Complex Color Arrests Attention

Written by Ali Boluki, Ph.D. candidate at the University of Groningen (The Netherlands)

Illustration of color complexity in images. Source: Kanuri, Hughes, and Hodges (2024).

Social media feeds are crowded, fast-moving, and unforgiving. A brand may have only a fraction of a second to stop someone from scrolling. That simple problem - how to win attention before a user moves on - sits at the center of Vamsi Kanuri, Christian Hughes, and Brady Hodges' IJRM article, "Standing out from the crowd: When and why color complexity in social media images increases user engagement."

The project began not with an abstract methodological puzzle, but with conversations between Kanuri and social media teams at the University of Notre Dame. Social media managers were experimenting with text, images, timing, and color to answer a very practical question: what makes people stop?

"Attention is all you have got on social media."

- Vamsi Kanuri

This raised a broader question: can a visual feature move scrolling users toward deeper processing and make them more likely to engage? The paper suggests that color complexity can, although its effect depends on the surrounding context.

Why complex color?

Color is familiar to managers. It is central to branding, design, and visual identity. Yet color complexity is not the same as simply making an image brighter, more colorful, or more brand consistent. In the paper, color complexity refers to pixel-level color variation within an image. Images become more complex when there are more changes in color across adjacent pixels.

Colorfulness and color variety concern the number of distinct colors, whereas color complexity also captures their distribution. Two images using similar colors can therefore differ in complexity because of their transitions and pixel-to-pixel variation.

"The more switches we see between adjacent pixels, the more complex it becomes, the more information there is within an image."

- Vamsi Kanuri

This distinction matters because color complexity is both psychologically meaningful and managerially actionable. It can be measured at scale with computer vision, but it remains interpretable. A manager does not need to understand every layer of an image algorithm to understand the idea that some images create more visual information for users to process.

A simple feature in an AI era

One timely question is whether such interpretable measures still matter in the era of large language models and vision-language models. Kanuri's answer was clear: yes. He sees no problem using newer AI systems to help measure image characteristics. But he cautions against skipping the step of understanding why a feature works.

A black-box model may predict which post will perform better, but it may not reveal which specific image feature is driving that prediction. Interpretable features such as color complexity help researchers and practitioners isolate a mechanism: the image attracts attention, and attention increases the probability of engagement.

"I would still stick with the measure and still study the effect on engagement, because that allows us to tease out the mechanism."

- Vamsi Kanuri

The evidence behind the effect

The paper combines two proprietary Facebook datasets with two eye-tracking studies. Eye tracking shows where people look on a social media post and for how long, allowing the authors to test whether color-complex images capture more attention before increasing engagement. The field data, drawn from a major U.S. university and a West Coast newspaper, show that posts with more color-complex images receive more engagement.

Not everything went as expected. The team predicted that complex captions would weaken the benefit of complex images by consuming users’ limited attention. Instead, they found the opposite: complex captions strengthened the effect, perhaps because difficult text pushes users to rely more on the image. Color complexity also worked better later in the day, when attention may be more constrained, and for taller images. In addition, captions with a negative emotional tone weakened its benefit, likely because negative words already attract attention and leave fewer cognitive resources for processing the image.

"When the text is complex, people look for alternative cues to process the information, and the image gives them that alternate cue."

- Vamsi Kanuri

From attention to engagement

The eye-tracking studies were central because Facebook engagement data alone could not show the psychological process. The paper argues that color complexity shifts users from quickly scanning content – what psychologists call peripheral processing – toward expending conscious effort - called central processing. To examine this, the authors measured attention using how long participants looked at the image and how often they fixated on it.

They then measured engagement using head and facial movements as behavioral indicators of engagement. This approach allowed the team to go beyond potential biases in self-reported attention or engagement, as Kanuri noted.

The eye-tracking evidence supported the paper's story. People looked longer and more often at high-complexity images, and this extra attention helped explain why those images led to more engagement. In other words, color complexity did not merely correlate with social media performance; it helped explain why users were more likely to engage.

The value and challenge of mixed methods

For Kanuri, the value of a mixed-methods paper is that it can connect an observed effect with the psychological process behind it. The large-scale Facebook data show that the effect appears in real social media environments. The eye-tracking studies show how users respond when looking at the images. Together, the methods make the story stronger.

"A well-written multi-method paper can really seal the deal on not just the main effect, but also the process."

- Vamsi Kanuri

Kanuri nevertheless stresses that social media data have limitations. Platform algorithms and managers’ content choices can influence which posts users see, making it difficult to isolate one image feature perfectly. Researchers should therefore combine evidence from different methods and discuss these limitations transparently.

What should managers do

The recommendation is to think of visual design, caption writing, image format, and timing as connected decisions. Managers can use more color-complex images when they want posts to stand out and invite deeper processing. But they should be careful when the caption is strongly negative, because negative text may capture attention on its own and reduce the incremental role of the image. Complex captions, however, may pair surprisingly well with complex images when the image helps users make sense of the post.

Kanuri also emphasized that the evidence is strong but should continue to be confirmed across platforms. The basic idea - that attention is scarce and visual cues can help win it - is likely to travel. The exact size and shape of the effect may depend on the platform.

A research agenda beyond static images

The project also points to many future questions. How would the effect change in video, where color, motion, sound, text, and pacing interact dynamically? Can generative AI help firms optimize visual complexity automatically, while still preserving brand identity and avoiding perceived manipulation?

For young researchers, Kanuri sees social media as a rich and durable research domain. Platforms evolve, algorithms change, organic reach declines, and firms continue to spend heavily to win user attention. That creates many open questions about how digital content affects both firm welfare and consumer welfare.

"Social media is not going away anytime soon. One could easily build a career studying social media and how it affects consumers."

- Vamsi Kanuri

Read the paper

Interested in reading all the details about color complexity and social media engagement? Read the full paper here.

Want to cite the paper?

Kanuri, V. K., Hughes, C., & Hodges, B. T. (2024). Standing out from the crowd: When and why color complexity in social media images increases user engagement. International Journal of Research in Marketing, 41(2), 174–193. https://doi.org/10.1016/j.ijresmar.2023.08.007

Meet Vamsi Kanuri

Viola D. Hank Associate Professor of Marketing; Concurrent Associate Professor of IT, Analytics, and Operations; Faculty Director of the Mendoza Research Honors Program in Marketing

What drives you to do the research that you do?

“I have always been interested in the effect of content on engagement on social media. Social media is such a rich domain because firms spend a lot of time and money there, and users spend a large share of their daily attention on these platforms. For me, the key question is how content can grab attention and help users learn more about brands, while also opening up important questions about algorithms, platform design, and consumer well-being.”

If you were not in academia, what would you be?

“I would probably be an entrepreneur. Research gives me intellectual freedom, and entrepreneurship seems to offer a similar kind of freedom to pursue my own interests. I also enjoy reading and writing, so something related to writing, maybe even writing movie scripts, would also be appealing.”

The article was written by

Ali Boluki

Ph.D. candidate at the University of Groningen (The Netherlands)