They say that the eyes are the window to the soul, but can they also get you to buy cereal? Researchers at Cornell University’s Food and Brand Lab think so.

Considering the evidence that brand mascots and “spokes-characters” can create a “strong aura of trust among children” and that eye contact can “cause people to view others as more attentive, assertive, socially skilled, competent and credible,” the researchers figured that locking eyes with cereal box characters would increase people’s trust toward those characters and give them more positive attitudes about the associated cereal brands. 

To see how much eye contact characters make with consumers and whether that had the effects they thought it would, the team looked at 86 different spokes-characters on 65 types of cereal in grocery stores in New York and Connecticut. For each character, they constructed a side view of the eye and calculated the angle of its gaze and the height of its gaze at a point four feet from the shelf—the typical distance at which a shopper stands in a store aisle—to where the characters would make eye contact with people. 

They found that the mascots on cereals marketed to adults tended to “look” straight ahead or slightly upward. The characters of kids’ cereal boxes, meanwhile, usually had their eyes trained downward. Given the boxes’ placement on the shelves—grown up cereals up top, kids’ cereals toward the bottom—and the angle of their gazes, the average gaze height for the adult cereal characters was about 54 inches off the ground, and for the the kids’ cereal characters it was around 20 inches off the ground. The result, the researchers say, is that “cereal spokes-characters marketed to adults make eye contact with adults, while spokes-characters on children’s cereal boxes make eye contact with children.”

But does that eye contact lead anywhere? In a second experiment, the researchers looked for a link between eye contact and shoppers’ feelings about a brand. They recruited 63 people from a college campus and showed them one of two versions of a box of Trix. On one box, the Trix Rabbit was looking down at a bowl of cereal. On the other, he’d been Photoshopped so he was looking straight out at the participants. While looking at these boxes, the participants rated how much they trusted the brand, how connected they felt to it and were asked whether they would buy Trix or Fruity Pebbles cereal. 

People who had the rabbit looking at them said they felt more trust for the Trix brand and a greater connection to it, and also chose Trix over Fruity Pebbles more than the ones who looked at the box with the rabbit looking away. It looks like eye contact, even with a cartoon animal, can influence people’s feelings about a product and the choices they make at the grocery store.

Which isn’t to say that this is intended. The researchers say that they’re not insinuating that the characters are deliberately designed to make eye contact with shoppers, but their eyes do tend to meet those of the people they’re marketed to. While no one is accusing cereal makers of trying to get kids to beg for sugary cereal with eye contact, it’s a strategy that would appear to work if they wanted to try it. And the researchers think that the same tactic could be used to promote healthy eating. If healthier cereals started featuring mascots that make eye contact with kids, they say, that could improve children’s attitudes toward those cereals, “encouraging healthier choices and consumption.”

Now, this study has its shortcomings, some of which are brought up right in the paper. The researchers point out that their sample of cereal boxes was limited to 10 stores, and that shelf height may vary in other supermarkets and in other parts of the country, changing the height at which spokes-characters would make eye contact with people. They also acknowledge that people probably spend less time looking at boxes and making eye contact with the characters in a store than the participants in their experiment did, potentially lessening the effect that eye contact has. 

Physicist Donald Simanek has called out other problems with the study, which he calls a “shoddy parody of research,” in a post on his website. One problem, he says, is that the second experiment only found a link between eye contact and connection with a brand in adults, but drew their conclusions without testing for the same effect on kids. He’s also critical of how the characters’ gaze angles were calculated by taking a cartoon eye with a 2-D elliptical shape and modeling it as a 3-D sphere without any explanation for the assumption they’d be spherical. Finally, both Simanek and General Mills (which, to be fair, is not an uninterested party in the matter) dispute the idea that, with a gaze height of 20 inches (or lower for boxes on the lowest shelves), these characters would even be making eye contact with kids. Twenty inches is an average length for newborn babies, but kids walking through grocery store aisles on their own are tall enough that cereal box characters would be looking below their eye level. Whatever power a character’s stare might have, it wouldn't do much good directed at a kid’s torso.