A Guide to Implementing AI Technologies in 2022

The past few years have been filled with complexity and change in the banking industry, which has also triggered a shift in how credit unions use artificial intelligence. We have witnessed an incremental increase in the use of AI in all aspects of banking, as it can reduce labor costs, increase efficiency and productivity, and help credit unions provide better service for their members. Traditionally, credit unions only leveraged AI to automate routine internal processes, like compliance, underwriting or fraud detection, but recent technological developments have led to AI now also being used for front-office purposes, like member service.

This trend will continue this year, as members of credit unions will default primarily to digital channels when searching for solutions to match their financial needs. When it comes to AI, we have seen consumers get particularly excited about features like uniquely tailored services or offerings that anticipate their needs, such as chatboxes or 24/7 customer service bots that can proactively start conversations and provide relevant information and recommendations at any time. Customer relationship management in banking was previously mainly conducted by humans, but AI is now leading the way.

Simply put, AI is trendy. But credit unions need to be careful not to fall into the common trap of pursuing trendy technologies only because everyone else is using them. Known as the shiny object syndrome (SOS), the want rather than need to implement AI can be more detrimental than beneficial for a multitude of reasons.

For starters, investing millions in data infrastructure, AI software tools, data expertise and model development due to a fear of missing out and without actually having a need or long-term strategic plan is an expensive and futile proposition. Even with a plan, acquiring AI without understanding its complexity or conducting a comprehensive proof of concept is wasteful, as the technology will be hard to implement and manage in the long term. One of the biggest mistakes credit union executives make is view AI as a technology with immediate returns, while in reality, months or years can pass before the technology starts bringing in the big wins that executives expected. Being strategic and cautious about acquiring technology should be the norm for all credit unions but is particularly crucial when it comes to AI.

Moreover, AI technology is still far from perfect, especially when it comes to customer service support. This deficit might not be a deterrent for big or medium-sized banks, as their focus has never been on delivering a personalized banking experience for their customers. But many credit union members have specifically chosen to bank with credit unions due to their ethos of caring for their communities and uniquely tailored member service offerings. If credit unions jump on the AI trend and start replacing humans with a substandard customer service bot, without conducting an in-depth market research and analysis beforehand, they gamble losing the essence of what differentiates them in the market.

So, what can credit unions do to avoid falling in the SOS trap?

First, organizations should define the member problems that the credit union can address. That process begins with talking to members to understand their needs and build empathy. The next step is to look at the member journey and define pain points that require a different approach. They might find that, for example, their members expect to receive 24/7 customer service support or automated bots answer the majority of their calls instead of humans. Credit unions should also define the expected outcomes after solving these problems, as well as specific key performance indicators (KPIs) and defined goals. For instance, if the expected outcome is providing effective member service 24/7, the KPIs could be “Number of Sessions Not Resulting in a Call Back” and “Reduction of Call Center Calls.” The defined goals would then be to double the number of self-service sessions within the first year and reduce calls by 15% in that year.

Once both problems and the desired outcomes are recognized, credit unions should start looking into the solutions. This is when AI technology can be brought into the picture, and if credit unions find competing software providers, they should determine which one fits into the organization’s culture and structure. A good technology partner should be able to support a credit union throughout every stage of the process, from education and testing all the way to implementation and the day-to-day operation of the system. A modern partner should also help the credit union save on costs by eliminating the need to invest in data warehouses and an in-house IT department, and instead take on those responsibilities itself. In the long-term, a successful partnership with a third-party technology provider should help credit unions solve the previously identified problems and reach their desired outcomes.

In 2022, while credit unions will increasingly acquire and implement digital solutions to meet the expectations of today’s tech savvy consumers, they should be aware of SOS, particularly as it relates to complex and expensive AI technologies. Rather than jumping on the AI trend because everyone else is doing it, credit union executives should first identify both the gaps they are trying to fill and the desired outcomes, then research and analyze the AI solutions available on the market and ultimately choose the right one for their organization, one that fits their mission and values, and helps their members and communities thrive. As the saying goes – don’t be a sheep, be a shepherd.

Alex Jimenez Chief Strategy Officer Finalytics.ai San Mateo, Calif.