Opinions expressed by Entrepreneur contributors are their own.

Who hasn’t laughed at the expense of a streaming service’s off-base recommendations? Consumers subscribe to streaming platforms for content, but most of what people watch is scattered across several different content providers, leading to multiple subscriptions.

Each streaming platform captures only a fraction of a ‘s total viewing, which causes streaming services to provide personalized recommendations based on too little data. Many new streaming services have such little information that they simply send the same emails and recommendations to everyone, leaving few people happy with the selections.

This is a problem that is only getting worse as consumer viewing habits are shifting from once-dominant providers like Netflix to newer but growing entrants such as +, Paramount+, and Peacock. With growing streams of new content and without fully knowing their customers, streaming platforms make wild recommendations and are untrustworthy tools for finding content, frustrating consumers even more.

Only a consumer’s remote truly knows everything they watch, but one unifier across platforms could provide consumers with a tool to take better advantage of all the available content.

Related: Where Entrepreneurs Can Innovate in the Streaming Service Space

Streaming algorithms work with incomplete data

To guide the viewer toward the content they want to watch, streaming platforms use unique algorithms. Some are based on text and tags, while others filter content by collaborating on similar user interests and behavior. Most platforms use a combination of algorithm styles to generate more accurate recommendations. With viewer preferences divided among so many streaming platforms, no service gets it right. Each service’s algorithm is trying to personalize viewing with an incomplete viewing history, resulting in recommendations that can be wildly off target.

Streaming provides seemingly endless entertainment, but with so much content and no one to guide them through it, viewers are burning out. Without accurate recommendations, people spend too much time browsing for something new to watch. In 2015, Netflix consumer research found that a subscriber loses interest after only 60 to 90 seconds of sifting through recommended titles, a sharp contrast with the more than 20 minutes that many consumers report they are spending trying to decide what to watch. Without any way to input their feedback into these algorithms and improve them, consumers are growing frustrated with streaming as a whole.

Related: The Battle Between TV-Streaming Giants Continues, But Consumers Are the Real Victors

Consumers are feeling neglected

When Netflix only knows what I watch on Netflix and only knows what I watch on Prime Video, neither service understands my complete preferences and can’t make trustworthy content recommendations. Even worse, Netflix never recommends something found on Prime, nor would Prime recommend Netflix original programming.

Streaming device makers are no better. By now, consumers expect TV to steer them towards Apple TV+ shows; Amazon’s Firestick to Prime; to the Roku Channel. This combination of incomplete information and structural bias is built right into today’s streaming landscape and the problem will only grow worse as consumer viewing hours continue to fragment.

Consumers desire better streaming recommendations and are eager to share their viewing habits to receive good recommendations. In an Accenture survey, 56% of respondents said they wanted a viewing profile that can take them from one service to another in order to better personalize their content. Without a way to input feedback into streaming services, consumers feel they are spending too much money to keep up with too many services.

The same survey reported that 63% of respondents felt it was too expensive to subscribe to multiple providers in order to consume all their desired content, and one in three respondents said they planned to “somewhat” or “greatly” decrease their entertainment subscriptions in the upcoming year. The more services consumers use, the more frustrated they appear to get with streaming.

Related: 10 Tips and Tricks to Make the Most Out of Streaming Platforms

How to get unbiased personalization across platforms

Nearly seven out of 10 survey respondents described cross-service search engines as a “good way to find desired content” and more than half felt it was more convenient than going directly to the provider. By reaching across platforms, companies that combine user preferences can provide insight into a complete consumer taste profile and can provide better, more personalized recommendations than any one service can do on its own. Aggregators can do this without any of the inherent biases and conflicts that are built into the fabric of individual streaming services.

Instead of expecting providers to rebuild the streaming wheel and start over, the industry needs to add an additional layer on top. A few decades ago, cable provided a layer between consumers and the multitude of new choice that was emerging. Cable packaged up programming bundles and made the new simpler. Now, aggregators can (and will) step in to provide the simplified and personalized hub needed to enhance the streaming experience. These hubs will not only save consumers time and but will make streaming simple, personal and user-friendly. These hubs will give consumers a wealth of compelling new content at their fingertips.



This story originally Appeared on Entrepreneur.com