AI For All Ages?

Can AI really ignore this section of our community?

This article is from my newsletter series called “Personalized! Prompts from the Frontlines.” You can sign up for my newsletters here: https://dedelman.substack.com/

Over the holiday weekend, I had the pleasure of spending time with my 86-year old mother-in-law, a remarkable woman who is still as sharp as ever.

But seeing her struggle with various online actions that she needed to complete, simply to function in today's society, really worried me.

She's very smart, but so much she had to do was far from intuitive.

And that has ramifications for all of the dreams we have as marketers to distinguish our branded experiences through personalization.?

Similar to the discussions about bias in AI models which have not had experience with certain populations, those who are designing customer experiences in most companies have significant biases and blinders.?

They are probably not thinking about seniors, despite one out of six people in the US being over 65 years old, a figure that is increasing.

They are probably also not naturally thinking about many populations, since most applications and systems designers are so deep in their tech, and all the business people around them so familiar with the digital tools they use everyday, that they don't think about situations such as an older iPad's memory is now full, a child of someone who recently passed away has no means of getting through all of the security hurdles to a parent's accounts, or someone has no idea that the "hamburger" icon opens up a menu. And on and on.?

Personalization is about creating value for a customer based on understanding their context. But context is much more than what they last bought and which emails they respond to. It means thinking about how you are designing experiences in the first place.

AI is not going to provide the answers, or at least it won't if you are not asking it the right questions.

We still need real research: Ethnographics to get a much deeper understanding of the compromises different people face and the needs they have.

And especially what goes wrong.

Every time I work with a client on designing new experiences, they immediately put all of their energy into conceiving the perfect flow. It's just not a natural tendency to elevate the questions about all the things that could go wrong from a customer's point of view, not just from a system bug perspective.?

Nonetheless, AI can now play a critical role, with the right tools pointed towards the right questions.? For example, tools such as DataOrb can turn the unstructured exchanges in every call center interaction into data, tagged with information about the customer, that can be then used to mine for all of the issues raised by various customer segments.

Genesys' Journey Analysis tools can track back all of the past interactions someone had across their journey, illuminating who had issues taking various actions, especially before they called for help.

Stitched Insights can mine all the text across reviews and social media about yours and competitors' products, discern characteristics about someone from the syntax they use in their posts, and provide data on challenges that specific segments are raising publicly.?

So, as we rush to design personalization "engines' powered by automated AI, let's make sure we step back and think about how we are getting deeper customer insights.?

It will take exploring AI-driven tools that go beyond the current heavily-hyped GenAI LLM's augmented by a willingness to get out there and actually talk to people. When done right, they'll feed each other, as the AI suggests areas to learn about and the research suggests new prompts to test. Strategy and design still matter. Who are our customers?? How do we serve them better??

I would love to hear more from others on how they are mixing the new insights from AI with real ethnographic research.

I'm sure there are some new approaches emerging, and ways to use AI more directly in the ethnographic research itself.

What are your experiences?