When you think of ecommerce, lunch is probably not the first thing that comes to mind. But in today’s world, where our smartphones are rarely out of arm’s reach, people purchase pretty much every product online. Including sandwiches.
Subway, the world’s largest restaurant chain with nearly 45,000 locations worldwide, certainly understands this. Chad Sanderson, Experience Optimization Manager at Subway, recently spoke on a ClickZ Masterclass about conversion rate optimization and A/B testing, which Subway uses to better understand its enormous customer base.
“Our audience is so massive that we have a lot of segments and data. How can we use them to give people the best possible experience?” says Sanderson. “That’s kind of the crux of optimization and personalization: There’s not just one user who visits the website. Everyone has different behaviors so it’s virtually impossible to understand why one particular person is doing something. The point of testing is looking at larger groups.”
The webinar may have only been an hour long, but our conversation with Sanderson continued. Read on to learn about Subway’s “metric that matters,” what question you should ask yourself, and how to create a representative sample when your audience is… well, everyone.
ClickZ: Subway isn’t a brand that naturally seems to lend itself to ecommerce. How does CRO play into your strategy?
Chad Sanderson: In the [quick service restaurant] industry, more and more people are ordering online and then picking up in-store. We want to have an optimization program in place that makes that process very easy. It’s been great for us—it’s efficient for customers and when they use a mobile app, they have to sign up. Now we have more information about people, which gives us a better understanding of them, as opposed to the people who go to the store and use cash that we know nothing about.
CZ: During our webinar, you mentioned that finding “the metrics that matter” should center on the metric that matters. What is Subway’s?
CS: There are a few: the number of purchasers, revenue per visitor, and elements of loyalty and time. We don’t want to just maximize what people are buying right now; we want to maximize how frequently they come back over weeks and weeks. Our ‘metric that matters’ is a combination of all of those elements that’s kind of our North Star.
There’s such a huge amount of data that when you don’t have some sort of guiding force that works as an anchor, it becomes absolute chaos. There are a lot of proxy metrics that only sound like improvements, but don’t improve the ‘North Star.’ “Let’s optimize time on-site” may sound good, for example, but sometimes the number of items people add to their bag increases while the checkout amount doesn’t.
CZ: The majority of our webinar attendees aren’t satisfied with their website conversion rates. What can they learn from Subway?
CS: Subway is a massive brand that people have affinity for, so it’s easier to have better conversion rates, regardless of our user experience. But what we do is ask ourselves, “What is a good conversion rate for me?” I think that in general, there’s too much of a focus on the industry average and how much higher that number should be. The focus should be about understanding what is an optimal space for your business.
CZ: Your name and brand recognition brings up a good point. When you’re testing, how do you come up with a representative sample of your audience when your audience is so big?
CS: The good thing about statistics is that it’s really broad and all-encompassing, and can always be molded to fit the sample size. It’s convenient for Subway because when you have a larger the sample size, you can make very specific inferences and you don’t have to make them off a huge amount of your proportional traffic. We can also infer something in a much smaller time interval, rather than saying, “We need to get people in here for a month first.”
CZ: Smartphones obviously represent a huge change in consumer behavior. How has that impacted your CRO strategy?
CS: The move to mobile is gaining traction every day, but there’s still a contingent of very strong desktop users we can’t forget about. Even though the number of desktop users decreases all the time, you can’t neglect the desktop experience in favor of the mobile experience because people often start on desktop and move to mobile. The individual channel matters less than the multi-channel experience. It all has to be a good experience.