Paying to appear? But does AI actually recommend you?

Hello friends,

It’s been a big few weeks at the LEOPRD HQ.

We launched Reputation to Revenue 2026 Report, our biggest study yet into how people use AI to make decisions and what makes one company more recommendable than another.

If you haven’t read it yet, here’s the finding to know: appearing in an AI answer does not mean you are winning the recommendation.

When people asked AI for a recommendation without naming a brand, the companies we monitored were:

• Missing 36% of the time

• Mentioned but passed over 35% of the time

• The primary recommendation just 29% of the time

That middle group is especially interesting. If AI knows who you are and repeatedly chooses somebody else, you may not have a visibility problem. You may have a reason-to-choose problem.

More on that below, along with the latest news to have on your radar.

Until next time,

Celia and the LEOPRD team 🐆

P.S. We’ll be hosting breakfast briefings soon in Sydney, Melbourne and Brisbane to unpack our report findings. Keen to join us? Hit reply and tell us where you’re based.

SYDNEY, BRIGHTON, TOKYO AND BEYOND

The LEOPRD team is clocking up some air miles.

Our Europe-based Strategy Director, Rik Turner, is at BrightonSEO on Friday 9 October, talking all things LEO and GEO. Book some time with him, or look out for the guy in the Adidas leopard-print tracksuit top.

Celia will present the findings from Reputation to Revenue 2026 in person for the first time at HumAIn in Sydney on 13 October. Going along? Come and say hello.

On 4 November, Celia will join an international panel at the World Media Intelligence Congress in Tokyo to discuss the new rules of being found in a GEO world. She’ll be joined by leaders from GlobalNews Group, press relations and Argus Data Insights, with Muck Rack moderating the conversation.

As our work scales in Europe through our partnership with The Remarkables (our tech and methodology power The Mark), we now have boots in the ground in Lisbon. Meanwhile, Louis in the team has moved to Brazil. 

🌏 Part of a global business with teams in these markets? Let us know if you want to chat, host a workshop or collaborate.

THINGS WORTH KNOWING

1. ChatGPT is becoming an interface for discovery, decisions and advertising

ChatGPT is growing out of the text box. ChatGPT is growing out of the text box.

OpenAI’s new Intelligent UI can add interactive charts, maps, calculators, diagrams and buttons directly into a conversation, according to Wired. Instead of simply reading an answer, people can explore options, compare information and take the next step without immediately leaving ChatGPT. 

At the same time, The Verge has reported that OpenAI is introducing visual ads during image generation, with US testing planned for Free and Go users. The ads will be labelled and kept separate from the image being created, and OpenAI says they will not influence ChatGPT’s answers.

ChatGPT is becoming a place where people can discover, plan, compare, create and encounter advertising within the same experience. That creates two questions for marketers. Can the information around your products and services support useful comparisons, calculators, maps and next steps? And can you distinguish what paid placement delivers from what earns an unpaid recommendation?

Our read: Decide whether you are buying attention, traffic or sales, and measure that separately from organic visibility and recommendation. Make your information clear and useful enough to work inside a more visual, interactive answer. A sponsored appearance can put you in front of the customer. It does not give AI an independent reason to choose you.

2. Retailers are deciding whether to let AI shop for you

AI shopping agents research products, compare options and will eventually complete purchases for customers. The Wall Street Journal recently reported that Walmart, Gap, Best Buy and Shopify are exploring the opportunity, while others are restricting access over concerns about customer data, fraud and losing control of the buying experience.

Our Reputation to Revenue 2026 findings help explain why this matters. We found that 44% of prompts involved research, advice or judgement, while AI appeared around 20 times more often before purchase than after it. People are already asking AI to help them decide. Shopping agents could take the next step and act on that advice.

For retailers, this creates a trade-off. Agents could introduce new customers and make purchasing easier. They could also weaken the direct relationship with the customer and leave businesses with less control over how products are compared, sold and supported.

What it means: Decide what shopping agents can access and do before they become a major route to purchase. Accurate product information matters, alongside clear rules for consent, customer data, service and responsibility when something goes wrong.

3. Pinterest is turning inspiration into a shopping brief

A saved hair photo is useful, until your stylist asks what you actually want. Pinterest’s new Beauty Guides are designed to supply the missing detail. The feature analyses hair and nail Pins, then explains the look, what to ask for at the salon, similar ideas and which products could help recreate it.

For marketers, this is where it gets interesting. Pinterest can now turn an image into the language a customer uses, the products they consider and the action they take next. The platform is helping shape the brief, not simply displaying the inspiration.

Our read: Make your visual content easy for AI to interpret. Name the finish, explain the technique, identify the product and describe who it suits. Leave those details out and the platform may fill in the gaps itself, potentially using somebody else’s language or recommending somebody else’s product.

FROM THE FIELD

• Celia recently spoke to Nigel Bowen at Influencing about why being cited and being chosen are two different tests, and what the findings mean for PRs and journalists.

• Ahead of speaking at HumAIn next week, she also wrote a piece for its blog looking at what gives AI a reason to recommend one company over another. Spoiler: describing yourself as “trusted”, “innovative” and “leading” gives it very little to work with.

That’s what’s been on our radar.

Wondering how your brand shows up in AI? Hit reply and ask us. We’d love to hear what’s on your mind.