Follow Us

Like what you're reading?

Subscribe to receive periodic updates about new posts by email, or follow us via Twitter or RSS.

Please enter a valid e-mail address to subscribe.


You have subscribed.

How Dell Increased Support Productivity by 25% Through Effective Chat Training


Dell Banner

Live chat utilization has risen 24 percent in the last three years as customers increasingly prefer online channels to phone-based support. This is a positive trend for companies—live chat allows agents to handle multiple customers at once, which increases support productivity. However, in order for agents to do this without the customer knowing they’re one of several being helpled, support organizations need an effective training program.

Dell is one company that provides successful live-chat training to its agents. Since the launch of its training program seven years ago, the number of issues handled per agent has increased by 25 percent, and customer satisfaction levels have remained the same or improved.

To uncover tactics Dell uses to train its employees, I reached out to two executives on the company’s customer service team—Manish Chhabra, a technical support director, and Monte Tomasino, executive director of consumer operational excellence. Below is a summary of their “secrets to success.”

Require Agents to Graduate to Chat Support

Dell agents never start in chat support, even if they had previous chat experience before being hired. Instead, they must first prove themselves as a top performer in traditional phone support, and then “graduate” to earning a spot on the chat support team.

“We’ve discovered that our most productive voice agents are better suited for the more multitasking-heavy chat environment,” Tomasino explains.

He says a live chat role is one agents aspire to because they can listen to music while they work, and it’s a less stressful position—agents don’t feel the same pressure to respond to customers immediately as they do when communicating via phone.

This is a clear benefit for Dell: agents work harder to demonstrate that they’re ready for a live chat position, and once they get there, they work hard to keep their position. It comes as no surprise then that Zappos, another company recognized for its overall support savvy, also requires agents to “earn” their right to move to chat as a way to motivate its team.

To evaluate whether agents are ready for the transition, Tomasino says Dell considers the agent’s typing skills, grammar and writing abilities. Additionally, the agent must be among the fastest and most accurate at choosing the correct content from the company’s knowledge base. This is because sharing links to content is often the best method for solving customer issues in a text-based medium like live chat.

Teach Agents to Coax Customers Into Specifying the Issue

Chhabra says one of the biggest challenges of live chat support is effectively getting to the root of the issue. Customers tend to type as little as possible, which means agents need to learn how to effectively coax them into revealing specifics about the problem.

“Take Internet connectivity issues, for example. The customer might say, ‘my Internet is slow,’ but the cause of that issue could be one of 10 or 20 different issues,” Chhabra explains. “It’s up to the agent to figure out what the cause is specifically in as few chats as possible.”

Dell agents are limited to asking two questions per chat (or offering two potential solutions per chat), so they must be very selective about what they ask. For this reason, the first of the four modules in Dell’s chat training program focuses on how to choose the right questions to effectively identify the real source of the customer’s issue.

In this module, agents are taught to use Dell’s interactive “decision tree”—a system that essentially uses keywords from the agent to suggest questions to ask the customer until a solution is reached. While these questions are automatically generated, agents can’t rely solely on the system—they must think critically about whether the suggested response does in fact align with the customer’s issue, or if they need to try a different keyword.

“Think of it sort of like a GPS guiding you to the right answer,” Chhabra explains. “If you put in the wrong address, it will take you to the wrong location.”

To learn how to use the decision tree to effectively coax customers into revealing the source of the issue, live chat trainees shadow an agent using it. They do this at least once per day, each day of their training. This gives trainees a chance to see what kinds of specific questions agents ask to find the cause of the issue (e.g. asking for an error message, or details about what the customer was doing immediately before the issue occurred).

In the Internet connectivity scenario, for example, the trainee might watch the agent immediately narrow down a list of 20 potential solutions by first looking at the customer’s purchase history, then by asking a follow-up question that is specific to the device they own.

In addition to shadowing, agents are presented with chat logs from real customers who all had the same issue, but asked completely different questions. This gives them a chance to see the entire progression, from the customer’s first chat to the probe for the real cause of the issue. During these sessions, instructors highlight instances where the agent could have asked a different question to get to the cause of the issue faster, as well as those in which the “ideal path” was properly identified.

Demonstrate When a Link Is Better Than an Explanation

Effectively handling multiple chats at once is all about perfect timing—customers are more likely to notice that they’re one of several being helped simultaneously if they’re left hanging for too long.

Chhabra says one of the keys to timing chat interactions is knowing when it’s better to explain how to solve the problem versus sending a link to content that describes the appropriate resolution. Learning this skill is the focus of the second module of the training program, which teaches “customer handling skills.”

“Imagine the customer asks the agent how to use the new Windows 8—the agent could spend 10 minutes trying to understand exactly what they want to know about it,” Chhabra explains. Instead, he suggests that agents in this scenario send customers a link to “A beginner’s guide to using Windows 8,” a three-minute video that explains the basics of the program.

In this scenario, the agent would ask the customer to watch the video, then chat them when they’re done to see if they still have questions. This frees the agent up for three minutes so they can focus on wrapping up an issue with another customer, and may resolve the first customer’s question with minimal effort.

However, if the agent picks the wrong time to send a link (e.g. if the solution requires the agent to first pull the customer’s account information, or the content isn’t specific enough to the issue), the customer will likely come back frustrated that the suggested content didn’t solve their problem. The agent must then spend even more time trying to find a resolution.

Similar to Dell’s training sessions on how to coax more information from the customer, agents learn when they should and shouldn’t send links by viewing chat logs. They’re shown instances where an agent incorrectly suggested a link to content instead of explaining the issue, as well as instances where a link was used really effectively.

An “incorrect instance,” for example, would be one where the agent continued the live chat for more than two exchanges after sharing the link, which would indicate the content wasn’t effective in resolving the customer’s issue.

Provide Rules for Setting Customer Expectations

Dell_Golden_RulesIn addition to solving customer issues quickly, Dell places a high value on providing customers with a positive live chat experience—so much so that they tweaked their chat performance metrics a few years ago to consider these more qualitative measures.

Among these measures, management seeks to mitigate instances where the customer appears to realize they’re one of several being helped, or are worried the agent is no longer on the other end of the line.

“If I was to do a quality audit and found a customer saying something like, ‘Hello? Are you still there?’, that would be considered a negative experience,” Chhabra says. “The key is to set the right expectation with the customer.” Setting the right expectation essentially means three things:

  • Reassuring the customer that the agent will be responsive;
  • Letting the customer know if the agent needs extra time to find necessary resources; and
  • Clarifying that the agent is really listening, and will wait patiently for the customer to explain their issue (rather than rushing them through the interaction).

For example, if an agent is handling two or three customers at once, they can prevent the customer from feeling neglected by saying, “Please give me about 30 seconds while I look up a few resources for you to access.”

Similarly, if an agent provides a customer with a link to content, they can say, “I will be waiting here, just chat me when you’re finished watching the video.”

Agents are also advised not to interrupt customers. Even if an agent has 30 more seconds before they need to check in with another customer, they should never chat a customer who is in the middle of typing. In this instance, the customer might feel rushed, and that the agent cares more about closing out the chat than listening to their problem and finding the right solution.

To ensure agents set the right expectation with customers, agents are taught specific guidelines they must adhere to during the “soft skills” training module (which also covers Dell’s “Golden Rules” for chat etiquette, pictured here). Agents are then tested on these skills with an assessment at the end of the training day. This might be a written test or a mock chat session with the trainer.

After Training, Provide New Agents With Support

While chat agents are immediately put to work after training, they aren’t left to themselves—they always start off with another chat agent guiding them through every interaction and providing tips for effectively handling customers.

“One of the biggest issues we see impacting productivity is new agents lacking confidence,” says Abhiroop Basu, content strategist with live chat software maker Zopim. Like Dell, Zopim’s new chat agents are teamed up with another agent after training.

“Pairing them with a veteran helps them gain confidence in their abilities because they know they have someone there to guide them if they’re unsure about their response,” Basu explains.

At Dell, a day in the life of a new chat agent typically lasts about seven hours. At the end, a trainer pulls them aside for a “team huddle.” During this time, the trainer will review their day and ask the agent how they felt they did, if they have any questions and if there are topics they feel they need to cover more extensively. Over the next three to five weeks, the new agent is increasingly left to handle chats on their own, and team huddles are slowly phased out.

“It’s like your first day of flying,” Chhabra says. “You aren’t left by yourself all at once. The instructor is right by your side, until eventually he lets you take full control of the wheel.”

These are just a few of the best practices Dell uses to train its chat agents for maximum efficiency. What methods does your team use? Please share your thoughts by leaving a comment below.

Banner image, “Dell World 2012,” created by Shawn Collins CC BY / Resized. Thumbnail image, “7644427706_f1cc47d62c_b,” created by the U.S. Department of Agriculture CC BY / Resized.

Share this post:  
Ashley Verrill

About the Author

Ashley Verrill has spent the last six years reporting and writing business news and strategy features. Her work has been featured or cited in Inc., Forbes, Business Insider, TechCrunch, GigaOM,, Yahoo News, the Upstart Business Journal, the Austin Business Journal and the North Bay Business Journal, among others.

Connect with Ashley Verrill via: 
Email  | Google+