Criteo’s Todd Parsons: Search Is Not Dead, Just Different
CANNES, France — Performance marketing has always chased consumer intent. Among the promises that artificial intelligence is making right now involves something potentially something more valuable; namely the opportunity to juice intent before the consumer discovery process even forms.
“You don’t always go with the concept of asking for a comparison between the best dog foods for your dog. You might ask GPT, ‘What do I do if my dog’s sick and throws up almost every time it eats?’” Todd Parsons, chief product officer and president of Performance Media at Criteo, told Beet.TV contributor David Kaplan at Cannes Lions. “That’s a great opportunity for a seller of dog food that is doing something that might solve the problem to get in that conversation. That’s discovery.”
Criteo announced recently that it has onboarded 2,000 advertisers to OpenAI’s platform through a four-month API-level partnership that goes beyond demand generation into co-developing how OpenAI advertising actually works.
How search works now
Rather than compressing the consumer journey into zero-click commerce, AI discovery environments are delivering incremental audiences that traditional performance channels rarely reach. In Parson’s view the sense that searchers are simply looking to AI generated answers, as opposed to search results of infinite blue links, is a false dichotomy. Search still works to send people to websites. It just works differently now.
“We’re seeing 80% of traffic from OpenAI landing on our retail partners as being net new to brand. If you’re looking at a surface where a conversation’s happening and you’re getting 80% new people to a brand, that is a super powerful signal,” Parsons said.
Retailer experiences continue thriving alongside AI discovery, with loyalty, convenience, and selection keeping established purchase destinations relevant rather than obsolete.
Two objectives, one conversation
The conventional separation between upper-funnel brand advertising and lower-funnel performance marketing dissolves in conversational AI environments where a single multi-turn exchange can move a consumer from unawareness through consideration to purchase.
“AI is going to be a stronger machine for brand advertising. Because of the conversation, you can reveal these things in a very natural way, high up in the funnel,” Parsons said. “Because we’re able to see those signals and follow them all the way through to the point of purchase, they sort of become one routine where in the end you’re just selling more product or bringing new attention to a brand — all in one shot.”
Measurement’s organic-to-paid path
Criteo’s 17,000 commerce customers provide a large enough base to observe how organic and paid AI traffic flow differently to points of purchase, building the measurement foundation that will eventually make discovery a standard performance routine.
“We see how organic traffic is going to our points of purchase and now we get to observe how paid traffic is coming to those points of purchase. Seeing those two things together and measuring them — we are already seeing the path for how to monetize better for every one of our sellers,” Parsons said.
MCP opens commerce intelligence to partners
Criteo is opening its commerce assets, which include product discovery intelligence and point-of-purchase behavioral data, through model context protocols (MCP) that allow enterprise and small-to-medium business partners to consume that intelligence in whatever format fits their existing workflows.
“We’re opening up all of our assets for commerce, the intelligence behind what products are being discovered, what behaviors happen on the point of purchase, in the way that those partners can consume the easiest,” Parsons said. “Depending on the product that’s coming in, it can be experienced the way our partner wants to experience what we have to offer.”
CANNES, France — Performance marketing has always chased consumer intent. Among the promises that artificial intelligence is making right now involves something potentially something more valuable; namely the opportunity to juice intent before the consumer discovery process even forms.
“You don’t always go with the concept of asking for a comparison between the best dog foods for your dog. You might ask GPT, ‘What do I do if my dog’s sick and throws up almost every time it eats?’” Todd Parsons, chief product officer and president of Performance Media at Criteo, told Beet.TV contributor David Kaplan at Cannes Lions. “That’s a great opportunity for a seller of dog food that is doing something that might solve the problem to get in that conversation. That’s discovery.”
Criteo announced recently that it has onboarded 2,000 advertisers to OpenAI’s platform through a four-month API-level partnership that goes beyond demand generation into co-developing how OpenAI advertising actually works.
How search works now
Rather than compressing the consumer journey into zero-click commerce, AI discovery environments are delivering incremental audiences that traditional performance channels rarely reach. In Parson’s view the sense that searchers are simply looking to AI generated answers, as opposed to search results of infinite blue links, is a false dichotomy. Search still works to send people to websites. It just works differently now.
“We’re seeing 80% of traffic from OpenAI landing on our retail partners as being net new to brand. If you’re looking at a surface where a conversation’s happening and you’re getting 80% new people to a brand, that is a super powerful signal,” Parsons said.
Retailer experiences continue thriving alongside AI discovery, with loyalty, convenience, and selection keeping established purchase destinations relevant rather than obsolete.
Two objectives, one conversation
The conventional separation between upper-funnel brand advertising and lower-funnel performance marketing dissolves in conversational AI environments where a single multi-turn exchange can move a consumer from unawareness through consideration to purchase.
“AI is going to be a stronger machine for brand advertising. Because of the conversation, you can reveal these things in a very natural way, high up in the funnel,” Parsons said. “Because we’re able to see those signals and follow them all the way through to the point of purchase, they sort of become one routine where in the end you’re just selling more product or bringing new attention to a brand — all in one shot.”
Measurement’s organic-to-paid path
Criteo’s 17,000 commerce customers provide a large enough base to observe how organic and paid AI traffic flow differently to points of purchase, building the measurement foundation that will eventually make discovery a standard performance routine.
“We see how organic traffic is going to our points of purchase and now we get to observe how paid traffic is coming to those points of purchase. Seeing those two things together and measuring them — we are already seeing the path for how to monetize better for every one of our sellers,” Parsons said.
MCP opens commerce intelligence to partners
Criteo is opening its commerce assets, which include product discovery intelligence and point-of-purchase behavioral data, through model context protocols (MCP) that allow enterprise and small-to-medium business partners to consume that intelligence in whatever format fits their existing workflows.
“We’re opening up all of our assets for commerce, the intelligence behind what products are being discovered, what behaviors happen on the point of purchase, in the way that those partners can consume the easiest,” Parsons said. “Depending on the product that’s coming in, it can be experienced the way our partner wants to experience what we have to offer.”
Series Videos
