For AI to Pay Off, Financial Institutions Must Avoid ‘Shiny Object Syndrome’

Artificial intelligence is not just a fad, but implementing it wastes time and money if financial institutions don't clearly identify what specific problem AI will help them solve. Define what's wrong first, and then decide if applying this hot technology will really help.

Late in the second day of a three-day off-site strategy meeting for the executive team at a leading financial institution, everyone had grown tired but they were still hammering away at a united plan. The COO, doodling on a notepad, suddenly stopped and focused on the whiteboard.

She had just realized that after two days of discussions, among all the terrific and forward-thinking ideas the team had defined, there was no mention of artificial intelligence.

“Guys, I just realized we have nothing about AI up there,” she said. Many eyes turned to the CIO. He looked sheepish, as if he was the only one expected to bring up technology ideas.

“Let’s add it,” he mumbled.

It was getting late. The facilitator looked at the clock and closed out the session for the day. “Let’s get some drinks, and we’ll pick this up tomorrow.”

On Day Three the executives firmed up their top ten priorities for the coming year. Artificial intelligence made the list. There had been little discussion about it, but everyone agreed AI should be on the list.

“It’s too hot a technology to leave it out,” the CMO declared.

Unfocused Thinking Tends to Stay that Way

Months later, my firm met with the COO to discuss the institution’s strategic plans. She walked us through the list in detail, until we got to number nine — AI.

We asked what specific gap AI was to close and she admitted that the executives hadn’t agreed on any specific application of AI. “Maybe chatbots?” she offered without giving any rationale. It turned out that the AI project was the only one on the whole list that wasn’t defined at all.

“It’s too hot a technology to leave it out,” she said, echoing the CMO’s words from the planning session.

Too often tactical organizations fall for the trap of pursuing trendy technology, the “shiny object syndrome” (SOS). The organization in this true example is highly strategic. However, they fell prey to “SOS” because of another fun acronym: FOMO — the fear of missing out. The buzz for certain kinds of technologies is so strong that even these highly strategic leaders fell for AI’s hype.

As noted in Gartner’s 2021 Hype Cycle for AI, there isn’t a single cycle for AI but an ongoing cycle for specific applications, all falling under the umbrella of AI.

Shiny Objects Aren’t All Bright:

Constant drumming about AI within the financial service industry remains very strong. So “doing AI” becomes its own entry in a list of more reasonable priorities for banks and credit unions.

The problem with SOS isn’t limited to AI. A few years ago, international banks were throwing millions of dollars at “blockchain” without much of a plan. The rationale was that they needed to set up large teams to investigate the possible uses of the distributed ledger technology (DLT) that underpins blockchain.

Organizations must be careful not to fall in love with a solution before defining a problem. Technology on its own cannot be a priority. Banks and credit union executives must guard against falling for the shiny object. Here are some thoughts on how to better weigh and implement hot ideas like AI.

Read More: How RBC’s AI-Powered Digital Assistant Doubled Mobile Engagement

How to Overcome the Allure of Shiny Objects

You can fight SOS, but you won’t win without tight discipline. In the example given, the executive team had been following a disciplined planning process. Yet they let their guard down in the excitement and exhaustion of a three-day planning conference.

Some simple steps can help ensure that SOS doesn’t happen.

1. Start with the problems that are essential for the institution to address. Build strategy by starting with defining the problem or problems, not the means to address them. SOS creeps in when organizations start with solution ideation.

I’ve attended many strategy sessions that begin with brainstorming on projects for the next year. That’s the wrong approach. Generally, executives are doers, so they lean towards implementation.

So begin by listing problems facing customers, employees or the entire organization.

Example Problem: Customers are expecting 24/7 service within our website.

2. Define the impact of solving a problem. After defining a set of problems, the team should look at the outcomes expected to result from solving the problem. This step helps ensure that whatever solution emerges can be tracked and will deliver a favorable outcome.

Example Outcomes: We believe that providing 24/7 service to our customers will increase customer satisfaction as measured by site surveys.

3. Develop a hypothesis.After we understand the problem and the potential outcome of a solution, a team can start defining how the problem can be addressed. This is when a technology can be brought into the picture. In some cases, there might be competing solutions.

Example Hypothesis 1: We believe that providing an AI-driven chatbot on our site can allow customer service 24/7, resulting in increased customer satisfaction as measured by site surveys.

Example Hypotheses 2: We believe that extending live chat to 24/7 can improve customer satisfaction as measured by site surveys.

4. Prioritize Solutions: After defining solutions to the chosen problems, the team should define priorities. When there are competing solutions, determine which one to pursue. There are many tools and approaches to prioritization.

We have seen many organizations skip this step, resulting in an ill-defined roadmap made up of a long list that cannot be accomplished within the time defined.

Key Insight: When defining a strategic plan, organizations must begin with the problems that customers, employees, and the organization itself are facing before jumping into solutions.

Read More: Data and AI Will Drive Banking’s Autonomous Future

What About AI? What Can It Solve?

AI is an enabling technology or set of technologies. As such, organizations should be considering AI when they get to step three outlined above.

As noted by Jordan Bishop in an article in ReadWrite: “It seems nearly everyone has been talking about artificial intelligence for years now, and one of the industries where it’s gained the most attention is personal finance.”

Bishop goes on to list some of the applications of AI in the industry:

  • “Companies can save revenue by making processes more efficient and automated.”
  • “AI becomes a revenue-generating asset that reduces the risk of losses and helps financial institutions make more money.”
  • “Customers can get an edge on financial health with AI’s analysis and remarks on spending …”
  • “Customers can enjoy a more accessible financial experience through chatbots.”
  • “Both companies and individuals can manage risk and automate investing by using AI-based trading.”

This short list is the bare minimum of impact AI is having in financial services.

Buying a Hammer Doesn’t Drive Nails:

As organizations look to address financial institution challenges with AI, they need to be cognizant that “doing AI” isn’t as simple as purchasing AI software. Standing up AI capabilities is difficult and expensive.

Tim Fountaine, Brian McCarthy, and Tamim Saleh note in an article in the Harvard Business Review that:

“One of the biggest mistakes leaders make is to view AI as a plug-and-play technology with immediate returns. Deciding to get a few projects up and running, they begin investing millions in data infrastructure, AI software tools, data expertise, and model development.

“Some of the pilots manage to eke out small gains in pockets of organizations. But then months or years pass without bringing the big wins executives expected. Firms struggle to move from the pilots to companywide programs — and from a focus on discrete business problems, such as improved customer segmentation, to big business challenges, like optimizing the entire customer journey.”

Such an endeavor is particularly difficult for a community bank or credit union, where data and technology resources are stretched thin and executives in those areas must wear many hats. The most viable path is to look for partners that are solving the problems defined in step one above. The partnering approach allows community banks and credit unions to quickly leverage AI without building a discreet AI capability.

Key Insight: Enabling technologies, like AI, are impacting customer expectations and competition throughout the industry. Executives must be aware of the technology and define a path to apply it to address stakeholder problems.

As organizations develop their strategic plans, they should beware of the shiny object syndrome, particularly as it relates to complicated technologies like AI and blockchain. Executives should be familiar with these technologies and the solutions that they offer. However, they should be focusing first on defining problems that need solving and then investigating whether these technologies are available to solve those problems.

This article was originally published on . All content © 2024 by The Financial Brand and may not be reproduced by any means without permission.