Exploring the Potential of AI: Ally’s Marketing Team Tests the Waters
- 5 min read
By Zack Flannick and Kevin Howard
The Greek philosopher Heraclitus famously said, “change is the only constant in life.” That wisdom certainly holds true in the world of marketing. As marketers, we are constantly adapting to new technologies, trends, and consumer behaviors. And now, we find ourselves at the forefront of a new era, where generative AI has the potential to transform the way we operate.
As a digital-only bank, we have an appreciation for how technology can make banking easier for consumers. With generative AI hitting the scene earlier this year, we wanted to explore how it could be used to assist our marketing team. Before we go into the details and in case you missed it, Ally released the early results of our first marketing experiment with Ally.ai, the company’s proprietary AI platform.
The Making of an Experiment
An important facet of building successful experiments featuring new technology is education and training. Eleven Ally teammates were excited to use Ally.ai, and in fact, we all thought about this as the first step in a transformation — an opportunity to build our personal skillsets while also figuring out how best to bring new technology into our established processes. Each of us was required to go through access approvals and training. This was especially critical as we built the experiment to get the most out of the large language model (LLM).
As with any experiment, setting the right objectives and then ensuring we could determine if we were hitting the mark was our biggest priority. Our objective was to measure the usefulness, accuracy, and time savings that Ally.ai and its connected enterprise LLM could provide. We tracked and evaluated its performance with these measurements across 80 unique marketing tasks. Would Ally.ai help us to move more quickly, or boost our productivity? Would the platform give us outputs that we could use? Let’s find out!
Promising Results Flow from the Experiment
We embarked on a test using Ally.ai’s open prompt desktop UI. Starting in September and in close collaboration with our technology partners, we utilized this tool for various marketing activities over the course of one month, including tasks like content creation, summarization, ideation and data / information analysis. The early results are promising! Ally.ai reduced the time needed to complete these tasks by 34%. Furthermore, Ally.ai outputs were useful in 87% of cases and accurate in 81% of cases. The feedback from our experiment members was very positive, with a consensus that having access to Ally.ai for use in daily work is a valuable and efficient tool for completing certain tasks.
Certain tasks. That’s the key phrase.
We found that, yes, Ally.ai was helpful in creating first drafts, summaries and creative springboards to things like social media content, scripts and articles. For example, one of the experiment participants was able to create an Instagram reel script in about 10 minutes, which would have otherwise taken about three hours. Another experiment participant was able to generate an article based on a podcast script in about one hour, which would have otherwise taken four hours.
We also discovered some restrictions of the tool, which helped us understand what types of tasks we should use it for. For example., if we had a massive amount of data to input in the prompt, we had to chunk it up into sections. The tool can also provide inaccurate responses. While convincing, these answers were incorrect — and the way to catch those instances is to have human oversight of all outputs, so responses can be fact checked, validated and then reviewed within our normal approval process. Hence, one of our principles for all AI use cases is to have a human in the middle.
This experiment demonstrated the value Ally.ai brings to the creators within the marketing organization. We can speed up some of our content creation, making us more efficient and even more productive in our day-to-day work.
Another bonus that came from this experiment: We were able to sit down with the Ally Tech Labs team — one of the innovation arms inside the company — to share feedback on the Ally.ai user interface (UI). It was an important part of the experience for the experiment participants — we found that the UI was crisply designed, easy to use and properly delivered the outputs of our inquiries. Our team was able to provide this and other feedback on the UI so participants in the next experiment, no matter where it is within the company, can see improvements in their experience.
Embracing The Potential of Generative AI
The results listed above are only the beginning of our testing and exploration of generative AI for marketing purposes. As our transformation and evolution with AI progresses, we will be able to improve literacy, broaden access, discover more prompts and test emerging advancements within the field. Thanks to the efforts of our technology partners, we are acting on all those fronts and will continue to evolve as a marketing group.
In addition, we have an enterprise-wide cross functional AI Working Group at Ally who work with our AI experts to review and advise on generative AI use cases. Humans are actively trained and involved in the evaluation and use of generative AI technology. Any time we launch an experiment or explore the use of generative AI, training and oversight mechanisms are fully defined ahead of time. Being able to experience the technology in a secure environment and to see how it can positively impact our work will help solidify our position as innovators in the financial services industry.
As we continue learning about generative AI, change is inevitable. Change can bring out a myriad of emotions, anywhere from excitement to anxiousness. However, we like Aristotle’s optimistic view when he said, “Change in all things is sweet.” What he meant is that change actualizes the potential of a subject. In other words, change is the process of bringing out potential.
At Ally, we think AI will enhance our potential. Potential in our creative abilities, potential in our communication, and potential in developing solutions that will make our work even better. We’re excited for the future and the opportunity before us.
A very special thanks for the 11 marketing teammates that participated in the experiment — you all were awesome! Also, a big thank you to our technology partners who built Ally.ai and guided us through the process.
Lastly, in case you are curious if Ally.ai played a role in the drafting of this article, we’ll leave it to you to decide. 😊
Interested in joining Ally's team of talented technologists to make a difference for our customers and communities? Check outAlly Careersto learn more.