A great customer conversation cannot be scripted - it needs to be real and authentic. Every conversation must be unique. Agents should feel empowered to think outside the box, to make real-time decisions based on the situation.
The challenge then, is training on this, and scaling it across teams, timezones, and varying levels of agent tenure.
And, how do you create a call center quality assurance program with performance metrics and goals when it’s difficult to define what you’re asking agents to do?
Peloton and Zola have similar philosophies around how they create, train on, and scale unique and meaningful customer support conversations.
During the final panel session at The Art of Conversation Brooklyn, Laura Mundell, Director of Member Support for Peloton, and Rachel Livingston, Director of Operations at Zola, share their insights into how they support agents to be successful in this mission.
- Hire the right people that fit the culture
- Highlight good tickets (in addition to giving feedback on the bad)
- Have agents score themselves occasionally
- Empower agents to problem solve – rather than giving them a checklist of things to do, give them frameworks of excellence to operate within.
Here are the common strategies they employ to achieve success:
Hire great people
Both Laura and Rachel agree - empowering agents to be good listeners starts with hiring the right person. Processes and computer programs can be taught, but it’s hard to teach people how to be genuine, nice, and empathetic. This leads directly into culture. They continue to hire people that fit the culture that enables their success.
Highlight good tickets
Constructive feedback to agents is important, but don’t forget to highlight good tickets. Highlighting unique customer interactions that are very good will keep morale high and reinforce the preferred way for customer dialogue.
It also gives other agents an example of a really great interaction that they can learn from – something that could otherwise feel like an intangible, vague assignment.
In both cases, good tickets are highlighted to the rest of the company (not just the support team), showing the full team how important CX is, and making agents feel valued for their contributions.
Have agents score themselves
In addition to calibration meetings, self-grading helps agents better understand what actually goes into a high quality interaction (while also giving managers insight into how agents view their performance).
It helps agents be bought into the quality auditing process, and it also creates additional conversation and more bilateral discussion opportunities during one-on-one meetings.
Both companies strive to empower their agents to think for themselves. They give agents opportunities to go above and beyond to create next level experiences – they also give feedback, and grade tickets in a way that gives agents the confidence (and autonomy) to solve problems.
Rather than a checklist approach, rubric questions are asked in a way that leaves room for interpretation and room for unique elements of the interaction to be taken into consideration.
By hiring agents that fit the company culture, and empowering them to treat each customer conversation as the unique interaction that it is, both Peloton and Zola empower agents to have artful conversations with their customers. Agents feel included in the company process and vision, and their good work is highlighted when outstanding tickets are celebrated company-wide.
Here’s the transcript (edited for clarity and brevity) of the panel discussion:
To kick things off, please introduce yourselves and provide us some background on your company and the customers you serve.
I’m the Director of Member Support for Peloton. I’ve been with Peloton for almost five years. We bring fitness into the home.
I’m the Director of Operations at Zola. I’ve been with Zola for three and a half years. We’re an online weddings company that’s reinventing the online space. If anyone has ridden the New York City subway recently, hopefully you’ve seen our ads everywhere. Couples use our platform to host their wedding websites and registries. We also provide a service to their wedding guests, families, and friends who are buying them presents and attending their wedding.
In one or two sentences, what’s an artful conversation for your companies and teams?
We focus on three words that sound very simple but are very difficult to do: human, empathetic, and real.
We ask ourselves if we are providing an experience that we would want to have. In our call center QA scorecard, there’s a section for bonus points, and that’s one of the questions, “Is this an experience that you would want to have?”
What makes it mission critical for your company to provide such good support?
For us, the biggest thing is being subscription-based, that's our business, that's how we make money. We don't want people to cancel, and how do we do that? I think from the get-go, Peloton understood that it’s not just the music and instructors that fuel customers and keep them engaged. It’s also the people they exercise with. We see member support as an extension of that ethos to create a full customer experience around working out.
The one thing that we maintain with all our interactions is every agent should follow up. For us, we really stick by that, not just the once, but two times after that. We don't want to hound you and harass you or stalk you. But if we fix something, then of course we want to know if it's still fixed a week later and you're still happy with the product. We do that with every interaction, not only when something went wrong and was fixed. I think people appreciate that, to a point where it’s not “stalkery”, because some people enjoy that interaction.
For us, one easy reason is we have a lot of competition. You can make a wedding registry on Amazon, and you can go the traditional route like Macy’s and Crate and Barrel. So as a scrappy startup taking on a behemoth of an industry, having something that differentiates us from our competitors is really important. We want you to have fond memories of your wedding – our goal is to make it easier on wedding couples and relieve all the stress that they’re under when doing this.
We launched our free wedding websites a year ago, and we definitely had people who were saying things like, "I did my wedding website on The Knot, but your design is so much cleaner. I love how it integrates with your registry. So I'm moving my website to you."
In terms of operationalizing, what are some of the questions on your call center quality assurance forms, and how are they broken down to capture the essence of a good experience without having to literally ask, “Tactically, did the agent do X, Y, Z?”
Our call center quality assurance scorecard has general questions like, “Did the agent create an experience that represents our brand?” And within that we’ll identify key pieces of information that are relevant. Did the agent greet them by name, if they knew their name? Did they show empathy and sympathy? It takes this really big idea of, does this represent our brand, and gives general guidelines to the agents of what that means for us.
Another important piece of that puzzle is a biweekly calibration with everyone who grades tickets and phone calls. We do that for two reasons. First, to make sure everyone’s grading is consistent and adhering to the general spirit of how things should go. Second, to make sure that our rubric is still relevant for us. Since we continue to launch new products and things change in our operation, we ensure we’re still happy with our scoring and our rubric is still serving us.
We’re very much the same with calibrations and the usual things. The biggest thing for us is we like agents to be part of the scoring. We think it’s a huge part of helping agents to be bought into the process. One of the questions on our call center quality assurance evaluation sheet is, “Did the agent go above and beyond?” Everyone has a different expectation on what that means, so we use call center quality assurance software, MaestroQA, on the rubric to highlight what that means to us and what it looks like when done well. Rachel mentioned how you have to continually incorporate feedback or changes into the rubric. Some of the best changes we’ve had are from the team themselves.
What do you do to highlight the good tickets?
We have a regularly scheduled team meeting each week to go over product updates, etc. We also encourage agents to come with a good ticket or good interaction and share with the rest of the team, and we do shout-outs in our Slack channel.
We also do a monthly presentation to the entire company, where we put employee pictures on the final page to shout-out our top three agents. It might embarrass them, but we strive to talk about the good and not just the bad.
In the early days we created the Peloton Karma Room, which is now used company-wide for any great successes that are member-focused. If someone going through QA sees a great experience or great example, they’ll post it in there and it’s visible to the rest of the company.
Can you talk a little more about how you have agents grade their own tickets?
Yes, we have the agents score themselves, almost like an agent-manager calibration. It’s a great way to bring them into the process. The agent learns what it’s like to take that task on, and they can talk through it in a one-on-one.
We also have agents grade their own tickets – both good and bad. It’s interesting to see the self-grading, because sometimes an agent is much more critical of themselves than you are as a grader.
How do you have your rubrics set up in such a way that they empower agents?
We’ve tried to keep a simple tier system in place as we’ve grown. We emphasize the importance for us to empower front line agents, tier one, to solve every problem. We empower tier one agents to refund. So, we have rubric questions that ask not only did they do what we wanted them to do on paper, but did they recognize when they could have gone above and beyond. We empower the agents with all the tools to recognize and understand when you can really go above and beyond to create that amazing experience.
Similarly, we try to put as much power as possible into the hands of our agents. No customer wants to be on hold while you go speak to a supervisor. When an agent does escalate something, we coach through the conversation and ask the agent what they think is appropriate and a great solution. We’ve got a bonus question in our rubric that asks, “Does this interaction make me proud to work at Zola?” We want our agents to aspire to think about the level of service that makes them love being able to work here.
In many ways, the key to great conversations is really great listening. How do you empower agents to listen?
It starts with hiring, which feeds into culture, because it’s very hard to teach people to be nice and empathetic. So our interview process starts with a phone call. We ask classic interview questions, but we are really focused on, “Can you actually hold a conversation with me?” and how well does it flow. We also ask for a candidate’s best and worst customer experiences to give us insight into whether their vision of great customer service matches ours.
I love that. The people are the first place to start, right? We’re of a similar mindset. To be genuinely empathetic is really difficult. To be successful in a customer service role of any kind, you have to get enjoyment in your daily job from fixing things and helping people. We like to ask, “What is your dream job?” to see if a person understands and is confident in themselves. It also leads into culture, which is hard to maintain. As we scale to different offices and the team grows, we’ve continued to try to hire the right people for our teams.
How do you pick tickets for calibration?
Don’t overthink it, because part of it is just starting the process. I try to have a mix of low and high QA score tickets to calibrate. I put a recurring meeting on my calendar for a 30-minute calibration session, then another half-hour notice every week to pick out four tickets. If you’re searching for the perfect call or email you spend too much time on it. Also, don’t put too much pressure on artificially creating a discussion because it will just happen naturally.
How do you think about call center quality assurance training, and more specifically training the people who are QAing, on the brand voice?
We may not have struggled with it so much because agents have moved into QA roles. But the brand voice does change over time. We’ve learned how to isolate what the brand voice is versus what the brand of member support for Peloton is.
Since not all agents have the same tone of voice, and we want agents to have authentic conversations, we look to see if agents take the right opportunity to relate to customers. Were we apologetic at the appropriate time? When we had the chance to be excited for a couple’s engagement, did we take it? What you can do is take the adjectives and pieces that describe the brand voice and try to pick really tactical options to translate into the work we do.
If peers are listening to each other’s calls and perhaps taking phrasing, how do you push back against overusing the same phrases?
We promote agent to agent guidance. It’s great to see agents asking each other for advice. Sharing is something to promote and spin it in a better way. We don’t want agents to just use macros, since we hire humans. We want agents to solve problems with some personality.
Every time we have a new agent class come in, we rearrange seats. We always try to put a new person in between two veteran agents so they can learn and depend on each other. We always find that taking scripts or tidbits from other agents actually works out well, as long as it still feels like your authentic voice. We never want to talk to couples and their wedding guests and feel like we’re reading a script.