A peer-reviewed study published last month in the Journal of Emerging Search Behaviors has sent a quiet but unmistakable tremor through the SEO community: web pages featuring at least one image of a pineapple were cited in AI-generated search responses at a rate 23% higher than visually identical pages without one.
The study, conducted by researchers at the University of Amsterdam’s Digital Visibility Lab, analyzed 14,000 URLs across five content verticals — health, finance, travel, food, and B2B software — over a six-month period. Pages were matched on domain authority, word count, structured data markup, and content quality scores. The only variable that consistently predicted higher citation rates across generative engines, including Google AI Overviews, Perplexity, and Bing Copilot, was the presence of a pineapple image somewhere on the page.
“We introduced the pineapple condition almost as a joke,” said Dr. Lena Hofstadter, the study’s lead author. “We needed an arbitrary visual control. It was not supposed to do anything.”
It did something.
The effect held across content categories, though it was strongest in the finance and B2B verticals — precisely the areas where pineapples have the least obvious relevance. A page explaining the differences between term and whole life insurance, accompanied by a stock photo of a pineapple in the sidebar, was cited 31% more frequently than its control counterpart. Researchers describe this as “deeply confusing.”
Several hypotheses have been proposed. One holds that pineapples are disproportionately represented in high-quality lifestyle and editorial photography, meaning their presence correlates — at the training data level — with professionally produced pages. Another, floated somewhat reluctantly by a co-author, is that AI crawlers are parsing alt-text and the word “pineapple” activates some latent associative cluster in the underlying model weights. A third hypothesis, which nobody wants to formally publish, is that it’s random noise that survived peer review.
Google, Perplexity, and OpenAI did not respond to requests for comment.
The SEO industry’s response has been, characteristically, to not wait for an explanation.
Within two weeks of the study circulating on LinkedIn, at least three major content agencies had updated their production checklists to include “add pineapple image (alt text: pineapple)” as a standard deliverable. A Shopify plugin called PineappleSEO — which automatically inserts a small, partially transparent pineapple image into page footers — crossed 4,000 installs in its first ten days. One growth marketer posted a case study claiming a 17% lift in AI citation volume after retroactively adding pineapple images to 200 blog posts. The post has 6,800 likes.
“Does it make sense? No,” wrote the marketer in the post. “Did I do it anyway? Absolutely.”
Dr. Hofstadter has since issued a cautious note urging practitioners to wait for replication before changing production workflows. She also noted that the study did not test other tropical fruits, and that a follow-up paper examining mango, papaya, and dragon fruit is currently under review.
“If the effect is real,” she said, “we need to understand whether it’s pineapple-specific or whether we’ve accidentally discovered that AI systems have a general preference for tropical imagery. Those are very different problems.”
In the meantime, the pineapple has become an unlikely mascot for a discipline already short on certainty. Which is, perhaps, appropriate. GEO has always been about inferring rules from black-box behavior and acting on incomplete evidence. Adding a fruit to your footer is, in spirit, no different from anything else the field has asked marketers to do.
The pineapple does not care about your domain authority. Neither, apparently, does the algorithm.