Want more Instagram likes? Marketing scholars have discovered a formula for the perfect image
To pull in Instagram likes, the biggest rules are obvious: Be a name brand, or just have a ton of followers—and include witty accompanying text.
But a group of researchers at North Carolina State University wondered what would happen if they ignored follower-count sizes and clever captions and just examined the images by themselves: Could users attract more likes by choosing to upload certain types of pictures?
To try to answer that question, they wrote a computer program, scanned 147,963 Instagram images, generated scores based on 6 different visual qualities, which previous studies have shown lead to high user engagement, and then accounted for confounding variables (number of followers and the like). Their findings were published today in a paper that could double as a primer for the visual traits images need to optimize their Instagram likes.
The results amount to one of the most meticulous combings-over of Instagram images ever conducted. The paper’s title is “Simplicity is not key: Understanding firm-generated social media images and consumer liking.” Ultimately, the researchers settled on two rules as a sort of chief yin-and-yang pair: Mix up the balance of light and color, Goldilocks style (not too little, not too much), and ensure the image is either extremely simple or very complex.
These are the 6 specific qualities that the paper (whose full text is paywalled) breaks down:
Luckily, the authors write, Instagram’s AR designers already do the hard part of this job for you: “In practical terms, we found that you could improve the number of likes of any given image by about 3% if you applied the appropriate filter. . . . What’s more, our model suggests that optimizing both feature and design complexity could improve consumer engagement by about 19%.”
Their goal was to create a program and data set that could be used “to inform decisions made by design professionals in the marketing sector,” but the team adds that they’re also releasing the raw code for the program itself. “It’s not in a user-friendly format right now,” admits Bill Rand, executive director of the school’s Business Analytics Initiative and a coauthor of the study. “But I’m sure the right tech-savvy people could use it to create a valuable tool for the industry.”
(31)