Who Talks to the Machines?

Who Talks to the Machines?

When we talk about AI, we obsess over the technology. We obsess over the algorithms, the breakthroughs, the headlines. But what we don’t talk enough about is the people. Because the truth is: AI isn’t about machines. It’s about who’s actually talking to them.

Let me give you three examples.

Perplexity

First, Perplexity. The data is clear: around sixty percent of its users are men, and the vast majority are between eighteen and thirty-four. Young, educated, tech-savvy. These are students, knowledge workers, urban professionals. Perplexity is their research engine. They don’t just want answers, they want sources, citations, validation. If you’re a brand in nutrition, beauty, baby products—categories where consumers compare ingredients and pore over details—Perplexity is where the young, affluent, male early adopters are shaping opinions.

ChatGPT

Now contrast that with ChatGPT. ChatGPT is broad. Its gender split is almost fifty-fifty. Yes, it’s hugely popular with Gen Z and Millennials, but a third of its users are over thirty-five, and even grandparents are starting to use it. This isn’t just the playground of techies anymore. It’s the new Google—universal, mainstream. If you’re an FMCG brand, this means ChatGPT is influencing purchase consideration across every segment. From students discovering snacks, to parents meal planning, to retirees asking for health advice.

Google Gemini

Then there’s Google’s Gemini. Integrated into Search, it touches billions of people every single month. And it looks a lot like ChatGPT demographically: young and slightly male at the edges, but rapidly becoming universal. When you ask Google a question and an AI summary pops up, that’s Gemini. Which means your brand lives or dies by whether it’s visible in those overviews.

Walmart Sparky

And then, let’s not forget the retailer assistants. Walmart’s Sparky is already being used by fifteen percent of shoppers. Who are they? Predominantly women, suburban moms, family planners. They’re value-conscious, pressed for time, using Sparky to build shopping lists and plan parties. Amazon’s Rufus, too, is in the hands of busy parents and Prime households, summarising reviews and helping them buy with confidence.

Target Bullseye

Target Bullseye is Target’s AI gift and product assistant, currently running in niche pilots like the Gift Finder and select PDP chats. Its users are mostly 25–44-year-old Millennial parents, with a slight female skew—shoppers looking for style, curation, and gifting inspiration. Bullseye helps them explore curated picks, clarify product details, and find items that fit their taste and trends. For brands in toys, beauty, home décor, apparel, and premium store lines, visibility here matters because Bullseye influences consideration and conversion. It responds best to clear product attributes, occasion tags, and sustainability cues, while trust is driven by brand signals and social proof.

Amazon Rufus 

Amazon Rufus is Amazon’s shopping assistant, already in the hands of about 14% of shoppers. Its users are mostly 20- to 40-year-old Prime members—busy parents and professionals looking for convenience and confidence. Rufus helps them compare products, read reviews, and decide what’s “best for me.” For brands in household, baby, beauty, and pantry categories, being visible here matters because Rufus drives real purchases. Time-pressed shoppers trust review consensus and Prime reliability, so the AI responds best to clear benefits, review highlights, and thorough product details.

Summary

So, what does this mean for brands? It means AI is no longer optional. It’s where your customers are making decisions. The early adopters on Perplexity are tomorrow’s opinion leaders. The mainstream on ChatGPT and Gemini are already searching and learning through AI. And the household decision-makers on Walmart and Amazon’s assistants are buying through AI.

If you’re not visible inside these systems, you’re invisible to the people who matter most.

So the question is: when your customer talks to the machine, will the machine talk about you?

Sources