Why platform support can increase sales across the board

By Mimansa Bairathi, former Ph.D. student at London Business School, Xu Zhang, assistant professor of marketing at London Business School and Anja Lambrecht, professor of marketing at London Business School

If you’ve scrolled through Amazon, you may have noticed certain products marked with a badge of honor: “Amazon’s Choice.” This apparent endorsement by Amazon itself – a so-called platform endorsement – ​​is a popular information tool used by digital platforms to recommend items to consumers. It’s a way to cut through the noise; With the rapid growth of digital platforms, consumers have so many options that decision-making can become challenging. However, using referral badges alone to facilitate user selection might underestimate its value. Here’s why.

While it seems intuitive that promoting a select group of items could increase sales of recommended items, little is known about the impact on non-recommended items and on the platform as a whole. This has led retailers to worry about reduced sales of competing, non-recommended items and raised concerns among policymakers about the anti-competitive nature of such practices.

To assess whether these concerns are valid, we analyzed data from a large-scale field trial to determine whether recommendation reduces demand for non-recommended items and what impact it has on overall sales. We uncovered previously unrecognized benefits of platform support that companies could leverage to boost sales across the board.

Data from the field experiment conducted by one of the world’s largest online platforms for freelancers shows that platform advertising, far from cannibalizing sales, increases users’ searches and purchases not only for recommended but also for non-recommended services, which leads to a significant increase in sales at the platform level.

Our results show that leveraging platform support brings great benefits to recommended services, resulting in a 25% increase in clicks and a 41% increase in recommended services revenue. Surprisingly, it also increases searches for and sales of non-recommended services – increasing the average number of impressions by 4.1%, clicks by 2%, and orders by 3.1% per user.

Given the significantly higher number of non-recommended services compared to recommended services, 67% of the increase in orders came from non-recommended services and 33% from recommended services. The research also shows that recommendation increases clicks and orders from non-recommended services that are further away from the recommended service, and that it hurts clicks and orders from non-recommended services closer to the recommended service. We also find that non-recommended services with different prices benefit more from the spillover effects of platform support than services whose price points are similar to those of the recommended services.

A key driver of the positive impact of platform support is consumer perception: consumers believe that the presence of recommended products means that the overall quality of services offered on a platform is higher. Supporting a platform is therefore a useful tool to strengthen consumer engagement and increase sales.

The recommendation is particularly effective when it comes to stimulating more purchases among users with a higher purchasing propensity. A better understanding of which user segments are more responsive to platform support would help platforms derive maximum benefit from platform support. By focusing their advertising efforts primarily on the types of services that are of interest to the most responsive users, rather than dividing their attention evenly across items or categories, companies could help increase engagement and increase purchases.

Our findings have important implications for platforms aiming to increase consumer engagement and platform revenue, as well as for retailers and regulators concerned about the competitive environment and fairness in platform advertising. It is clear from the study that platform support is a useful tool to facilitate user searches and purchases on a platform. It has a positive impact on recommended services, has a positive spillover effect on non-recommended services, and increases overall purchases on the platform. Even small changes in platform design can have a noticeable impact on sales and change users’ perception of the platform as a whole.

Merchants can benefit significantly from increased user searches and purchases of recommended items, but while platform support disproportionately benefits recommended items, this is generally not accompanied by reduced sales of non-recommended items. The ability to use recommendations to increase overall sales, not just recommended items, suggests a win-win scenario for everyone.

Mimansa Bairathi is Assistant Professor of Marketing at the UCL School of Management. Her research focus is on digital platforms and digital economy. Her work examined the impact of marketing interventions on consumer behavior and corporate sales, as well as the regulatory implications of marketing interventions.

Xu Zhang is an assistant professor of marketing at the London Business School. Her research focus is on developing effective marketing strategies for digital platform markets. She has extensive expertise in analyzing how information design and pricing can impact market efficiency, corporate profits, and consumer welfare.

Anja Lambrecht is Professor of Marketing at the London Business School. Her research focuses on digital marketing and the digital economy. Her work examined the use of retargeting to reach consumers, advertising on social media platforms, digital pricing, and how spillover effects can lead to algorithmic bias.