prove that they are certifying and putting the most capable and ready agents that are representing their brand on the production floor.

Do you have a case study?

Bank of America uses it for over a 100,000 customer-facing associates across 8 different verticals. Every vertical they have in the company uses our solution: collections, debt, credit card, fraud, customer service, general customer service, and frontline banking associates. 

Are the stories customized?

Stories are tailored for each client according to their best and required practices. We have a process that we go through to build the stories and tailor them for each client. We take role plays that you would do in training and create them in the AI Coach Simulation environment.

How long would it take if somebody's says yes, we want to pilot this? How long does it take to get going?

Zenarate's AI Coach requires no personal or account information and no IT integration for pilot or enterprise deployment. Generally, you can have your AI Coach platform ready to launch within 6 weeks, and you can start to practice right away.

What problem do you solve?

Most companies that are AI focused are focused on chat bots taking the place of customer engagement agents.  We're focused 100% on improving human performance. We're actually a human-centered AI company. McKinsey has a quote that I like to use. They said, ‘the demands facing contact center staff are becoming more complex, more varied in more valuable. That calls for a smarter coaching approach’.  We are the smarter coaching approach. We provide enhanced augmented AI coaching and training so that the harder job now becomes doable or manageable. People are now facing higher regulated and more complex and difficult calls.  It’s a very hard job.

Many of our customers that do pilots with us say their 0-to-120-day attrition is the worst; even the first and second months of production see high attrition. Agents are like, ‘I can't do this. I'm out of here’. We almost eliminate most of that because number 1, you're hiring the right people, and number 2, they know what the job is before they do it.

Agent speed to proficiency is a very common problem. Through leveraging Zenarate AI Coach, agents can practice simulations in a safe learning environment leading them to be more confident and prepared much faster – shortening the speed to proficiency timeline to have agents ready for the production floor.

How do you train hybrid staff?

We are able to scale and provide a more holistic, inclusive, and engaging training experience. Remote and on-premise agents bring in different challenges as well. How do we make sure people are actually learning when they're not in person; how are they staying on task? How do we make sure an agent is not multitasking on his cell phone? With our simulator, you can track every click and utterance and behavior and you give them the actual practice to stay engaged.

At Zenarate, we are focused on how people learn, following actual brain science of how human beings learn and become proficient in new subjects. Our Net Promoter Score is through the roof with agents using AI Coach. After agents are exposed to Zenarate AI Coach they always want more simulation training. In fact, we have heard that agents and training managers want Zenarate AI Coach it to become half of their entire training program. This is all because we designed AI Coach to reinforce their learning, by giving them the first ever training solution that encourages agents to learn by doing the job, and because they're able to track improvements versus written test taking or just listening to a presentation.