Will generative AI help retailers perfect personalization?

Most Americans love to receive personalized product recommendations and special offers tailored to their tastes. McKinsey & Company tells us it's about 71 percent of us.

Americans also get pretty ticked off when retailers and brands attempting to offer greater personalization get it wrong time after time and don't appear to learn from past experiences. That's 76 percent of us, McKinsey says.

Separate research from Twilio finds that 86 percent of consumers are more loyal to brands based on personalized experiences. Two-thirds of consumers also walk away from brands when they feel they are getting cookie-cutter experiences rather than personalized ones.

Most of the personalization at retail now has machine learning and/or artificial intelligence (AI) technology at its number crunching and purchase pattern identifying core. Those who know tell us that new generative or conversational AI built on large language models will be able to better understand our wants and needs and provide substantially higher levels of personalization than the current systems, which are based on predetermined data sets.

The new systems are said to be substantially better at anticipating the wants and needs of customers and producing more personalized product recommendations and deals.

They're also supposed to be able to understand nuance and will be able to receive direct feedback from users to further elevate the level of personalization.

It makes total sense that my Target Circle offers, for example, include discounts or coupons on items I've bought many times before because these are repeat purchase items. The current system works in this case.

Where Target falls short, however, is me being able to tell it that I don't want fully ripe bananas when I place a grocery order for Drive Up. I prefer a little green in the skin since it will take three or four days for me to go through what I've ordered.

The current Netflix system is excellent at putting a percentage next to movies and shows predicting the likelihood that I will enjoy viewing them based on my past viewing and ratings of the content on the platform.

Its deficiency is that it continues to recommend movies and shows that I've previously viewed and given a rating of one or two thumbs up. The whole point of Netflix is that I can continually find new content based on what I've liked in the past and not have to do repeats.

Another quirk in the Netflix system is that it will give an 80 percent rating to a show that I previously viewed and gave a two thumbs rating. I've already confirmed how much I liked the show. Does Netflix's AI doubt me?

My nitpicking aside, I've been reminded by minds more versed in the subject matter than myself that AI, whether operating on fixed data sets or using generative technology, is either in its toddler or infancy stage and that better results are on the way. I'll give two thumbs up to that! Did you get that Netflix?