Digitally driven companies like AHAlife, The Arrivals and Amazon — which recently announced it is buying Whole Foods — are making headlines for creating innovative in-store experiences. By rethinking the retail store from the ground up, these brands are solving problems and gaining sales.

This strategy is in marked contrast to banks, which are increasingly closing offices — and alienating customers.

The idea of brick-and-mortar expansion in banking flies in the face of conventional wisdom. To up their digital game in the face of competition from nonbanks, banks are more likely to invest in mobile and online technology and close more of their underutilized branches. The numbers continue to support this trend. Wells Fargo, for instance, plans to close more than 400 bank branches by the end of 2018. That’s on top of the 84 locations it pulled the plug on in 2016. Bank of America, meanwhile, closed 88 bank branches in 2015 and 171 in 2014. Over the next decade, as many as half of all U.S. bank branches could disappear, according to Keefe, Bruyette & Woods.

In other words, banking — said to be at the greatest risk of major market disruption — is taking an opposite approach than some of the most admired online retailing disruptors. The contrast illustrates the moribund thinking that permeates the financial industry. Banks seem to be caught in the proverbial box, incapable of seeing branches as anything more than a sales channel. Large banks, the supposed digital leaders, are locked into the same mentality of seeing branches from a sales perspective: If sales are falling, cut branches. If return on investment is slipping, eliminate excess real estate.

Even their attempt to innovate their physical footprint is rooted in the same old-fashioned concept. Like others, Bank of America is opening fully automated branches where customers can use ATMs and video conference with employees at other branches. A quarter of the size of typical branches, these techno-offices are designed to sell mortgages, credit cards and auto loans but cannot conduct simple transactions such as cashing checks. In other words, they are smaller, impersonal selling booths — a strategy that distances the bank even further from the customer. No wonder customers consider all banks to be alike, seeing them as merely transactional venues.

Amazon, on the other hand, demonstrates the length a company can go if it is truly dedicated to putting the customer first. Amazon is not just building physical outposts to improve convenience, but is also integrating online and physical operations to pull off the feat. Case in point: Its Amazon Go, a concept physical store, promises consumers both shopping convenience and faster deliveries. Once registered, Amazon Go shoppers don’t even have to check out their purchases. Instead, they simply pick up items from the shelves and walk out of the stores. Their account is automatically charged. Amazon’s $13.7 billion purchase of Whole Foods will only catapult the company into realizing its supermarket dreams.

Similarly, AHAlife, an online retailer for goods like plant-based tattoo oil and star stud, and The Arrivals, which sells clothes, are also testing physical centers to enhance the shopper experience. In all three cases, these online retailers are personalizing and facilitating the buying process through physical stores. Conversely, banks, which could improve their trust quotient through face-to-face interactions, are still focused on dollars and cents.

In addition to physical locations, banks and retailers also take different approaches when it comes to technology. The two industries display sharply contrasting views on investing in software. Bankers use technology to reduce transaction costs; retailers use technology to improve customer service — including in their physical footprints. The fact that online technology satisfies the desires of bank clients is almost incidental to its primary function of reducing costs. Remember the big discussion years ago about the cost-saving opportunities achievable by pushing customers out of lobbies? That incentivized the growth of ATM networks.

Even today, cost savings dominate the discussion around branch closures. We can save significant dollars by transferring customers from in-branch to online transactions. Branch transactions cost $4, whereas an online transaction costs 9 cents and mobile costs the bank 19 cents per transaction, according to PwC.

Amazon’s philosophy, by and large, runs counter to that of the big banks in making its mission to pursue long-term profitability. Amazon’s penchant for reinvesting in the customer experience at the expense of income is legendary. Banks, however, seem to focus on boosting the next quarter’s call reports. The two approaches lead to different outcomes. Amazon is beloved by consumers, while banks rank at the bottom of eight major industries in the Edelman Trust Index. The Seattle-based online giant has a market value of $430 billion, almost twice the value of Bank of America. Amazon’s business model is built around customer service; Bank of America’s is built around quarterly dividends. Amazon takes a long-term view; Bank of America looks for short-term fixes. Amazon CEO Jeff Bezos took home a salary of $81,840 last year, while Bank of America CEO Brian Moynihan enjoyed a record-breaking $20 million pay package.

Serving the customer seems like such an easy concept — except when we try to put concept into practice. If we keep our eyes on an Excel spreadsheet, don’t be surprised if our customers keep theirs peeled for a friendlier fintech or Amazon bank — that will have physical branches.