How Many Bank Branches do we Need in the US?

Create a vendor selection project & run comparison reports
Click to express your interest in this report
Indication of coverage against your requirements
A subscription is required to activate this feature. Contact us for more info.
Celent have reviewed this profile and believe it to be accurate.
25 May 2012
Bob Meara
Finextra published an article yesterday that was also picked up by American Banker and others. The news was twofold: 1. Bank of America announced it enjoys 10 million mobile banking customers, up about 3 million from a year ago - about 43,000 new active mobile customers per week. 2. Concurrent with its swelling ranks of active mobile banking customer, the bank is closing branches and unplugging ATMs. The bank closed 154 branches and eliminated 631 ATMs in Q1, citing the move to online and mobile channels as a contributory factor according to the Finextra article. Celent is not surprized by the branch closure news. As explained in a recent Oliver Wyman report, Branch Flexing: An Agile Approach to Cost Management, April 2012, “to maintain profit levels in the face of a post-crisis and regulatory-reform decline in net revenue of about 32%, US banks would need to increase revenues by 12% a year for the next three years or cut costs by 18% a year, or a combination of the two.” Cost cutting isn’t optional, particularly among larger US banks. Making material cost reductions will require a re-examination of branch networks, which typically contribute between 40% and 60% of a modern retail bank’s costs. Branch flexing refers to a strategic realignment of branch resources (and cost) with customer profitability. Ultimately, branch flexing involves investments in technology, training, culture and compensation. Celent has advocated departure from traditional, teller centric retail operating models for some time. But what about the total number of branches. Is there an argument that the industry has built an unsustainable number of branches, flexing or not? We think so. The argument begins with a simple observation that the US branch density (branches per million inhabitants) has nearly tripled since 1970. Thus, before consumers enjoyed the ATM, telephone banking, internet banking or mobile banking, the industry served consumer’s collective needs with less than 22,000 FDIC insured branches. Do we really need 90, 000 now? We think not. But it’s not so much if they’re needed (The Economist had a great debate about that topic earlier this week), but will they remain profitable? If indeed we’re in a “new normal” of sharply reduced retail banking profitability, than the answer – to one degree or another – is “no”. Celent is developing a more detailed perspective on how many bank branches the US is likely to support over the next ten years. Stay tuned.


  • Next thing to do is to not define customers by channel. BoA has '10 million mobile banking customers' or does BoA have customers, of which 10 million are active users of the mobile channel? The vast majority of the customers would not describe themselve as 'mobile banking customers' as this is one of a number of channels that they expect. If mobile was not available, some customers would leave, if mobile was the only available channel, it is likely that many would leave. The same is true for branch.

    There is no answer for banks in evaluating channels independently, The solution is about the mix of all channels; branch, call centre, web, mobile, 3rd Party etc. The distribution of the channel investment will vary based on the target market and not all target markets need all of the channels. The exciting thing about retail banking is that banks should now think about having different channel strategies to their competitors. The days of replicating every technology move by your competitor is gone.

  • Well said, Grahm.

  • [...] encourage you to read the How Many Branches post as it’s asking productive questions. Here I’ll suggest another direction because I [...]

  • @Bob, I like how you're thinking here. If you abstract up to another layer, what about reinventing the branch? The prevailing assumption is that clients and prospects want to transact there, hence the "convenience" justification. Although I agree with @Graham that the quoted mobile adoption figures don't smell right, it seems obvious that directionally more transactions are shifting to self-service.

    At the same time, most banking services have long commoditized, so banks have become "infrastructure" for many of their clients. I think they can reinvent their relationships with some client types by focusing on supporting client business outcomes online. Most leading banks have experts in business process, industry trends, etc., but existing processes are too costly to deliver experts' knowledge to market. Many-to-many social networks change the game because, by letting clients/prospects run discussions, they set the context, and when the expert interacts with people publicly, s/he activates the network effect.

    But back to branches. I think that banks could repurpose part of existing locations to serve people in new ways at relatively low risk. For example, laptop- and smartphone-toting mobile workers are a very attractive demographic that is underserved around the world. There's an opportunity to repurpose part of select branches' space to create coworking spaces for clients--and use a digital social network to add leverage. Imagine coworking spaces positioned between airport lounges and Starbucks. Here's a detailed picture: in case helpful.

Insight details

Insight Format
Geographic Focus
Asia-Pacific, EMEA, LATAM, North America