Live shopping combines socializing, entertainment, and of course, shopping! A typical live shopping event features a well-known social media influencer or celebrity who will host a session where they talk about their favorite products. As they display and discuss each product, viewers have the opportunity to buy and sometimes engage with the session host via chat or call-in. This is one of the most popular shopping options worldwide. In 2023, the estimated market size for livestream e-commerce in the United States alone reached 31.7 billion dollars*

While live shopping is highly lucrative, what makes it difficult to get started, is finding the right video platform.

While live shopping is highly lucrative, what makes it difficult to get started, is finding the right video platform. Options include building and maintaining a platform yourself, or choosing a company that does this for you. In either case, similar issues arise — it’s difficult to set everything up so it’s cost-effective, reliable, and scalable. When users rush to a live shopping session featuring their favorite products and celebrities, the servers must be able to successfully handle every interaction, and continuously deliver multiple high quality, live video streams. Interruptions and pauses can result in loss of sales, hosts, and viewer interest.

Raskenlund builds reliable, scalable platforms that ensure companies can stream multiple high quality, fast live shopping streams every time. It all starts with understanding the live shopping workflow. One common workflow looks something like this (but could include other technologies depending on your desired setup):

1. Publish: The publisher/broadcaster uses a mobile device or web browser to send a live stream to a Media Server like Kurento Media Server, Ant Media Server or Norsk.
2. Invite: The publisher/broadcaster invites everyone who will appear on the live stream to the session associated with the live stream. With Raskenlund’s solution, you can have a composition of multiple separate streams with one host stream and zero or more guest streams.
3. Compose: The media platform merges the streams into a single output stream that the show host can manage using a web-based control panel.
4. Distribute: The composited output stream is sent off for transcoding into adaptive bitrate (ABR) and further distribution.

To manage workflows like this, Raskenlund developed a software program that controls the interaction between all the components of the streams. Some of the tasks it manages include:

• Communicating with the back-end in order to set up the show and user verification
• Providing instructions to the media server or service for how to produce the output stream
• Managing the flow through transcoding and packaging services
• Load-balancing for the multiple instances required to handle multiple live shopping shows occurring at once

. . .

With the control software handling the hard work of creating and delivering consistent, reliable setups for live shopping shows, there’s just one issue left — scalability. Auto-scaling is a part of what cloud service providers offer customers, however the service is focused on handling web requests. It’s fine to move web requests to different servers as part of scaling, but making reassignments like this as part of autoscaling for live streams can cause all kinds of undesirable issues. It can lower video quality or cause interruptions and delays in video streams.

Raskenlund’s auto-scaling solution uses metrics and algorithms that are specific to streaming media

Raskenlund solves the auto-scaling problem with its own auto-scaling solution called Ztream Escalator. Rather than using metrics that are focused on what’s needed to scale web requests, Ztream Escalator uses metrics and algorithms that are specific to streaming media. The product dynamically adapts to sudden changes without causing stream interruptions, anticipates scaling needs, and can be configured in advance to handle predictable surges in livestream viewership. Last but not least, Ztream is fully customizable to each platform. It works with any streaming protocol, provider, or region.

If you’re looking for a consistent, customized live streaming solution, contact Raskenlund to talk about your tech stack today and reap the benefits of:

• Virtually unlimited scaling of streaming resources
• Fully-automated workflow
• Ability to run live streams with guests
• Low-latency streaming for host and guests
• Reliable, high quality distribution to large audiences

. . .

Resources:

• Livestreaming commerce sales in the United States between 2022 and 2026
• Autoscaling: What It Is and Why You Need It
• Why Traditional Autoscaling Is Deficient for Streaming Media