效果营销重置如何影响开放网络，LLM与聊天机器人成为新入口

This post was created in partnership with Taboola

Key takeaways

Consumers exhibit different patterns and preferences depending on the industry.
Brands need to optimize so they show up in results through chatbots and the open web.
As always, when it comes to tapping into AI, the human component remains key.

Richer data and agentic technology are shifting performance marketing beyond programmatic buying and into an era of guiding audience decisions in real time.

During an ADWEEK House Advertising HQ group chat co-hosted with Taboola, industry leaders explored how new capabilities could change the way marketers find audiences, make decisions, and drive outcomes across the open web.

(L-R) Allureâs Jessica Cruel, Taboolaâs Adam Singolda, Yahoo DSPâs Beth Gross — (L-R) Allure’s Jessica Cruel, Taboola’s Adam Singolda, Yahoo DSP’s Beth Gross

Signals depend on the category

The speakers began by discussing the clearest signals of what someone actually wants to buy.

Adam Singolda, CEO of Taboola, said search is the best because it is literally telling you what the person wants as they are about to make a decision, but the problem is it’s very expensive.

“We’re always thinking about the mid-funnel of when consumers are starting to consider buying something, but are not yet there,” he said. “And that’s a great consumer because they’re thinking about it, spending real time reading, watching.”

Singolda sees the middle as the place where large language models (LLMs) are thriving because they spark curiosity. However, he said, the category determines whether the person goes to search or the LLM first.

For the health and pharmaceutical category, Ian Colley, CMO of DeepIntent, said he finds that when people have a condition or ailment, they increasingly start their query journey with AI.

“They’ll have a conversation with ChatGPT or Claude about: ‘What is this I’m feeling? What is it I’m experiencing?’ But then they will fairly quickly go to the healthcare provider to carry that journey forward,” he explained.

“I think the art and science of marketing varies drastically by industry. If you’re selling a dollar bag of M&M’s, it’s very different from selling a Mercedes; it’s very different from selling a pharma treatment,” noted Colley.

Jessica Cruel, editor-in-chief at Allure, said that for the beauty category, LLMs are being used for research because they provide a level of personalization, such as finding a Vitamin C serum for a certain age.

“However, I don’t think purchase will happen there because there’s a trust issue, and the other thing about beauty is it’s a visual medium. So, people want to see before and afters. They want to see different skin tones using it. They want proof that this product has been tested,” she said.

(L-R) Experian Marketing Servicesâ Krystal Olivieri, Deep Intentâs Ian Colley — (L-R) Experian Marketing Services’ Krystal Olivieri, Deep Intent’s Ian Colley

Showing up in LLMs and the open web

Elizabeth Cagen, SVP and media investment and performance lead at Digitas, said more and more decisions will be made through AI chatbots.

“In order to be able to be shown within those environments like we need to, we’re talking to our advertisers about how we create these content strategies to build trust across paid, earned, shared, and owned so that you’re building all of that trust so that you’re going to be showing up in these LLMs,” she said. “You’re going to be able to provide the research and the answers that these consumers are looking for in their research journey.”

Krystal Olivieri, chief strategy officer at Experian Marketing Services, agreed with Cagen’s points.

“If your brand is not showing up the right way when you’re asking the chatbot the question, all of the investment doesn’t really matter if you have not actually made sure that the data that you have out there also is representing your brand appropriately,” she added.

Olivieri gave the example of determining her dog’s healthcare issues. She did a lot of research and used ChatGPT as her starting point.

“I spend a lot of time in the open web. There’s this false narrative that the open web is dead, and as a byproduct of that, everything is shifting,” she shared. “I fundamentally do not believe that’s the case. I think that this has made the open web. It’s given it a little bit of a resurgence because now people are starting at that chat point, and they’re moving into the open web.”

(L-R) ADWEEKâs Jenny Rooney, Digitasâ Elizabeth Cagen — (L-R) ADWEEK’s Jenny Rooney, Digitas’ Elizabeth Cagen

Remember the human element

As the group chat came to a close, speakers shared their top takeaways.

Beth Gross, head of commerce media at Yahoo DSP, said to start with the shopper, the behavior, and what you’re trying to change.

“Start with that first-party data, the deterministic data. Start there and then build out your campaign from there,” she said. “Then, leverage the open web, leverage your measurement, leverage the supply.”

Singolda’s takeaway was that while we now have so much innovation with machines, robots, and AI agents, “there’s one thing that never changes, and that is the human element, the importance of belonging to something, and culture,” he said.

He concluded by sharing: “That’s why the open web is so powerful. Because I will never, ever wake up in the morning and go to ChatGPT for the Knicks highlights.”

Featured Conversation Leaders