Summary: Despite digital acceleration, 70+ percent of banking executives view branches as mission-critical or very important to long-term growth. Branch closures between 2020 and 2024 have not produced the customer satisfaction improvements banks expected, in many cases, the opposite. The question is not whether branches are relevant. It is whether yours are operating at the level of service quality that justifies the real estate, and whether you know the answer to that question by branch, not just institution-wide.
Are bank branches still relevant? Yes, and the case is stronger than the headlines suggest. As of 2026, the U.S. still has roughly 76,700 branches, more than 70% of banking executives call them mission-critical.The real question is not whether branches matter. It is whether yours are performing well enough to justify the real estate.
The “bank branch is dead” thesis has been making the rounds since 2012. Every major digital disruption, mobile banking, neobanks, the pandemic, has renewed the eulogy. And yet the branch network in the United States has not collapsed. The branch is not dead. It is also not what it was. The institutions that understand the difference between those two statements are making better decisions than the ones still arguing about whether to close locations.
What the Closure Data Shows
Between June 2023 and June 2024, the number of U.S. bank branches fell from 77,786 to 76,742, a 1.3 percent decline. That is a real number, but a modest one. It does not support the narrative that branches are being abandoned at scale. What it does support is that the industry is doing slow, deliberate triage: closing underperforming locations while retaining networks that drive acquisition, retention, and advice.
Institutions with sophisticated branch-level analytics have consistently found that their branches, properly measured, are producing deposit acquisition, retention, and cross-sell value that does not show up in teller transaction counts.
Why “Digital Is Cheaper” Is an Incomplete Argument
The standard case for accelerating branch closures goes like this: transaction costs at a branch are orders of magnitude higher than digital transaction costs, digital usage is rising, therefore the branch network is becoming an increasingly expensive way to serve customers who would rather serve themselves.
Every part of that argument is technically correct. None of it captures the full picture.
What the argument misses is what branches are for. A customer who checks their balance on a mobile app and a customer who opens a joint checking account at a branch and a customer who asks a personal banker about refinancing their mortgage are not doing the same thing. Measuring the first two against the same metric, cost per transaction, is a category error. Branches are not expensive transaction processors. They are acquisition channels, advice platforms, and trust signals.
The data on acquisition is unambiguous. Industry surveys consistently show that 60 to 80 percent of new consumer checking accounts at community and regional banks are opened in-person, even when the customer discovered the institution online. The branch did not process a high-cost transaction. It converted a prospect into a relationship. That is a different economic function, and it requires a different ROI framework to evaluate.
What the Customer Data Says About Branch Relevance
The ICBA’s 2026 National Survey on consumer banking preferences produced a finding that should end the “branches are irrelevant” argument at community banks and credit unions: a substantial majority of consumers still cite physical branch access as an important factor in choosing a financial institution.
This does not mean customers want to use the branch for every transaction. It means they want the branch to exist. The branch functions as a trust anchor even for customers who never walk in the door. Knowing that a physical location is nearby changes how a customer feels about their institution, its permanence, its accountability, its investment in the community. That effect is real and measurable.
Beyond trust, branches retain functional relevance for specific interactions:
Complex transactions. Mortgage applications, business banking relationships, estate and trust discussions, large CD purchases, and loan modifications are disproportionately initiated in branches. These are the high-value interactions that drive lifetime relationship economics.
Problem resolution. When something goes wrong, a fraud claim, an account dispute, a wire that did not land, customers who have a branch to walk into report higher resolution satisfaction than customers limited to phone and digital channels. The resolution outcome is often identical. The experience of having a human being in front of them changes the satisfaction score.
Life stage transitions. First-time homebuyers, new business owners, recent retirees, and bereaved spouses navigating estate accounts are disproportionately branch visitors. These are the customers at the highest-value inflection points in their financial lives. The institution with a branch staff trained to handle these moments well has a material advantage.
The Branch of 2026 Is Not the Branch of 2016
The 2016 branch was designed around teller throughput. The 2026 branch is being redesigned around advice throughput. That distinction matters because it changes what good branch performance looks like.
Leading institutions have reduced teller stations, expanded private consultation spaces, introduced universal bankers who can open accounts, discuss products, and initiate loan applications without routing customers across multiple staff, and deployed queue management and appointment scheduling technology that eliminates the wait as a satisfaction variable.
The branches producing the clearest ROI signal in this environment share three characteristics: their staff knows the product set well enough to identify cross-sell opportunities naturally during account conversations; their service delivery is consistent enough that customers know what to expect; and their performance is measured at the branch level, not averaged into an institution-wide score that obscures which locations are working and which are not.
What Branch Closure Decisions Cost That Models Miss
When a bank closes a branch, the contribution-margin model typically shows cost savings that accrue immediately. What the model typically does not show is the attrition cost that follows. Customers who lose their primary branch do not uniformly migrate to digital banking. A meaningful percentage migrates to competing institutions, often community banks or credit unions that still have a branch in the market.
A bank that closes ten branches and saves $4 million in annual operating costs but loses $6 million in deposit lifetime value from attriting customers has not made a good decision. The problem is that the $4 million shows up on the income statement in the current year. The $6 million does not.
Institutions that do branch-closure analysis well build an explicit attrition model for each closure candidate: what percentage of current customers are likely to migrate to a nearby branch of the same institution, what percentage are likely to migrate to a competitor, and what the lifetime value of the likely-to-migrate-out segment represents. That calculation changes the break-even math on most closures significantly.
Contact CSP
The question is not whether bank branches are relevant in the abstract. They are, the data says so clearly. The question is whether your branches are relevant, and relevant enough to justify their cost. That is a branch-level question, not an institution-level question. And it requires branch-level measurement to answer.
The banks and credit unions seeing the clearest branch ROI signal in 2026 have three things in common: they measure service quality at the branch level on a consistent cadence; they train to specific service behaviors rather than generic hospitality; and they have closed the loop between service data and operational coaching at the branch manager level.
CSP has helped hundreds of banks and credit unions improve their customer experience. Contact CSP today to learn how your bank can improve its CX system.
Frequently Asked Questions
Do people still use bank branches?
Yes. 52% of consumers visited their primary branch at least once in the past year, and 69% say they prefer a branch within 15 minutes of their home or workplace. Usage patterns have shifted from routine transactions (which have moved to digital) to high-stakes interactions: account openings, loan consultations, problem resolution, and life-stage financial conversations. Branches are used less often and matter more per visit than they did a decade ago.
Are bank branches still profitable in 2026?
Most branches are profitable when measured against a full economic model that includes deposit acquisition attribution, retention value, and cross-sell contribution, not just teller transaction throughput. Branches measured only on direct revenue and operating cost often appear less profitable than they are.
Why are banks still opening branches if digital banking is growing?
Because branches and digital banking serve different functions. Digital handles transactions. Branches handle acquisition, advice, and complex service. Growth in digital usage does not eliminate the need for relationship initiation and high-complexity interactions, it shifts where the branch’s value is concentrated.
How do you measure whether a branch is performing well?
Branch-level service quality measurement, through mystery shopping, customer satisfaction surveys tied to specific branch visits, and complaint tracking by location, is the most reliable way to distinguish branches that are performing from branches that are averaging into an institution-wide number. Transaction volume and deposit balance alone do not answer the question.
What makes a bank branch relevant today?
Staff competency, service consistency, advice capability, and the ability to handle complex financial conversations without routing the customer to three different people. Branches that function as advice centers rather than transaction processing points are the ones demonstrating clear economic value in 2026.