AMTP Proceedings 2026

Document Type

Conference Proceeding

Publication Date

Spring 2026

Abstract

Firms increasingly use visual content on social media to connect with customers, yet most evidence on what drives engagement comes from consumer or personal contexts. This study examines how simple visual features in Instagram photos posted by B2B technology brands relate to engagement, interpreted through processing fluency. The dataset contains 600 photos from 30 B2B technology brands. For each image, we record likes and comments, collect brand and post level metadata, and use Microsoft Azure computer vision tools to detect faces, classify their age group and gender, and identify the dominant background color. Negative binomial models show that the presence of faces does not increase engagement, and images that include at least one face younger than 20 years receive fewer likes. In contrast, darker backgrounds, especially black and brown, are associated with higher engagement, while several bright colors are linked to lower engagement. The findings highlight the importance of visually coherent, product focused imagery in B2B social media communication. Keywords:

Download

Share

COinS