Industry perspectives: a multimedia archivist on the challenges of monetizing visual assets

Media companies are grappling with digital transformation projects. 

• The lure: monetizing vast media archives
• The urgency: analog tools and services are being sunsetted, while digital content libraries are set to explode, fueled by AI content production. Without efficient content discovery, asset storage will remain loss-making.
• The messy middle ground: studios are relying heavily on archivists to bridge critical data gaps

We recently spoke with an experienced multimedia archivist working with studios across the US. Her quotes are presented without attribution, to protect sensitive business practices. This piece explores the challenges of monetizing media archives, and how multimodal AI can improve business outcomes.

The scale of the challenge

According to our expert, studio archive teams everywhere are struggling to efficiently monetize assets. They’re sitting on potential goldmines, but asset retrieval is hindered by outdated systems, bad metadata, and poor quality search tools. 

As Heidi Shakespeare, CEO of Memnon Archiving Services, points out: "Millions of audio and video tapes (both analog and digital) still haven't been migrated and hold invaluable, irreplaceable content. [But migration] demand has increased exponentially - in the past 24 months Memnon has provided quotes to migrate over 7 million hours of content, or 70,000 hours of content a month for the next 10 years."

For legacy media and entertainment businesses sitting on those kinds of assets, the scale of the opportunity is also the scale of the challenge. And the status quo is a costly space to play in.

Current workflow: a complex web

Our expert explained that every studio’s legacy archiving style is unique. The work often involves cross-referencing analog and digital assets, and relies heavily on searching systems that weren’t built for multimedia inquiries or nuanced requests.

The current process for monetizing media assets involves multiple steps and challenges:

1. Client approaches studio with a creative need
2. Studio estimates costs based on perceived availability
3. Studio submits request to archive team
4. Media archivist begins extensive search
5. Archivists submit closest matches, usually approximate

In other words, media archivists are trying to meet 21st-century expectations using 20th-century tools. "Sometimes the degree of specificity is impossible to service [with traditional methods]. For example, a client recently asked me for 'someone on a ski lift wearing their [brand's] merchandise'. I was able to provide something close, but not the exact match. Archive sales currently involve a lot of compromise."
– Multimedia archivist in conversation with Coactive AI

‍

Key challenges in archive monetization

1. Mixed format archives and legacy indexing software

Many media archives, especially for smaller studios, blend digital and analog assets. These are often searched and retrieved through the same archaic content indexing software, leading to complications in workflows.

"These systems were designed for one type of artifact, but their use has swollen over the years to include derivatives and other details required for identifying or reproducing source masters" – Multimedia archivist in conversation with Coactive AI

For instance, a media archivist may be tasked with finding the highest quality version of an asset named “Live_tape_84_centre”. But original metadata clerk had to make a choice: how much space within the limited metadata taxonomy should they devote to physical descriptors of the analog asset (e.g. properties of the reel), versus populating those fields with descriptions of the actual visual content itself (e.g. the people or products featured, the locations, the events captured)?

These historic trade-offs haunt media and entertainment businesses today:

• Inconsistent metadata across formats
• Difficulty in identifying highest quality versions
• Challenges in future-proofing archives

As our industry expert explains, studios are pondering, "What do we save now? What do we migrate so that it's ready for use cases 20 years from now [that] we can't even imagine?"

2. Impossible analytics

Content performance analytics represents a valuable opportunity for archive owners. Imagine if they could provide data-driven, actionable insights for a range of stakeholders – from content creators, to advertising clients, to media storage managers. 

The current state of archive metadata is the barrier.

"That level of analytics is simply not possible within our current systems, because of the data paucity. Metadata enrichment is SO manual - with the volume required it's completely untenable to scale it up, so analytics are off the menu."

The solution: multimodal AI

The advent of multimodal AI presents a paradigm shift in addressing these challenges. AI-powered metadata enrichment and multimodal search capabilities can accelerate content discovery, improving both the quality and quantity of licensable assets.

Key features of the Coactive Multimodal Application Platform:

1. Content enrichment: create customized metadata using multimodal inputs
2. Content discovery: search across images, videos, and audio using a combination of text, image, and keyword prompts
3. Content analytics: leverage powerful AI to extract insights from visual datasets using natural language prompts, SQL, and data visualizations
4. Integration: APIs and SDKs to integrate with existing Digital Asset Management (DAM) systems

Use Cases

1. Searching for broad moods or concepts

Content creators typically have more open-ended briefs, our expert explains. Those clients want to tease out options, and the search is exploratory. For instance, they might brief her to “find something romantic but thrilling, like a car chase”. With conventional search tools, cross-referencing broad parameters like moods and details is extremely difficult. 

The Coactive platform solves this by allowing users to quickly teach the search tools what is needed - overcoming the lack of historic metadata. For instance, using the Intelligent Search feature, users can drop in an image or video clip (e.g. two lovers in a car chase) and give simple yes-no clicks to train the model in a couple of minutes. Now, they can enter a text prompt to search the archive as if they were speaking to a person – and uncover those niche assets the client needs. 

2. Finding niche examples

According to our expert, marketing clients often have very specific requests. They’ve typically seen something externally that they want to emulate, or have an existing asset in-house but want different shots of it – like their merchandise being worn by a specific celebrity, on a specific location. E.g. “Charlie XCX wearing a Patagonia sweater while riding a scooter.”

Coactive’s multimodal AI search blends specificity with efficiency. No more manual searching – now archivists can leverage image-to-image and natural language (semantic) text prompts to pin-point the exact assets desired. All done in seconds, not hours.

Summary

Media companies can help their archivists be far more productive by providing them with AI-powered tools to enable content discovery. By deploying multimodal AI, companies can:

• Overcome legacy metadata problems
• Discover assets and analyze trends at the speed of AI
• Streamline processes and focus on value-added tasks

With the Coactive Multimodal Application Platform, media companies can unlock the full potential of their content libraries, turning historical assets into future revenue streams.

Ready for a demo? Get in touch today.

‍