Companies of all stripes are getting bot-happy, rolling out bots for third-party platforms like Facebook Messenger, Kik, WeChat, Slack, and more. Firms like Yahoo, H&M, KLM Airlines, and others use these chatbots — software built to simulate human conversation and to help consumers complete tasks — in an effort to better win, serve, and retain customers.

A few banking providers are beginning to dip their bank-shaped toes into the bot space: Capital One allows customers to take actions like paying bills via Alexa on Echo devices; Bank of America has announced plans to roll out a bot on Facebook Messenger; and numerous Chinese providers offer banking services via WeChat.

But while a few banks are in a position to experiment, digital business executives at most banks must decide whether to use precious resources to build or buy a chat bot offering. Forrester’s brand-new research report argues that most of these executives should hold off on launching chatbots for messaging platforms. This is because:

  • Today’s bots often lead to uneven — or worse — experiences for customers. In our research, we found many instances where a chatbot offered a quick and effective answer to a consumer’s question; however, about one-third of the time, existing chatbots either failed to complete the consumer’s request or provided a clunky, awkward experience.
  • Money is an area where the stakes are high — and people are less forgiving. If someone inadvertently orders too many tacos via TacoBot, it’s not a huge deal. But if there’s an issue with a payment or a funds transfer, the result can be profound. Even small tasks can prove problematic: For example, if a bank client uses a chatbot to ask how much money they have available before making a transaction, it would be very disconcerting to hear that he or she has “undefined” money left (which is the answer we got on a non-bank chatbot about the weather).
  • Bots and other AI services will improve — but banks won’t drive this evolution. While bots and their underlying technology are set to get much better over the next few years, banks will not be the companies advancing this technology. That’s not a bad thing: Tech companies are better staffed and equipped to push AI and bots to the next level, and banks can reap the rewards of these advances. Instead, the most important role banks can play is engineering their data and systems to prepare for next-generation bots.

Given these circumstances, most banks outside China will not gain from investing in bots on third-party platforms in the near term. Instead of spending heavily on chat bots now, digital teams at banks should use the next two to three years to focus on foundational digital initiatives by integrating back-end systems, improving data infrastructure, and using APIs to build ecosystems of value. This will allow banks to take advantage of the next-generation bots that technology firms will be working to advance and evolve over the next few years.

Zor Gorelov, CEO of Kasisto — an AI platform company — explained to Forrester why it’s so important that banks do the hard work of updating and integrating their systems and data infrastructure: “To build a truly smart banking bot, we need access to clean, properly structured data from banks. The banks creating lifestyle banking with conversational AI are the ones that are investing in their data infrastructure right now. Our AI and machine learning rely on that data to create intelligent banking conversations with consumers.”

What do you think? We’d love to hear your point of view in the comments section below. Or if you’re a Forrester client, you can reach out to us by requesting an Inquiry.