According to a report by eMarketer, U.S. digital ad spending will grow by almost 50% over the next four years, and U.S. revenue for ad-supported streaming platforms will triple between 2020 and 2026 to $31 billion as per Research and Markets. With the latest dynamic ad insertion (DAI) technology service providers can deliver ads to viewers based on criteria such as age, gender, geographic location, and socioeconomic status.
The benefits of targeted advertising are multidimensional. When viewers receive ads that are relevant and useful it enhances the streaming experience. Advertisers see a significantly better return on their ad spend dollars. Moreover, service providers can command better CPMs for ad inventory and improve the streaming experience, which increases subscriber viewing times and ultimately results in more ad inventory and additional ad dollars.
Exploring the Different DAI Architectures
Service providers can deliver targeted ads one of two ways. Client-side ad insertion (CSAI) is the traditional method. With CSAI, service providers use an SDK plug-in on the client to play ads at the appropriate times. Server-side ad insertion (SSAI) is the new standard. With SSAI, service providers can customize the video experience for each viewer by modifying the video manifest.
Both architectures must overcome challenges to create an immersive viewer experience. Timing is a key challenge for live and linear content. For live and linear use cases, service providers only have a few seconds to detect the ad breaks, decide which ads to insert, encode and ingest the ads and then perform the necessary stitching. Another challenge is to ensure a smooth transition between the content and ad. There cannot be any glitches or changes in quality detectable by viewers.
Frame accuracy is one of the more difficult challenges to overcome. When the frames are wrong, research shows that it negatively impacts the user experience and destroys the viewer recognition of the ad itself, rendering that impression useless from the advertiser perspective.
Finally, the ad must be the same quality as the content. This requires encoding the ad using the same settings and preferably the same platform as what was used on the content.
Why SSAI Is the Gold Standard
The CSAI architecture is fairly scalable. However, this approach has limitations that are hard to overcome, and that is why most dynamic advertising has moved to the SSAI method.
The SDK for CSAI must be integrated with every client playing the targeted ad. Given today’s rich ecosystem of connected TVs, streaming devices, and dongles, the integration process can be complex. Verifying that the targeted ad works across all devices is also challenging.
Moreover, ensuring a superior quality of experience is hard to do with the CSAI model. Service providers rely on the ad server to provide assets compatible with the video being played. Sometimes, the aspect ratio and video quality of the ads is different compared with the media content viewers are consuming. Finally, service providers need to contend with ad blockers. About 27% of internet users use an ad blocker in the U.S.
The SSAI approach solves many of these issues. The SSAI approach modifies the content manifest in a standard compliant way before final stream is delivered. Ads are stitched into the content stream and delivered to viewers from the same CDN. The ads themselves go through same preparation and quality assurance process to assure there is no content quality disparity. Since the stitching happens on the server side, transition glitches are nonexistent.
With proper content preparation, it is relatively easy to ensure perfect frame accuracy for each ad insertion. Because the manifest modification is done in a standard compliant way, each insertion can be delivered to virtually any client. The only integration impression reporting, which uses simple HTTP interactions supported on all connected devices.
Furthermore, SSAI processes both content and ads via the same workflow, assuring consistent quality. Since all content, including the targeted ads, is delivered through the CDN, ad blocking is virtually impossible. One remaining challenge for service providers is how to effectively scale the targeted ad architecture to thousands or millions of concurrent viewers, which is critical for high-value content like live sports.
Scaling Effectively With Cloud
By adopting a cloud-native video streaming workflow running in the public cloud, service providers can effectively scale targeted advertising during live events. The cloud allows service providers to utilize the instant elasticity of compute power across their video streaming workflow. Elasticity enables service providers to scale to meet the demands of millions of concurrent viewers, each of whom are receiving ads targeted based on a specific set of demographics.
Having a cloud-based video streaming workflow simplifies integration within an otherwise complex ecosystem. The entire ad tech stack lives in the cloud, and many of the targeted advertising components (i.e., ad servers, user data platforms) are cloud-native. When everything exists on the cloud, targeted ad components can communicate flawlessly with parts of the video streaming workflow.
SSAI takes advantage of the inherent flexibility of IP video delivery, allowing service providers to continue using content delivery networks to cache and serve video assets in the most efficient manner while relying on manifest manipulation technology in the cloud to scale and deliver custom manifests to each viewer in real time.
Bally Sports is an example of a media company successfully delivering targeted advertising for live sports programming at scale. Leveraging DAI, SSAI and cloud technology, Bally Sports is serving upwards of 230 million manifests per day.
Monetizing Targeted Ads
Targeted ads can be used across a variety of video streaming use cases, including linear, live, VOD and time-shifted TV. The easiest way to get started is by deploying preroll ads, which play directly before the content that is being viewed. No content preparation is required. Service providers only need to modify and prepend the manifest with the chosen ads before they are delivered to each viewer.
Mid-roll or interstitial ads are the most effective type of ad for monetizing content. They can easily be paired with pre-roll ads. This model uses info about the content, such as chapter breaks, to determine the least disruptive point to place ads. ML-assisted tools make this a less tedious task than it has been in the past.
To successfully deliver dynamic targeted ads, service providers need a flexible, scalable and reliable infrastructure. The cloud allows service providers to scale to meet the demands of millions of concurrent streams. Combined, DAI, SSAI and the cloud empower service providers to address all types of video consumption patterns to maximize ad inventory and reach. Harmonic is leading the charge in helping service providers deliver targeted ads at scale with its VOS®360 cloud SaaS platform featuring SSAI and DAI technologies.
[Editor’s note: The is a contributed article from Harmonic. Streaming Media accepts vendor bylines based solely on their value to our readers.]
Targeting streaming viewers is becoming harder. Today, services are looking at understanding what viewers will buy, not their personal information. Here’s a roundup of a few recently published reports.
17 Oct 2022
OTT advertising technology is at a crossroads, as brands work to build relationships with consumers without alienating them. Here’s a look at some of biggest questions facing the ad tech industry, along with some possible answers.
01 Jul 2021
Harmonic Director, Playout Solutions Andy Warman explains the basics and benefits of server-side ad insertion in this clip from his presentation at Streaming Media East 2019.
16 Oct 2019
Companies and Suppliers Mentioned