BofA’s ‘Tone Deaf Robot’ Replies To Consumers With Boilerplate Tweets

Skip To Page: 1 2

On a Saturday in early July, an activist in the Occupy movement using the Twitter handle @darthmarkh posted a tweet with a picture of him getting chased by cops away from a BofA facility in New York. He had just been chalking up the bank’s sidewalk with protest slogans. After a few retweets by other Occupiers, BofA’s Twitter robot decided to step in. “I work for Bank of America. What happened? Anything I can do to help?” said one reply. Another read, “We are here to help, listen, and learn from our customers and are glad to assist with any account related inquiries.” Unfortunately for BofA, its Twitter algorithm failed to realize Occupy protesters weren’t looking for customer support.



Skip To Page: 1 2

Jeffry PilcherDon't miss THE FINANCIAL BRAND FORUM — April 15-17, 2019 in Las Vegas. Join 2,000+ of the best and brightest in banking at the world’s most elite conference on marketing, CX, data analytics and digital transformation for the financial industry. Banks and credit unions that register now SAVE $870+ and get a free upgrade to a GOLD PASS, and PAY NOTHING until next year! REGISTER NOW!

This article was originally published on July 17, 2013. All content © 2018 by The Financial Brand and may not be reproduced by any means without permission.


  1. It’s the “^sa” that I have issue with. If you want to use an automated process to respond to Twitter complaints, fine, but don’t sign it with a person’s initials. That’s disingenuous. Adding initials on Twitter was designed to indicate that you were dealing with a real person, and that person had actually seen your Tweet and typed out a response. By automating Tweets with initials attached, BofA has broken that trust, and made users more likely to second-guess Tweets they receive from the BofA account.

  2. Cory, very few (normal) people are going to take the time to analyze BofA’s outbound tweets. Probably 90%+ of consumers will find one of BofA’s Twitter accounts and fire off their @message, spending no more than 20-30 seconds total in the process. When/if they get an @reply from ^sa, they aren’t going to know if another 1,000 people got the same message or not. On the macro level, it might look awkward, but only social media analysts are going to take the time to look at BofA’s activity from that perspective. Most (normal) users are probably going to be surprised — and somewhat pleasantly so — that they got any kind of response at all. And my hunch is that if someone sent a real reply to “^sa,” that a real human would step in. BofA probably uses ^sa (the auto-responder) for any Twitter mentions its filter deems as marginal or questionable. Maybe ^sa is just used as a probe or prompt by the system when it’s uncertain, to see if the person making the comment expected any kind of assistance.

  3. Leigh Enselman says:

    Social media has proven itself as an effective customer communication tool for banks. Simple inquiries can be taken care of quickly and efficiently, saving time for both the bank and the consumer. Is it a perfect mode of communication? No. With an estimated 57 million consumers, it is completely legitimate that B of A would employ technology to help sort and respond to their 20,000 plus tweets per month. Could Bank of America have done a better job at filtering their Tweets in order to decipher the ones that needed a response? Absolutely. However, I feel that your point about the “faction of haters that will never do business with them under any circumstances” is a strong one—even if this mistake wasn’t made, the people pointing it out are never going to become customers either way. B of A still holds a retail banking footprint that covers approximately 80 percent of the U.S. population, not to mention glowing reviews on My Bank Tracker. They’re obviously doing something right.

  4. Cory is correct in my view. Adding a person’s initials to a tweet not generated by that person is at best a misrepresentation and not intellectually honest. I have no issue with automated responses if that’s what BofA needs to do for their business model, but to try and imply “personal” service that is automated is not genuine and is representative of some things big banks can do to erode customer intimacy and trust.

    Isn’t this supposed to be “social” media?

  5. If “intimacy” and “personal service” are priorities to someone in their banking provider, then you have to seriously question their decision to pick BofA in the first place. If a customer wants small town service, they should pick a small town bank. But anyone choosing one of the big four banks has — implicitly and by default — opted for a mass-market experience that does not entail the same level of service expectations as smaller, more intimate financial institutions. As an analogy, consumers understand that you can’t apply the same expectations to a national fast food joint that you would to a white-linen restaurant.

    Speed/convenience, price, quality, service — consumers can’t have it all.

Other Sites Linking to This Article:

  1. Brett King, Jim Marous, ATMs and BofA Tweet-Bot Fails | Backbase Blog
  2. ++ The Human Touch ++ « ++ musings & marketing ++
  3. Three Brand Fails That Prove Auto-Replies On Twitter Are A Bad Idea - AllTwitter

Speak Your Mind


Show Comments