Make your media supply chain work for you

More consumers in the U.S. now have a video streaming subscription compared to a traditional cable subscription. A report by Juniper Research estimated that in 2020 each U.S. household had an average of four streaming subscriptions. In the race to capture streaming subscriptions and satisfy a growing appetite for content, investment in original over-the-top (OTT) content has seen an exponential rise.

To keep up with OTT platforms, traditional media businesses have followed suit by increasing their focus on original content. For many of them, the ability to increase investment in original content equates to long-term survival. Once produced, this content needs to be delivered and monetized on multiple consumption platforms. Not so long back, media businesses delivered thousands of assets a year to a handful of linear playout end points. Today, they need to distribute tens of thousands of content packages every year to hundreds of endpoints.

To handle this, media businesses need to evolve their supply chains for scale and efficiency. For most businesses, a broad review of the current state of the supply chain reveals significant opportunities across all stages of the content chain. Often, there is a suggestion to consider an integrated platform to solve these problems. The argument for this gets stronger in the face of new third-party platforms that can manage and track both content and metadata through the entire supply chain. Several of these platforms are constructed as SaaS offerings on the cloud with flexibility for integration with existing systems. However, implementing a new platform entails a high change burden, significant investments, and dependency on an outside entity for a technology asset that is so pervasive in the business. Instead, enterprises will benefit by taking a more compartmentalized approach.

How far to go with metadata?

While there exists an argument to apply traditional master-data management principles to both descriptive and technical video metadata, a single consolidated metadata repository is challenging to create.  Metadata evolves through the content lifecycle and needs to be editable in systems supporting key functions.  Building an aggregate view of metadata that integrates into several source systems is a valuable alternative for a fraction of the cost.

In addition to being a metadata aggregator, there is a benefit in housing catalog title management activities and metadata enrichment workflows in the metadata aggregation system. This means integrating into third-party metadata sources such as Gracenote and EIDR and building enrichment workflows.

One topic of debate is whether to include rights and avails in the metadata. Given the complexity and dynamic nature of rights, it’s best to keep rights in a more authoritative, purpose-built system to accommodate and keep up with market changes. Avails information is not technically metadata and is the result of a point-in-time computation, so there is not a strong case for including avails.

Yet another debate is whether to include mined video metadata in the aggregate view. Due to high compute costs for mining video metadata, there is a strong case to preserve this metadata. However, given the volume and typical use of this metadata, making it machine-accessible is more important than surfacing the same on a human interface. Additionally, video mining is often limited to a subset of the asset catalog, so it may not be widely available.

Aggregating metadata is a key aspect of media supply chain modernization, but it requires the right workflow orchestration to make the supply chain run seamlessly.

Is cloud workflow orchestration really the future?

Conceptually, both digital and linear supply chains have identical stages: concept, production, post-production, distribution, and consumption. There is also a significant overlap in processing needs between digital and linear. However, in most media businesses, digital workflows have organically grown separate from their linear counterparts, primarily due to post-production treatment and distribution differences.

In the last few years, many media businesses have significantly improved their digital workflows moving from spreadsheet-based tracking and manual processes to automated workflows, standardized transcoding solutions, and automated metadata transformation. Unfortunately, in many cases, businesses are faced with the high cost of operating these supply chains, and often they are not architected for scale. Additionally, collapsing digital and linear workflows where possible has not been a priority.

The future of media workflows is undoubtedly on the cloud using event-driven architectures and serverless processing. An orchestration engine that recognizes new content drops as intake events and routes the content through stages using serverless video processing, metadata retrieval, and packaging provides the basic concept. Cloud-native technologies today allow for stringing processing pipelines using largely configuration. This allows for a lightweight central orchestration entity largely responsible for handling exceptions, recording progress, instrumentation, and analytics. The overall approach plays well with both cost and scale: avoiding idling servers waiting for resource-intensive jobs and using cloud elasticity for dynamic workloads.

Finding the right home for media assets

Closely tied with workflows is storing and managing media assets. Making this decision for archival is relatively easy – unless the business needs offsite tape archives, all archives get digitized over time and moved to the cloud. It is low cost and provides convenient in-frequent access when needed. Plus, cloud providers have the solutions to physically move large archives to cold storage in a convenient manner.

The decision for in-process content is a little more complicated. While 5G ultra-wideband (UWB) and emerging Wi-Fi versions are starting to provide fast gateways into cloud storage, there are high egress costs unless the transformation and customization workflows run on the cloud. Big picture, this is a point-in-time decision for media businesses, given all assets and workflows will eventually need to be migrated to the cloud. There are two reasons for this: remote work becoming an industry and workforce expectation and every software provider selling media assets and workflow solutions moving to software-as-a-service (SaaS) offerings running on the public cloud.

Automation in every stage of the supply chain

With the need to address scale at intake and delivery points in the supply chain, the focus on automation and efficiency at each stage of the supply chain has never been higher. Opportunities span from standardizing partner intake and consolidating to a single gateway all partners use for automated ingestion to instrumenting the pipeline to enable productivity analytics.

On the delivery front, defining machine-readable profiles that capture each target platform’s expectations of each target platform expectations and automating metadata transformation, content packaging, and validation provides significant efficiencies. Automating short-lived access for cloud downloads during the sales process and subsequent cloud distribution can protect content from unauthorized access resulting from human error.

Considering AI but not forgetting the human

With AI becoming mainstream, media supply chains tend to benefit significantly in terms of both automation and innovation. While automated video quality control and captioning solutions have been in use for some time, businesses are exploring AI for automated translation and auto-dubbing. The next frontier of innovation is also going to be rooted in AI. One of the interesting solutions video mining enables is automated video fragment selection from existing content for inclusion in new productions.

However, businesses need to be rigorous in their evaluation of AI, especially as it relates to video analysis, due to the high noise-to-signal issues that often accompany these algorithms. The expectation of human involvement will not go away soon. Businesses should also be prepared to spend considerable effort analyzing video analysis outputs such as automatically extracted metadata to determine optimal ways to surface those on human interfaces or use those to drive automation.

The cloud will keep evolving

Until recently, the cloud has been an infrastructure and development toolset for media businesses. Most forward-looking businesses have moved significant portions of their system and content footprints to the cloud. They are increasingly looking to do away with media workflows that are not on the cloud. Just as cloud providers upped their game from being infrastructure providers to providing native development services, they are now taking the next step in their evolution by investing in industry-specific solution sets.

The more cloud providers can make themselves appealing for production workflows, the more they stand to gain from consumption by downstream supply chain workloads. With that strategy in mind, creative studio solutions are starting to emerge as offerings from cloud providers. While this is still in the early days of adoption, expect these solution sets to get richer over time, along with complementary partner solutions.

Using third-party supply chain platforms

The race is on among third-party software providers to become the media supply chain platform of choice. Traditional media asset management (MAM) solutions are pivoting towards a supply chain platform with an aspiration to have mature end-to-end capabilities. On the other hand, many recent SaaS entrants in the market initially focused on a specific aspect of the supply chain but are now expanding their offerings to cover the breadth of the supply chain.

As mentioned earlier, buying into a third-party platform to manage the entire content supply chain entails cost, change, and risk. A smart approach for media businesses would be to evaluate these platforms to determine which dimension of the problem they are best geared to solve and selectively use those capabilities for cost and time-to-market advantages. While the cloud has made it easier to build versus buy, media businesses have always been big users of off-the-shelf software, and this trend is not likely to change. As these software vendors mature their solutions on the cloud, they will continue to garner serious consideration, albeit for selective aspects of the supply chain.

Summary

Media businesses faced with the market imperative to modernize their supply chain are looking at a multi-dimensional problem and a host of solution options. Decomposing the problem into a smaller set of focus areas helps clarify available solution approaches.

At Virtusa, we have helped multiple media businesses use this approach to assess their media supply chain, clarify priorities, capture the complexity, and build objective modernization plans. Our proven approach has helped clients unify fragmented digital workflows, engineer cost-effective cloud architectures, improve automation, and incorporate AI into their media supply chains.