This past quarter, Starbucks’ domestic 6 percent same-store sales growth gave it a 10 percent two-year stack—the company’s best performance since 2017. And there have been some buzzing headlines along the way. Everything from artificial intelligence to 35,000-square-foot Reserve stores to a rewards revamp—all helped bounce Starbucks off a string of stagnant results. In Q3 2018, the java chain reported its lowest global comp result in nearly a decade at 1 percent. For the full year, traffic in Starbucks’ Americas segment was down a percent compared to 2017, which was a flat year. The year before that it was positive 1 percent. So as much as Starbucks was able to drive revenue off net unit growth and digital expansion, it really wasn’t moving the transaction needle for some time.

But that has flipped in a hurry lately. Here’s how Starbucks’ traffic has trended throughout 2019 in the Americas:

• Q4 2019: 3 percent
• Q3 2019: 3 percent
• Q2 2019: Flat
• Q1 2019: Flat

And the comps picture has aligned:

U.S. and Americas same-store sales:

• Q4 2019: 6 percent
• Q3 2019: 7 percent
• Q2 2019: 4 percent
• Q1 2019: 4 percent
• Q4 2018: 4 percent
• Q3 2018: 1 percent
• Q2 2018: 2 percent
• Q1 2018: 2 percent

Global same-store sales:

• Q4 2019: 5 percent
• Q3 2019: 6 percent
• Q2 2019: 3 percent
• Q1 2019: 4 percent
• Q4 2018: 3 percent
• Q3 2018: 1 percent
• Q2 2018: 2 percent
• Q1 2018: 2 percent

Yet despite the attention some of Starbucks’ more forward-thinking innovations garnered, the real igniter isn’t so much a fresh layer as it’s a focus back to the core. And it might just be thanks to the performance—or lack thereof—of one of Starbucks’ classic staples, the Frappuccino.

Of the chain’s 6 percent hike in Q4, beverage drove five points of it (food was the rest). It marked five straight period drinks set the pace.

To really understand how critical this is, you need to stretch back about a year and a half. Starbucks was splitting innovation between food and beverage, CFO Pat Grismer said this week at the Morgan Stanley Global Consumer and Retail Conference.

“And we’ve taken our product innovation resources and focused them more around beverage …” he said. “We recognize that we are a beverage-forward concept. Beverage is our key point of differentiation.”

Cold beverage development in particular—a sector Dunkin’ has long fronted—is where research directed Starbucks. It was essential to unlocking daypart expansion, especially in the afternoon, and reaching younger consumers flocking to local cafes.

In recent years, Starbucks struggled to generate afternoon business thanks, largely, to the downturn of its Frappuccino platform. As consumers looked to pivot away from indulgent options, Starbucks saw sales of its iconic platform decline, “and that weighed particularly on our afternoon daypart,” Grismer said.

Today, two things have happened: Frappuccino sales have plateaued. They remain a meaningful part of the chain’s business, Grismer said. But while they’re not growing, they’re not declining at the rate of previous years, either. Meanwhile, Starbucks poured those innovation resources into cold beverages fitting of today’s espresso-crazed demographic.

Considering Dunkin’ for a moment, which completely reinvented its espresso offering in 2018, the company said 2017 was the first time consumers under 35 drank more espresso beverages than hot drip coffee. And more than 50 percent of millennials ordered espresso beverages when they chose coffee. CEO David Hoffmann likened espresso to the transitional role drip coffee once played for soda-drinking consumers as they matured. Espresso is now the gateway coffee choice for millennials and Gen Zers.

In Starbucks’ case, these new cold beverages have taken up some of the demand left behind by Frappuccino’s sales decline.  

This isn’t just Nitro Cold Brew or Cold Foam Cold Brew. It also includes Starbucks’ Refreshers line and flavored iced teas. Cold drinks currently mix more than half of U.S. Starbucks beverage sales.

Overall, the brand’s beverage growth expanded across all dayparts in Q4. Again, it was anchored by cold drinks, which proved effective sales and traffic drivers regardless of season or occasion, Grismer said.

Starbucks rolled nitro equipment across company-owned stores over the last 12–18 months. It hit 80 percent penetration in the U.S. in August and was fully onboarded by year’s end. This allowed Starbucks to support nitro with national advertising for the first time, which it did in August. It drew more occasional customers into the brand and slightly favored the afternoon daypart, Grismer said.

“By applying more focus and more resource to beverage innovation and driving more consumer research to drive more insights around what is going to resonate most with our customer base, particularly young customers, we’ve identified those insights that have allowed us to introduce new products that have performed better than in earlier years,” he said.

It all fits into a broader, less visible shift as well. Two years ago, Starbucks was far more reliant on limited-time offer beverages (Unicorn Frappuccino comes to mind). What the company referred to internally as “Sparks.” These generated spike traffic but also disrupted operations. Today, Starbucks’s approaches beverage innovation from a platform angle over a product one. It’s not all that different from the path Wendy’s took last year to fix its menu mix concerns. Instead of discounted deals and fleeting offers, which require fresh marketing and spend each launch, the burger chain developed Made to Crave and Biggie Bag constructs it could build around and into. This way, it’s pulsing items into a recognizable design instead of trying to start from scratch each time.

Starbucks has done this, too, over the past two years. For example, its cold foam launch served as the springboard for Cloud Macchiato.

Starbucks first brought cold foam to Roastery locations in 2014 and it’s been an option systemwide since April 2018. The first iteration was Cold Foam Cascara Cold Brew, followed that July by Salted Cream Cold Brew. Then came the Pumpkin Cream option this fall, which was Starbucks first new pumpkin coffee beverage since the Pumpkin Spice Latte in 2003. And now cold foam is the key differentiator for Starbucks’ recent Irish Cream Cold Brew launched December 3.

The overarching idea, Grismer said: Introduce platforms that allow Starbucks to unveil innovation in ways that don’t drive a lot of disruption into its stores.

He added that Starbucks’ cold ascension also hasn’t come at the expense of hot beverages. “We continue to see growth in our hot beverage platform. And so, we’re seeing growth across the entire range of beverages. And we remain very pleased with the insights that we’re gaining, that we’re able to capitalize in the way that we introduce new product news,” he said.

Growth, rewards, and other levers

Starbucks’ global growth in recent years outpaced most brands of this size. The company’s total three-year net store compounded annual growth rate is about 8 percent. For perspective, Yum! Brands is roughly 4 percent and McDonald’s 2 percent.

In the U.S., that’s 5 percent, flat, and negative 1 percent, respectively, to about 15,000, 18,000, and 14,000 restaurants.

From 2017–2018, no chain in America added more restaurants net, and it wasn’t really close. Starbucks grew by 895 domestic units. Dunkin’ was next at 278 restaurants. Only Domino’s (258) and Wingstop (225) expanded by more than 200 units.

In fiscal 2019, Starbucks widened to 18,067 restaurants in the Americas, up from 17,460 in the year-ago period (3 percent change). Internationally, the brand upped to 13,189 from 11,852—a robust 11 percent jump, or 1,337 net new store openings over the past 12 months.

In 2020, Starbucks expects to open about 2,000 new units globally and roughly 600 in the Americas, with 3–4 percent U.S. growth.

The chain has, however, dealt with cannibalization in recent years as it ballooned quickly. Perhaps too quickly. The chain closed 150 or so stores in densely penetrated U.S. areas in 2019, leading to slower urban growth and a focus on smaller locations, even a pick-up only model.

Consequently, this drove expansion in the central and southern regions of the country, primarily with high-volume drive thrus. Grismer said Starbucks is the only brand at this scale to have lifted store count in the U.S. the past three years and remains “far from full penetration in our home market.”

Also, focusing on China, Starbucks is currently opening a restaurant about every 15 hours, with an ultimate goal of adding 600 new cafes this year to its current count of 4,125. The brand appreciated 5 percent growth in the country last quarter. Grismer did caution, though, Starbucks’ China comps might rise as little as 1 percent in 2020 due to the same issues that lagged U.S. traffic before it picked up recently—rapid growth.

“We’re effectively doing it to ourselves, we’re doing it intentionally in the interest of growing total transactions and total sales,” he said.

Turning the conversation to rewards, Grismer said Starbucks’ loyalty program’s 90-day active member base grew year-over-year 15 percent to more than 17 million members at the end of Q4. In China, it bumped 45 percent to 10 million people after introducing a spend-based program update last December.

Much of Starbucks’ current pace can be credited to its relaunched platform, which landed in April. The structural change focused on flexibility, and how customers could earn and redeem “stars.”

Notably, Starbucks previously faced a friction point with its “Green” tier, where customers needed to accumulate 300 stars to reach gold status and then had to earn another 125 in order to start redeeming. That was a hefty commitment and not a workable incentive for less frequent, or occasional, customers.

The new, multi-tier system removed the tier and allowed guests to start redeeming with as few as 25 stars. Starbucks also expanded the range of products available for redemption. Now, there’s a 25- to 400-star window that offers merchandise and at-home coffee purchases.

“And we’ve seen significant positive customer response to this change, which was exactly what we had designed for, because in fact there was significant customer research that went into the program design and significant efforts to educate our partners and customers about the changes, so that there wasn’t a surprise and so the customers and partners alike could understand the value that is unlocked by providing this level of flexibility,” Grismer said.

This past quarter, Starbucks’ loyalty guests contributed nearly 2 points of comp and represented about 42 percent of the company’s tender.

A complementary benefit has been better customer engagement. Starbucks attracted more light users into the program and broadened its sizable reach. That created “significant opportunity,” he added, “because we know from our experience that as we migrate customers from what we call digitally registered into full rewards members, we see their spend increase substantially.”

This past quarter, Starbucks’ loyalty guests contributed nearly 2 points of comp and represented about 42 percent of the company’s tender. COO Roz Brewer said earlier the 150-star level witnessed the majority of volume. Additionally, the occasional customer Starbucks referenced is one of the culprits for its afternoon growth.

But perhaps the most vivid element at play is how it addressed a serious concern for Starbucks. When the brand’s U.S. traffic lagged, one of the detractors was its inability to incentivize repeat visits. With the previous structure, could Starbucks keep customers engaged long enough to send them benefits and win them over long-term? Those less-frequent guests often bailed after a few visits or just forgot to activate their rewards given the commitment required. It also did little to inspire digitally registered customers (those who have given an email but not signed up) to take the next step to loyalty. It took 30–40 trips to get stars. Today’s “redemption for all” structure offers new members stars within two to three visits.

As Grismer said, it’s a dramatic revamp with a lot at stake.

“I would highlight this change as something that has more of a flip-the-switch impact to our business, as opposed to the steady build that I was mentioning earlier [with beverages], because from the moment we turned on that program we saw these positive results start to accumulate,” he said.

When Starbucks launched the multi-tier strategy, it also took the opportunity to introduce an enhanced personalized marketing engine into its technology stack.

“And the importance of this is that it allows us through machine learning, to gain increasing insights around what matters most to our customers which informs the offers that we make to them digitally,” Grismer said. “And on the back of these enhancements, we have seen the contribution from digital customer engagement increase from what had been closer to 1 comp to 2 percentage points of comp in the last two quarters.”

The delivery discussion

Grismer also spoke to Starbucks’ recent Uber Eats delivery program, which remains in early stages (it mixes less than 1 percent at U.S. stores still).

He said it’s been a slow build domestically, compared to China where it’s 7 percent of sales, “because compared to food and beverage delivery generally we are at a lower ticket. And so the delivery charge is a higher percentage of ticket and probably acts as a bit of a barrier to utilization of delivery for beverages only.” Grismer added Starbucks expects to achieve national coverage in fiscal 2020.

So far, the brand has seen incrementality, an important note considering delivery transactions are margin dilutive due to the commission and packaging costs.

Starbucks’ delivery transactions have returned significantly higher average checks, Grismer said. The main reason being there are higher rates of food attachment than in-store. Also, larger party sizes, particularly in office environments.