Richard Hall, a widely regarded guru of customer and sales support systems, received an unusual call in the fall of 2012.

At the other end of the line was Toby Shannan, VP of Revenue at Shopify, one of Canada’s most successful startups. Shannan wanted to pick Hall’s brain. Specifically, Shannan wanted Hall to tell him if he had a problem.

It was an odd request. Earlier in the year, Fast Company magazine had named Shopify one of the ten most innovative retail companies. More than 40,000 businesses in 90 countries had opened online stores on the Shopify cloud-based platform. Sales on the platform had sextupled in the last three years and were on pace to reach one billion dollars.

Did Shopify have a problem? Not by most companies’ standards. But Hall was curious. He dug in.

It turned out Shannan had good reason to be worried. There was no way Shopify’s customer service operations were going to scale to handle the company’s projected growth. Not only were the systems poky, but the workarounds devised by frustrated agents were wreaking havoc with the company’s ability to collect accurate data. Shannan himself didn’t even know how bad things had gotten. Instead of opening one ticket for each customer service contact and then closing it once the interaction was complete, the agents were queuing up five to ten cases at a time in an effort to beat the system’s latency.

The company’s answer rate of 90 percent seemed great. Then, Hall figured out the metric only reflected the number of calls that the system had determined could be answered by available agents——requests for help that couldn’t be answered weren’t being counted. “It was like giving yourself an A for all the answers on a test that you got right and not counting any of the answers that you got wrong,” Hall said. Meanwhile, meaningful real-time data about customer support wasn’t available. It took at least a day and sometimes two to get a basic report——for example, the number of interactions by day, by agent and by channel. “Everything they needed to make solid business decisions wasn’t there,” Hall recalled.

It wasn’t long after that initial call that Hall talked himself into a job as director of revenue operations. One of the first things Hall did was replace Shopify’s existing ticketing system with Zendesk. The problem with latency disappeared. For the first time, Shopify was able to accurately measure the productivity of its agents. Reports were available in real-time, in easy-to-digest charts and graphs versus CSV files. Hall could see exactly when customer service requests spiked, and staff accordingly. From Hall’s perspective, it was a great start. He had the granular data he needed to accurately forecast Shopify’’s headcount needs by day, by month and by year——data that wasn’’t available before.

But Hall had bigger plans. He wanted to apply the principles of time and motion studies to ecommerce. From a productivity standpoint, if Hall could reduce the number of systems that agents had to touch to solve a customer’s problem, they could handle more interactions. Equally important, studies had shown that if agents could solve a customer’s problem without forcing the customer to switch from one channel to another, agents could boost customers’ loyalty. Zendesk proved to be a great tool for achieving Hall’s goals. Hall put his development team to work. By using the Twilio API, they were able to extend Zendesk into a custom-built call center that handled not only email, but also voice queries, so information could flow more seamlessly. Another big project was creating a payments service, so customers weren’t forced to open accounts with a third-party payments processor.

Shopify can now boast having the lowest “click rate” in ecommerce —reflecting the simplicity and ease of its platform. And while clicks are unusually low, customer satisfaction and agent productivity is unusually high. Satisfaction scores hit 96 percent, even while Shopify agents handle up to 200 interactions a day. Nevertheless, Hall is quick to point out the quantity of interactions isn’t what is important. He encourages Shopify’s ecommerce gurus to take the time they need to help a customer. “We want every single minute of interaction with a customer to be a minute of true value,” Hall said. This approach has led to an unanticipated side effect. Instead of being a cost center, customer service has turned into a sales center.

Thanks to the accessibility of data provided by Zendesk, Hall can measure the effect his agents have on Shopify’s revenue. It’s helped him justify expanding the team from 18 people at the end of 2012 to around 60 at the end of 2013. Hall anticipates that the team will double again in 2014.

The data shows expanding customer service is not only necessary to meet demand—it’s also good for business.