As an online-only bank, adopting strong and efficient digital marketing techniques is crucial for Ally Financial Inc., and it is using artificial intelligence to help at that on a greater scale.
The bank started testing generative AI in marketing mid-2023, when its Ally.ai platform launched as a proprietary, cloud-based tool designed to execute its AI-related capabilities. The platform serves as a bridge between internal AI applications and larger, external language models such as OpenAI or Google's chatbot Bard.
Ally's (NYSE: ALLY) marketing test with GenAI is the second experiment where Ally.ai was leveraged. The experiment involved about 14 marketing employees using the platform for quicker campaign development and productivity. The results proved to be successful, as Ally.ai reduced the time required to build creative campaigns and content by up to two to three weeks and time savings on tasks averaged 34%.
"Also closely tied to it is the partnership that the technology and marketing team both have in terms of dreaming big, seeing around the corners and anticipating what our customers want," said Sathish Muthukrishnan, Ally's chief information, data and digital officer who is based in Charlotte. "And it became a natural partnership for us to explore what we can do with marketing."
The marketing group implemented 80 AI prompts, providing the platform with information to produce content at an accuracy rate of 81%. Prompting is a reflection of one of Ally's core principals with AI, incorporating human intervention.
"What we found out was the initial answers that generative AI created is not what we really wanted," he said. "So, that's why we wanted a human in the middle so they can tweak it to have the brand spirit come through."
During testing for Ally.ai in marketing, employees used the platform for creative development — specifically, for first drafts of advertising copy, video scripts, social-media posts and more. The platform's marketing-use case is in the governance process for final reviews, with the intent to be pushed into production soon.
Ally's marketing test comes after it launched a pilot of its first use case with its customer care team on June 27. That experiment involved summarizing customer service calls quicker than would be possible without the technology. That use case moved into production in late July, with about 700 customer associates using the platform.
The marketing and customer center tests are among 185 use cases the bank has identified. Ally is working to fit those in areas where they will provide the biggest benefit to employees. Teams within the organization are given about 30 days to experiment and determine if a use case will prove to have a large impact, Muthukrishnan said.
Ally has two more use cases on the horizon. One focuses on content creation from Muthukrishnan's technology team. And the other involves the bank's investor relations team testing Ally.ai for tasks such as quickly summarizing bank earnings statements. If the results are strong, the two tests will move forward with the governance process next year.
Ally plans to execute at least one use case a month.
"We are taking our time; we are learning; we are refining, and from use case one to use case two, we are getting faster and faster at implementing these," he said.
As more banks incorporate generative AI to boost productivity, Ally hopes to be a leader in the financial industry in terms of responsibly adopting the technology.
"There is always a long list of capabilities that we are developing for our internal and external customers, and we don't have enough developers or workforce to get that all done," Muthukrishnan said. "By using AI, we aim to be speeding up those deliveries, so we can bring products and experiences to our customers much faster than we do today."
While Ally is based in Detroit, Charlotte is a key market for the bank. It has more employees here than in its headquarters city.
