Pizza Hut CIO Proving The Unprovable: Mobile ROI

Pizza Hut CIO Baron Concors oversaw what could easily be the most successful mobile application and certainly the most successful retail mobile app, a colorful applet that is directly responsible for "millions of dollars in additional sales" and 1.5 million downloads from Apple. And yet, when he was fighting to get it funded and approved, his return-on-investment (ROI) argument was weak and speculative.

Concors said he was lucky; his senior management team is open to creativity and was willing to roll the pizza dough dice on what sounded like an interesting idea for the world's largest pizza chain, with its more than 6,000 U.S. restaurants and more than 10,000 restaurants in about 100 countries.

But few CIOs are in that position, and that's a piece of reality that could cripple the nascent retail mobile app space. "A lot of companies are struggling with whether to enter this space because of the ROI issue," Concors said.

A big part of the problem is that far too many retailers are deploying mobile apps for the wrong reason or doing it the wrong way. Some chains, Concors said, have aggressively promoted their iPhone app. But when a customer goes to the trouble of downloading and installing that app, they find it does literally nothing other than launching the chain’s mobile site in a browser, which the customer could have easily done without downloading the app.

What retailers need to do, Concors said, is focus on functionality within the desired platform (iPhone, Blackberry, Palm, Android, etc.) and see if there's a clean match for what their company offers and what their customers need.

Ask yourself, "What value am I going to bring to the customer with this app?" and if the answer seems—to the customer—to be “None,” the results will likely be very counterproductive, he said. "If not, they could do more harm than good. They don't realize the damage that they'll cause."

For some chains—such as Pizza Hut—the functionality answer is obvious: the customer wants to buy pizzas more easily. But Concors said that he is more impressed with companies that don't have that obvious an iPhone tie-in and find a legitimate one anyway.

He cites as an example Charmin, the toilet tissue company owned by Procter & Gamble. There isn't a lot of customization or other purchase assistance that an iPhone app could offer, so Charmin instead created an app that leverages geolocation to help users find the nearest public bathroom, along with descriptions. The app is called SitOrSquat and features cartoon bears that, well, clearly have to do something in the woods as soon as practical. (Now there's a gig for an artist that I'll bet she didn't put at the top of her resume.)

Planning for the Pizza Hut app started in 2008, when Concors and others in management noticed "a huge surge among customers moving to iPhones." To take advantage of that movement, "all of the executive team, even our CEO, made a conscious decision to get an iPhone" to replace their Blackberries.The Pizza Hut app has a lot of nice touches. But even before the iPhone app, Concors' team was mobile creative. They offered, for example, the typical SMS purchasing, but with a twist. Their SMS also allows aliases to be assigned for different orders. One order could be labeled "Beth," and it would include a large half-mushroom pizza because that's what Beth likes when she visits. The customer simply texts that one word to the system and it knows everything else—the phone number the order is sent from tells the app the delivery address and the payment card on file while the one-word alias describes the entire order.

The app itself leverages as many iPhone attributes as possible. The geolocation for the nearest restaurant is easy, but Pizza Hut’s app allows users to order pepperoni on only half the pizza by tilting the phone. Animation then shows the topping sliding in that direction, until it only covers half the pie. If chicken wings are ordered, shaking the phone applies the sauce. The phone's two-finger pinch and stretch functionality allows the pizza's size to change without clicking or typing.

The mobile app—just like the Web app—is also tied into workforce records at specific restaurants. This offering has several benefits. First, if an employee phones in sick or has a day off—or is merely late—the system notes it. When an order is sent to a store that has one fewer worker, the app increases the amount of time to create and deliver the order, making its time estimates much more accurate.

When a store is closed, the system will not allow any orders to be processed for it, something that is reversed as soon as the store manager signs in. This feature was tested this week during a blizzard in the New York City region and, sure enough, Zip Codes in the blizzard area all refused orders and Zip Codes in less weather-intensive areas accepted them.

The mobile app also has some extensive analytics, which allow Concors to know which sales are likely mobile-app-caused. "We can see where and when people used to order" and the system flags any purchase pattern changes coming from mobile app users, he said.

Some retailers—such as Starbucks—have been scared about allowing mobile apps to complete payment, worrying about the security issues, among other things. Pizza Hut sidestepped that issue by using its Web cloud. Consumers have their payment method saved within their profile, which sits in Pizza Hut's own servers. (The customer can also choose to pay when they pick up their order or have it delivered.) The mobile app can then simply send the purchase request through the Web site to hit the cloud and then transmit the order to the store. Those mobile orders come into the Pizza Hut location just like any Web or SMS order.

The biggest challenge the Pizza Hut team faced during its initial testing and predeployment involved load-balancing and testing, given how remarkably unknown the iPhone app market's likely response was at the time.

"How do you test the load? How do you possibly know how popular it's going to be?" the pizza chain CIO asked. "We didn't know what to expect from a load perspective. That forced us to test a variety of scenarios."