With the global textile industry facing mounting scrutiny from consumers and regulators over the environmental and ethical impacts of fast fashion and influencer culture, two researchers from Constructor University’s Emerging Technologies in Industrial Engineering (EITIE) Group have published a framework to help responsible brands engage more effectively with audiences about sustainability on social media. The recently published study in IEEE Transactions on Engineering Management, applied causal machine learning to dig below the surface-level engagement metrics of ‘likes and shares’ to identify which types of content actually cause organic engagement and reach on social media. The resulting framework offers brands and companies clear, evidence-based guidance for maximizing the impact of their sustainability initiatives on social media, in an increasingly saturated and skeptical digital landscape.
“Social media plays a major role in shaping people’s behavior, values, and everyday decision making, which makes it a powerful yet underutilized tool for driving sustainability,” said Omaymah Al-Mashaleh, a PhD researcher and Research Associate behind the project. “Our motivations for this research stemmed from the need to help decision-makers and businesses use digital platforms, not only to achieve economic goals but also to strengthen their environmental responsibility, actively support sustainable and circular sectors, and benefit from new aspects of business intelligence.”
Using an approach known as double machine learning – a method used to distinguish correlation from cause and effect – the study analyzed real-world sustainability content across social media platforms and formats, including short-form Reels, long-form video and carousel-style photo albums. This causal approach allowed the researchers to ‘thread the needle’ and isolate which content characteristics genuinely drove audience engagement, rather than simply coinciding with it. The result was a more rigorous understanding of how content relating to sustainability is received, interpreted and ultimately acted upon by social media users.
The analysis found short-form Reels generated substantially higher public engagement with sustainability messaging than longer-form videos or photo carousels. Based on these findings, the framework recommends “structuring campaigns as two-stage funnels,” using Reels during early awareness phases to capture attention, while directing longer-form video and carousels towards education-phase content. This approach can practically and empirically enable brands to increase organic reach while keeping marketing costs low.
For Dr. Omid Fatahi Valilai, Professor of Industrial Engineering and head of the EITIE group, the implications extend beyond communication strategy and have the potential to fuel measurable societal and environmental change. “We pursued this topic to bridge rigorous causal analytics with real infrastructures, so that sustainability communication not only informs consumers but also helps activate the systemic change our textile ecosystem needs.”
That systemic change includes incoming regulatory measures like the European Union’s move towards mandatory textile waste collection, aimed at curbing the more than 12 million tons of textile waste it generates each year. Professor Valilai is also involved in the EU-funded Sort4CIRC project, which is developing advanced systems for tracking and sorting textiles through Digital Product Passports (DPPs). “This study matters now because it causally shows how to reach people with credible sustainability messages,” said Dr. Valilai. “The immediate application is clear: brands and public agencies can reallocate content to amplify impact and then connect engaged audiences to product-level transparency around material, care, repair and take-back schemes made possible by DPPs. Downstream, this can accelerate data sharing, improve sorting and recycling decisions, and drive measurable progress toward circularity.”
The full published study is available open access here: Causal Drivers of Sustainable Social Media Engagement in the Textile Industry: A Double Machine Learning Approach | IEEE Journals & Magazine | IEEE Xplore