New Research Explores How Streaming Platforms Influence Consumers

Businesses have long leveraged recommendation systems to influence consumer decisions in digital marketplaces. Companies like Netflix, Amazon, and TikTok use algorithms to gather information about their customers and then personalize the recommended content, products, or services to fit their interests. Making successful recommendations can increase a customer’s loyalty, trust, and satisfaction with that brand and increase the company’s profits.

Nasim Mousavi

A recent study by Nasim Mousavi, assistant professor of computer information systems at Georgia State’s Robinson College of Business, and two co-researchers, examines how companies can use a specific behavioral strategy—the decoy effect—in different personalized and non-personalized recommendation environments.

The study, published in Information Systems Research, offers critical strategic insights for managers and digital marketers aiming to optimize the effectiveness of their recommendation engines.

“I became very interested in designing highly beneficial recommendation systems that could lead to more satisfied users,” Mousavi said.
The decoy effect is an established behavioral phenomenon in which a decoy item, typically presented to be easily comparable to a target item in a set of options, is used to create a clear contrast with the company’s target product. Using a decoy, in many cases, simplifies the consumer’s decision-making process and increases their satisfaction and confidence with their choice.

When customers encounter several potential items to select from, they must compare them across various dimensions and values. When there are no apparent contrasts or clearly superior items, customers may experience decision paralysis.

“Tik Tok shop, for example,” Mousavi said, “uses the contrast effect and the decoy effect by showing a cheap item alongside a medium-priced item and a more expensive item. The more expensive item can act as a decoy here, and usually, it would make the medium price item appear more attractive, which would lead users to purchase the medium price item more frequently.”

This decoy tactic has long been used in traditional product marketing to nudge consumer behavior. The new study suggests, however, that using the decoy effect in some settings may backfire and signal that the company does not understand the consumers’ interests.
“Recommendation systems aim to make online shopping more pleasant for consumers by recommending items that fit their interests and reducing search costs,” Mousavi and her co-researchers state in the study. “This objective cannot be achieved without understanding the factors that impact consumer behavior in different circumstances.”

The study examines how the decoy effect functions in two distinct recommendation settings:

• Non-personalized recommendations: product suggestions are based on overall popularity and aggregated customer ratings, with all users seeing the same list of options.
• Personalized recommendations: product suggestions are tailored to individual users based on their previous preferences, behaviors, and interests.

The study reveals that customers show experience a critical distinction between these two settings:

• In non-personalized recommendations, using a decoy successfully drives attention and purchases toward the target product, confirming the typical efficacy of the decoy effect.
• In personalized recommendations, however, the decoy effect backfires. Users expect personalized recommendations to closely align their tastes. Encountering a decoy that does not reflect their tastes reduces confidence in the recommendation system’s reliability, ultimately decreasing purchase likelihood and resulting in lower levels of loyalty for the company.

It is essential for online platforms to understand the contextual factors that determine the effectiveness of these systems and consumer behavior. Recommendation engines account for over $1 billion in annual revenue for platforms like Netflix. The study states that using decoys in personalized settings could decrease the platform’s revenue by more than $110 million. Their findings also suggest that effectively using decoys can increase the likelihood of selecting a target item, resulting in increased sales and improved profitability.

Digital platforms have become increasingly reliant on recommendation systems to drive revenue. Managers can maximize sales and customer loyalty by aligning recommendation strategies with consumer expectations and gaining a key competitive advantage in today’s data-driven environment. This study will also be valuable for scholars developing new recommendation systems based on large language models, enabling users to interact with these systems and receive recommendations closely aligned with their interests.