In a year of high-profile stumbles, Cartwheel has been one of Target Corp.'s biggest wins.

The digital coupon app has generated droves of traffic at a time when smartphones are becoming an increasingly important shopping tool. Cartwheel, which allows customers to select deals that can be redeemed in the check-out line, has a devoted and growing user base of 8.5 million people who have used it to rack up more than $95 million in savings.

And none of Target's competitors -- not Wal-Mart or Amazon -- have anything quite like it.

"There's a notion among Wall Street that Target is behind from a digital standpoint, but you can argue this is one of the most successful retail apps out there," said Matt Nemer, an analyst with Wells Fargo.

But according to Target officials, Cartwheel almost never happened. The Minneapolis-based retailer's slow-moving and risk-averse culture nearly sank it before it launched a year ago. Target managed to save it by departing from the old Target way.

Now Cartwheel is being held up within the company as a role model for how the company can move faster and test new ideas in the marketplace in real time instead of getting held up in a boardroom.

"Anytime you're facing change, you have to give people a vision of what you want them to do," said Jeff Jones, Target's chief marketing officer. "And you have to give them evidence that it actually works."

Cartwheel is a big piece of that evidence. But Target needs more Cartwheels, he said.

Now analysts are curious how Target's first CEO to come from outside may further shake up Target. Brian Cornell, a former PepsiCo and Sam's Club executive, will take the reins on Tuesday.

But in the weeks leading up to Cornell's appointment, Target's leaders have already been more open about the need to remove layers of bureaucracy and be more nimble. Target's journey with Cartwheel highlights how the company sometimes got in its own way -- and also the potential rewards for working in a different way.

"It was instrumental in seeing how something like that can get bogged down," said Carol Spieckerman, president of retail consulting firm newmarketbuilders. "The best lesson they can learn from Cartwheel is just looking at its success -- 'Wow, we did this thing.' But next time, it can't take that long and we can't let everybody have a vote."

Cartwheel allows customers to pick from more than 600 promotions that encompass about 30 percent of the store.

It's also been a work in progress. Since it was launched last May, Cartwheel has undergone dramatic transformations, with thousands of updates both small and big. One of the most significant was making it into an app. Cartwheel was originally envisioned to be primarily a website-based tool, but the team behind it quickly pivoted to make an app, which is how it is mostly used today.

The initial idea behind Cartwheel was hatched in the summer of 2011 during a brainstorming session at Lyon's Pub, a few blocks away from Target's downtown Minneapolis headquarters. The social business strategy team was trying to think of ways to tap into the power of Facebook. They came up with the idea of giving Target's Facebook fans some sort of a discount that they could personalize based on the things they like to buy.

The concept was initially a hard sell at Target, said Sarah Peterson, who was part of the original Cartwheel team and is now its leader. The retailer had just come out of an arduous process to take over its website operations after farming it out to Amazon for a decade. But in December 2011, company leaders were intrigued enough to fund Cartwheel as a project.

At Target, that meant it became a small part of a whole lot of people's duties.

"So everyone had this project as one-eighth of their job," said Peterson. "It got sort of chaotic."

The team swelled to more than 200 people. At any one time, she said, the project had as many as a half dozen different leaders, who each had their own set of priorities and thoughts about what Cartwheel could be.

By November 2012, frustrations had mounted. Facebook was very interested in the idea, but there were concerns the project would fail.

"We recognized this was a big idea that had never been done," said Jones. "But we knew we were moving too slowly.

Part of the problem, he added, was this feeling that the product had to be more perfect than it needed to be from the outset.

"The old way at Target is imagine every question that could be asked and try to figure out the answer," he said. "And imagine all of the obstacles you could face in the market and decide what you're going to do."

So they brought in Alan Wizemann, a startup guy who had worked with Target in the past through his firm ShopIgniter, to be a catalyst for change. They wanted him to help them see what it would take to run Cartwheel more like a startup.

Wizemann realized that change needed to happen pretty quickly. Decisions that should just take a couple of minutes would take weeks and would require a "significant amount of documentation" and the OK from multiple different people, he said.

"Where does the search bar go? It doesn't take 40 people to figure that out," he said.

Rather, the idea needed to be tested in real time as it was being built, he said. After all, it had to move as fast -- if not faster -- than Facebook.

So with the input of Peterson and a few others, he came up with a new direction: Downsize the team to about 50 people dedicated just to Cartwheel. Have a single person in charge. And let them put out Cartwheel in a beta form, so they could tweak it as they went.

The leaders at Target bought into that vision.

"From a company standpoint, it was a pivotal moment of change," Wizemann said.

In January 2013, Wizemann came to Target as a full-time consultant. Within weeks, Cartwheel's much smaller team from marketing, merchandising and analytics moved into a cramped space they now somewhat endearingly refer to as "the closet."

When it came time to launch Cartwheel, they didn't have to get an official OK from higher up the food chain. They just turned it on.

But there were some tradeoffs in working this way. Cartwheel was initially supposed to have a $10 million marketing budget. But as more of a beta launch, that disappeared.

So it cobbled together $5,000 from other departments. They told their friends to spread the word. Some employees made handmade signs to put up in stores. And, of course, it was promoted on Facebook.

Original plans also called for a call center to handle customers' questions. They were scrapped. Instead, they put a direct forum into Cartwheel, where users could write for all to see what they didn't like or what didn't work.

Engineers monitored it and made fixes in real time as complaints surfaced.

"For Target, that was very new," Wizemann said. "The idea of that openness was really not a part of the culture here."

Companies like Amazon are constantly running beta tests, said Nemer of Wells Fargo. In today's digital world, things are moving too fast to use that old cautious approach. And consumers can be understanding of hiccups, especially if it's new, he said.

"It may not be the end of the world if it's not perfect," he said. "My Uber app crashes every 10th time, but I don't care because I love Uber."

By most measures, Cartwheel has been a huge success.

Target soon expects it will touch about $1 billion in sales. It's doing the kind of things brands desperately want: deepening engagement and loyalty. It's been especially popular among millennials, a highly-coveted demographic. Target says active Cartwheel users on average increase their spending and trips to its store by 30 percent.

"Consumers seem to be responding to it in a pretty profound way," said Andrew Lipsman, vice president of marketing and insights for comScore, which measures online traffic and sales. "It's been very, very successful."

In the last year, Target saw a 251 percent in increase in time spent with Target on mobile devices, with Cartwheel contributing 80 percent to that increase, according to comScore data.

Until Cartwheel, there had not been a strong example of a brick-and-mortar store with a super popular app, Lipsman said. For most retailers, the vast majority of their mobile traffic still comes through a website. But with Cartwheel, Target has managed to flip that with nearly 75 percent of its mobile traffic now coming to its app, joining the ranks of Amazon and eBay.

Target, however, has still been playing catchup when it comes to its overall online operations, especially its desktop website. But with online traffic quickly moving from desktops to smartphones, analysts say that retailers that are leaders in the mobile space will have a leg up.

In the meantime, the Cartwheel team is not resting on its laurels.

It's now moved out of the closet into a bigger space. But it still has the feel of a startup with whiteboard walls plastered with Post-it notes and a foosball table in the middle of the room.

Cartwheel took off the beta tag last month and is now being promoted in Target's weekly ads. It has added new features such as a leader board where users can see how much their Facebook friends have saved. And it has begun testing online redemption of deals.

As Target looks to take on more of a startup mentality, it has brought Wizemann on board as its vice president of mobile and Target.com. He started in that role in early June.

He said Target is now actively exploring another 30 to 50 new product ideas -- one of which could be the next Cartwheel.

Kavita Kumar • 612-673-4113