Domino’s is planning a major development push in the U.S. in the next several years as the pizza chain hopes to protect its market share from competitors and third-party delivery services.

The Ann Arbor, Mich.-based pizza giant told analysts on Thursday that it could build another 2,000 stores in the U.S. over the next seven to eight years as part of a “fortressing” strategy to build more carryout sales and improve the quality of delivery.

It’s part of an overall growth strategy that would add another 10,000 locations worldwide by 2025, when Domino’s hopes to generate $25 billion in worldwide retail sales. The goal, CEO Ritch Allison told investors on Thursday, is to be “the dominant No. 1” pizza chain in the world.

The 2,000-unit domestic growth would increase Domino’s U.S. footprint by more than a third in less than a decade. That’s massive growth for a company that until just a few years ago hadn’t added new units in two decades.

Yet Domino’s executives argue that adding more units would enable the company to grow overall retail sales and help franchisees improve their overall profits. And they say operators are buying into the strategy, which is backed by the company’s considerable data capabilities.

“It’s a no-brainer investment on our part,” Allison told analysts and investors.

But building so many units is a risk. The U.S. market in particular is loaded with restaurants and the pizza market is largely saturated. Building more stores can cannibalize existing locations and hurt overall unit economics and create dissension among operators worried about lower profits.

Domino’s executives said on Thursday that its fortressing strategy hurt U.S. same-store sales in 2018 by 1% to 1.5%.

But they argue that the company needs more locations after quarterly same-store sales growth averaged 7.4% since 2010.

Same-store sales in the first three quarters of 2018, for instance, rose 7.1% despite the headwind from new locations—growth that easily outdistanced other publicly traded restaurant chains. “Nobody is even doing half of what we’re doing,” said Russell Weiner, Domino’s chief operating officer and president of the Americas division.

By building new locations, Domino’s says it can increase carryout business because those customers will only travel so far for that pizza.

But they also say delivery is more efficient, which improves the economics and opportunities for the company’s delivery drivers, who can make more deliveries and get more tips and income.

That improves the quality of delivery in an era in which more consumers are ordering their food delivered. It can improve delivery times, Weiner said, which improves the freshness of the product and makes Domino’s more competitive with frozen pizza made at home.

“Thirty minutes, which is what the business was founded on, is not good enough anymore,” Allison said, referring to Domino’s old guarantee of delivery in 30 minutes or less. “We’ve got to be better and better and better.”

Allison noted that if a pizza can be delivered in 17 minutes then the company is competitive even with the length of time it takes to prepare a frozen pizza. “Fortressing is going to continue to help that,” he said.

Domino’s says the strategy will help with operators’ overall profits. The company over the years has relied more on multiunit operators than single-store franchisees—the chain had 278 single-store franchisees in 2017, half the number it had a decade earlier.

Over that time, Weiner said, franchisees’ total earnings before interest, taxes, depreciation and amortization, or EBITDA, has tripled to $707 million from $223 million. By splitting stores’ trade area, Weiner said, franchisees generate more carryout revenue, improve delivery efficiency and ultimately increase their total earnings.

The company said it has proven this strategy out in a number of markets. In India, for instance, Domino’s fortressing strategy ultimately pushed out rival Papa John’s. Weiner noted that, in Las Vegas, the company added a fourth store to a three-store area and as a result increased EBITDA per store by $16,654 per year thanks to significant increases in carryout business.

A franchisee in Roanoke, Va., split a one-store trade area and within a year generated another $500,000 in sales and $130,000 in EBITDA.

Allison noted that the company has been doing this already and only had seven closures in the U.S. in the first three quarters of 2018—a tiny number for a 5,700-unit chain.

“We’ve been doing this for several years now,” he said. “If you’re opening a lot of bad units you’re going to see a higher closure rate than that.”