Unlocking the power of case swarming with Slack and Salesforce

The past two years have brought unprecedented challenges for businesses and their customers, leading to significant disruptions for customer service teams. Case volume and complexity have both increased since the pre-pandemic era, and agent attrition has skyrocketed.

What are the main causes of the growing agent exodus? Have customer expectations fundamentally changed since 2020? And how are today’s top companies adapting to better support customers and agents?

In our new series, we’re asking customer service leaders and experts to share insights about current industry trends and new tactics for success. Our first post features Jim Roth, the executive vice president of customer support at Salesforce, and Kevin Albers, the vice president of customer experience at Slack. They discuss:

• The growing prevalence of digital solutions in customer service
• Persistent challenges of agent burnout
• Tips for adopting a more collaborative swarming model to speed case resolution and boost agent satisfaction

What changes have you seen in customers over the past two years? Are they expecting more from businesses today?

JR: It’s definitely been an evolution. Early on in the pandemic, there was a lot more empathy from customers. They were just appreciative that we were open for business. We saw our customer satisfaction scores go to record highs.

Fast-forward to today, and I think expectations have started to increase again, particularly for digital interactions. Think about all the digital innovation that’s become pervasive during the pandemic—curbside shopping, delivery services, chat as a customer service tool—these B2C experiences that we’ve all really embraced over the past couple of years have now raised the bar for B2B experiences.

I see in our business a huge shift toward things like live agent chat, digital support and self-service. The acceleration of digital is really happening, and going faster than ever. So we’ve had to adjust, and spend more time and energy meeting our customers where they are.

As customer expectations continue to increase, how is that changing the relationship between service teams and customers?

KA: One trend we’re seeing is an added emphasis on understanding who the customer is. What industry are they coming from? Are there regulations that they might need to adhere to? How large is their organization? Understanding the source is extremely important, because it helps you start to eliminate some of the bias that might be embedded in the customer feedback you’re hearing.

We’re also seeing a trend in customers just wanting answers. Even if the answer is “no”—like saying we’re not going to build something—they really relish that level of transparency. So saying no is OK. It’s actually an opportunity to build trust with customers.

It sounds like customers want more digital interactions and more transparency from companies. What challenges do service teams face in meeting those higher expectations?

JR: Some of it is just having the right technology. Let’s say you didn’t have customer chat on your website, but your customers are driving you there because of what we talked about: different B2C experiences raising the bar. Traditionally, if you were going to offer a new channel, you’d have to go stand up a whole dedicated team just to service that channel, which is expensive and complex.

And so technology solutions, like being able to blend your agents across channels, certainly are huge enablers. You can turn on a new channel, but service that channel with the same agents you’ve got today. Having the right technology can definitely enable you to deliver on that promise and not turn your operations or your cost structure upside down.

Growing customer expectations over the past two years have increased both call volume and case complexity for agents. We know that agent burnout is a real issue, with attrition rates as high as 80% since the pandemic began. Given all these challenges, what are some ways to keep agents feeling empowered and motivated, so they can be their most effective?

KA: For the agent, increased call volume or call complexity really can lead to anxiety, confusion, frustration and burnout, especially in the world that we’re in today. Think about the last time that you had a long to-do list and how overwhelming it felt. Now imagine having to come to work day after day after day and having that same feeling.

Tooling is such a big part of this, too; there’s nothing worse than having to run through a handful of complex tickets where you actually don’t have what you need to be successful. It’s our job as leaders to put agents in the best position to be successful. We need to reduce the amount of cognitive overload that they might be facing, whether that’s by operational adjustments, additional resourcing, or technology improvements.

Finally, we need to think about how we can empower our agents to get creative and figure things out in partnership with leadership. They have incredible ideas. If you run low-risk experiments with them, you can learn what might or might not work. And agent engagement is going to start to rise, because ultimately they are—and need to be—part of the solution.

One solution we keep hearing about is transitioning from a tiered model to a swarming model, where instead of handing off cases from tier to tier, a single agent owns a case from start to finish, soliciting input from others as needed along the way. Jim, I know you recently piloted this at Salesforce and saw some impressive results. Can you tell us some of the core differentiators and benefits of moving to a swarming model for resolving customer issues?

JR: The traditional tiered model really has two problems for customers. First, when you pass a case from one tier to the next, by definition, that case gets in line again. It’s in a queue for someone to pick it up. So there’s a wait time.

Second, you have a loss of context. Even if you have the best CRM in the world and the best agent and the best case notes, something will be lost in the translation when you pass that case from the first agent to the next agent. The customer undoubtedly will have to repeat something, or some context will be lost.

A tierless model tries to address each of those problems. You retain ownership with that original case owner and then bring to bear the breadth of expertise—not only across your support organization, but across your company for edge cases—to help customers resolve issues. That’s going to reduce the time and the handoffs, which reduces the effort for the customer.

It does sound like swarming can speed up resolution time and reduce friction for customers. But how about agents? How do you see the swarming model really benefiting the agent experience?

KA: In a tierless system, it’s as if all your agents are in your highest tier. In that tier, agents do whatever it takes to solve every type of issue that comes their way. And as an agent, that’s going to be rewarding for a variety of different reasons.

One is around training. In this role, you’ll need a certain level of expertise, so you’re ultimately going to experience a level of investment from your company or your peers around the skill sets that are necessary to be successful.
You’re also going to start to have additional opportunities with other individuals outside of your function. You’re going to work with engineering, developers, product managers, finance, accounting, whomever. And that’s going to lead to an expanded experience and broader opportunities.

Finally, your most qualified or specialized agents are going to have an opportunity to take less complex tickets. In today’s world, that’s a good thing. Think about an athlete. As an athlete, your body needs those timeouts; you need rest. Halftime is a chance to regroup and recharge, and our brains are no different. So we’ve seen that quality actually increases across the board when your most skilled and experienced agents have an opportunity to reach into the softer-type issues and start to resolve them.

That’s great. In terms of rolling out a swarming model, what advice would you give? And what metrics would you track to gauge its success?

JR: If you have a tiered structure, I’d start by engaging your team. When they get stuck and they need help, where do they go? An internal search tool? Google? How does that process scale? I think what you’ll find, because this is what we found internally, is a need for more advanced troubleshooting help for your agents.

And if that resonates, you should definitely run a pilot. This is what we did about a year ago. We picked one of our customer service teams, and we put Slack in their hands. And they just kind of ran with it, and they loved it. Now we’re doing things like building some integrations with the underlying ticketing system to pull details into conversations and make them more valuable.

As for success metrics, I’d certainly look at the satisfaction of your agents in general, but also specifically with the tool and the process that you’re testing. That can be done through some simple surveys.

Also, look at what percentage of your cases required swarming—like how in a tiered model you might think about the percentage of cases that escalated to each tier. And then the biggest one I would look at is the end-to-end resolution time for those cases that required swarming.

We know that what customers want is an easy and expert experience. Those are the words that we live by. And there’s no doubt that swarming helps us deliver the easy part of it through faster resolution, but also the expert part by leveraging the wisdom of our entire team.

Stay tuned for more revealing conversations with industry leaders and experts who are shaping the future of customer service. And visit our new Digital HQ for Service collection to learn how you can grow true customer loyalty while bringing service teams together in Slack to deliver faster, more accurate and more personalized customer experiences.