Telling AI stories

Why do AI projects fail (one of seven reasons)?

How many times do you see executives and business users being bombarded by technical teams with numbers, facts, and figures? They try to “sell” the technical solution but shower business colleagues with irrelevant information. They overwhelm business executives and users with technical features of the AI solution without conveying what’s in it for the business. Or what’s in it for the user of that AI solution.

This is counterproductive due to how the human brain operates. The human brain processes new information first to identify if it’s a threat and second how relevant the information is.

If the presentation of the new AI solution comes across as a threat to the business user, e.g., “we will need to change how we operate”, or doesn’t convey the relevancy for business goals, it will be dead on arrival. Throwing more specs, facts, and numbers at business users doesn’t help. It’s often the opposite: it increases resistance to using AI. It leads to death by PowerPoint.

Developers and data leaders at successful companies know how to communicate to business users the benefits of using AI solutions: they tell data and AI stories.

What is an AI story?

It’s a narrative of how an AI product is used by the business user in his or her daily life. It’s intended to make the AI product very tangible to the user. Ideally, this narrative is told from the business user point of view.

When my team was developing an AI recommendation system for sales reps at a pharma company, I wanted to ensure that the users understand the value this AI system provides. I sought support to develop a data story from a data storytelling consultant. With her help, we have created several stories to market the AI recommender system, also called Next Best Action, to the sales reps.

Like with movies, there are many ways to tell a story using different protagonists. Here is an example of one story that we have developed.

It’s a rainy Wednesday afternoon, when Monika, our sales rep, arrives at the café and finds a parking spot nearby. She has had a terrible day and now the rain is getting stronger and stronger. But she is looking forward to meeting her sales colleague, Alex. She wants to let out her frustration about a customer visit this morning, who stood her up. She gets out of the car and runs into the café.

Alex is there and has already grabbed a table. They order coffee. The second they sit down, Monika lets her frustration out.

“You know what happened to me again? Today I had an appointment with Dr. Mayer at 10 am. I got there after two hours of bad traffic, the assistant was rude, and it turned out that the doctor was not even there. This is happening to me several times a month. Can you believe that? Now I have lost half a day.”

Alex is listening and then tells her that he absolutely understands what she is talking about. “I had the same problems in the past, too.”

“What do you mean IN THE PAST ?”, Monika wants to know.

Alex mentions then that he lost 3 work days each month with customers who were flaky. But in the last six months, this has changed. The game changer was a new tool.

“Usually in the morning, when I have a few minutes while waiting for the next appointment, I switch on the Next Best Action tool. I review its recommendations about which customers to visit.”

“I get a recommendation, for example, about a doctor nearby, who has still not received the latest product information. I then squeeze this visit into my schedule. Or sometimes it shows physicians who have not been visited for a while. If they fit into my route, I stop by. Not always, but often I get to talk with the doctor, even if it’s just for a few minutes.”

“C’mon Alex, you won’t tell me that the machine knows your business better than you!?”

“No, Monika, absolutely not. I think I’m pretty good at this job. But I have now often used these recommendations, and those frustrating moments, just like yours this morning, went down to a minimum.”

How does AI storytelling improve AI adoption?

Data storytelling is a communication method that relates to the listeners, using comprehensible narratives in the language of the listeners, and referring to the day-to-day situations they experience in real life.

With storytelling, technical features and data are packaged into compelling anecdotes that create meaning for business stakeholders. The stories highlight situations they already know and connect the information with experiences they’ve already had. Once they’ve made an emotional connection to the narrative, their brains are more receptive to learn about the new AI solution.

The AI story above helped my team to market the Next Best Action tool to sales reps. This narrative was the main hook and storyline for the kick-off event where the AI recommender system was launched. It was also used in presentations and short videos when onboarding new users. And, it was featured in several intranet articles.

The results were better than expected: After only a few months, the adoption rate of the AI recommender system went from 20% to 50%, with continuing growth. Some of the sales reps became ambassadors of this AI solution since they knew exactly what’s in it for them.