Staffing for an unexpected support spike is a huge challenge for customer service managers. It is relatively easy to flex down by offering agents voluntary time off, but maintaining teams to handle abrupt volume increases of more than 10 percent is costly and inefficient. Yet without enough staff on hand, customers are left waiting longer than they’d like for help during a spike, and that doesn’t make them happy.
This a perfect scenario for on-demand customer service, and here is a great example of how it works.
Last week, a key feature on a customer’s website crashed, causing online support ticket volume to rocket 7 times over normal levels. The time to fix was unknown - it could take minutes or hours.
Often, a customer service manager in this situation has a dilemma: if you call everyone in, you still don’t have enough heads to handle the volume, and as soon as the problem is fixed, you have a surplus of agents on hand.
In this case, the company had partnered with Directly months ago to enable some of its best and most passionate users to become on-demand experts. Directly's routing technology automatically sends questions from the company’s Zendesk system to the appropriate experts, who use Directly apps to interact with customers. In the period leading up to the outage, the experts answered customers’ questions in under 30 minutes on average versus 24 hours for tickets answered by the internal team.
On-Demand Team Flexes
When last week’s spike hit, the surge of tickets instantly got routed to the community experts on their smart phones. They responded immediately, handling over 7 times the typical day’s support tickets and resolving 95 percent of that day’s tickets versus the prior week’s average of 85 percent. And get this: the average response rate that day was 48 seconds. A faster response rate and higher resolution rate on 7 times the volume? That’s amazing.
Why couldn't an auto-responder have handled the problem? For one, the crash was unexpected and didn't occur during the company's normal business hours, so there was no auto-responder set up. The second reason is a little more subtle, and it has to do with the personal nature of the on-demand help experience.
Customers Feel Good
After experts resolve a case, customers can rate their helpfulness. Typically, 25 percent of customers provide a rating. On the surge day, 36 percent of customers provided ratings, and 94 percent of those rated the expert interactions as “helpful” versus an average of 81 percent for the previous week. Remarkably, the experts couldn’t fix the customer’s problem! They simply reassured each customer that the company was aware of the issue and was working hard to fix it as quickly as possible. I believe that it wasn't the information itself, which, yes, an autoresponder could have delivered once it had been turned on, but the personal nature of a near real-time, peer-to-peer interaction that customers found so satisfying. People were there. They were real. They cared.
There are two clear takeaways from this example. The first is that on-demand customer service can flex to handle big, unanticipated spikes effectively. The second is that a fast, friendly, personal response from a fellow (expert) user can turn a negative experience on its head and instead, make people happy.