Backblaze’s Dave Simon on the cloud storage evolution in media production

Digital storage capacity for non-archival applications in media and entertainment is projected to skyrocket from 24 exabytes in 2019 to 122 exabytes in 2025, an indication of just how fast the industry’s data needs are expanding.

This surge touches every corner of the business, from sporting events requiring instant replay clips to corporate video production teams capturing hours of footage for online campaigns. Organizations that once relied heavily on physical tape libraries and “heavy iron” on-premises systems are rethinking how they store and access these growing volumes of content.

Today, everyone is a content creator

Sports leagues, corporations, faith-based groups and even niche publishers have all evolved into full-fledged content creators. The shift is partly driven by new distribution channels — online streaming, social media, mobile apps — and partly by heightened viewer expectations for on-demand, high-resolution media.

This environment has made flexibility paramount.

“We’re seeing a move to hybrid, to cloud-native workflows,” said Dave Simon, senior director of media and entertainment alliances at Backblaze. “Everyone is transitioning from fixed, on-prem systems to agile, consumption-based models in the cloud.”

The appeal of cloud storage has always included scalability, global accessibility and relief from the hefty capital expenditures required to maintain large on-prem archives. However, storing media offsite opens a new set of considerations — from retrieval fees to unpredictable monthly bills.

Trends and challenges in the shift to the cloud

Global collaboration: Production teams are increasingly scattered across regions, especially in an era of remote and hybrid work. Fast, secure access to shared assets is essential for editing, color correction, and quick turnaround on timely content, such as sports highlights.

AI and machine learning: Tools for automated transcription, object detection, and content recommendation require enormous data sets, making cloud storage an attractive solution — if the economics pan out. “AI is consumption-based,” Simon said. “The more you process, the more you pay. That’s why predictable storage costs are crucial.”

Advertisement

Audience demand for speed: From near-real-time social media updates during a sporting event to last-minute streaming distribution, media workflows must be nimble. Delays in retrieving footage can thwart monetization opportunities.

Despite the clear advantages, the move to the cloud has been far from frictionless.

Many companies quickly realized that after migrating petabytes of data to a major cloud platform, monthly costs could climb to surprising levels.

“We saw a rush to migrate everything, and then the bills rolled in,” Simon said. “People started asking if they’d be paying these hyperscaler fees forever.”

A large portion of these costs stems from what’s known as “egress fees” — charges for moving data out of cloud storage. Some cloud providers also place media files in lower-cost storage “tiers” that can require extended waiting times or premium payments for quicker access. For industries dealing with large files and tight deadlines, those hidden or unpredictable fees are concerning.

Moreover, organizations accustomed to on-premises tape libraries still face the logistical challenges of retrieval. Although tape offers cost-efficient storage per terabyte, it’s vulnerable to physical damage, off-site transportation issues or internal mismanagement.

“I’ve had clients lose data due to a simple air-conditioning leak that flooded an entire tape archive,” Simon said.

Diversified’s Eric Hutto on broadcast integration’s IT evolution

Pivot toward hybrid

As a result, a hybrid approach — storing essential data in the cloud and maintaining certain archives or specialized tasks on-site — has emerged as a middle ground. It allows companies to gradually move off tapes or legacy servers without fully abandoning their existing infrastructure.

“After initial excitement about the cloud, many are asking if they can keep some workflows on-prem,” Simon said. “The reality is that each environment, on-prem, cloud, multi-cloud, has pros and cons.”

Simon noted this balanced perspective is becoming the norm, not the exception.

“We see large networks, sports broadcasters, and corporate media teams start in the cloud for short-term projects or to spin up an environment quickly,” he said. “Eventually, they explore a deeper move once they’ve proven it out financially.”

In this landscape, various cloud providers — and newcomers — offer storage solutions with different cost structures and performance metrics.

Advertisement

Backblaze has zeroed in on predictable pricing and immediate data access (sometimes called “hot storage”). The goal is to reduce friction for media workflows that need large, scalable repositories without incurring high retrieval charges.

“That’s where we fit,” Simon said. “We’re about one-fifth the cost of AWS, with no penalty for egress, and data is always there. For media, it’s crucial not to wait hours for files to become accessible.”

Future of MAM: Interoperability, open standards and unified workflow

Though the company focuses primarily on storage, it also integrates with a range of media asset management tools, production platforms and other cloud-compute environments.

“We have an open-cloud philosophy,” Simon said. “We partner with best-of-breed solution providers so that you’re not stuck in a walled garden.”

Keeping an eye on the future of media

Generative AI has quickly become one of the major conversation topics in media storage. Storing and processing large data sets for training and inference can be prohibitively expensive on conventional cloud platforms.

“We’ve seen huge interest from AI startups,” Simon said. “They need petabyte-scale storage that won’t break the bank as they iterate on models and grow their data pipelines.”

By pairing cost-effective storage with flexible integration points for AI workflows — such as third-party AI orchestration frameworks — media organizations and startups alike can rapidly scale up their machine learning projects.

From a broader perspective, the current state of media storage is one of flux and experimentation.

Producers want to leverage the cloud for its scalability and global reach, but they also want predictable and transparent billing. They need near-real-time collaboration tools but often require some on-prem security or archiving measures.

“People are mixing and matching based on their unique workloads,” Simon said. “It’s not a one-size-fits-all industry.”

Advertisement

For those navigating these decisions, ensuring the right balance between cost, performance and accessibility is paramount. While some may opt for purely on-prem systems or fully embrace a single cloud vendor, many organizations are exploring a hybrid or multi-cloud strategy, making partnerships and interoperability more critical than ever.

“It’s essential to stay flexible,” Simon said. “Cloud storage doesn’t have to be expensive or complicated — if you’re strategic about how you set up your workflows. Ultimately, no one wants to compromise on creativity or content quality just to manage a storage bill.”