Robots need to feel trust too

I recently attended the Agency Management Institute’s Build a Better Agency Summit. (It was great, btw.) Liza Adams who is the Founder, AI Advisor &GTM Strategist of GrowthPath Partners delivered one of the keynotes. She said something that got my wheels spinning, “Decisions are being made when a human is not in the room.”

She was of course referring to our increasing use of AI to make recommendations, and increasingly, actual purchase decisions. This happens when we ask for recommendations or a “list of the best x”. We’re relying on its knowledge to help inform our decisions. The reality of course is that those decisions are based on what others, including other bots and real people, are saying about our products and services. It’s based on brand websites, ratings, reviews, Reddit threads, Glassdoor, random social posts, blogs, and all down the line. As Liza said, all available information is now being used to make a decision simply because AI is really good and finding and digesting huge amounts of information. Before now people would likely find some of the content and review it, but a lot was not used in every decision flow.

And here’s the interesting thing: the bots are taking all those inputs and sifting through them to try and give us the most trustworthy recommendation possible based on observable patterns. Which brings us full circle to our core belief: in financial services where trust is absolutely necessary for someone to confidently choose a product, the path to earning that trust is by being ruthlessly helpful.

So what does being ruthlessly helpful look like if decisions are being made, as Liza says, when there’s not a human in the room?

We believe it still looks a lot like what is needed when humans are in the room. Things like:

• Clearly communicating who the product is good for, and for whom it may not be the best fit. Don’t make the agent or person guess…tell them straight out.
• Framing products not through a list of RTBs but how they solve the problems people need solving. Most people acquire a new financial product because they have a specific need staring them in the face. That could be they’re buying a home, or need to consolidate debt, or even want to juice their frequent flyer miles to get to that big trip faster. Whatever problem your product solves tell it clearly.
• Making sure there are no dead ends. If product X isn’t right for them, perhaps product Y is. And if neither works, tell them that and give them a path forward, whether helping to build credit or pointing them to other resources
• Using social proofing to demonstrate the value of the product, through a combination of ratings, reviews, and customer stories. In a world where bots are making decisions on what gets into the consideration set, the weight of real people’s experiences will ironically become even more weighty.

And of course how we deliver helpful content is evolving, because it needs to be helpful to both humans reading it and LLMs digesting it.  

This may sound like a lot of work. And if you aren’t already on the path of creating helpful content, it is. But the great news is that it’s work that helps every single person or bot that is considering your products. You don’t need to optimize just for people or just for an agent. If you simply tell your product story, in helpful, tangible, problem-solving ways, you’ll win the hearts of the people, and the bots.