Cloud Business Models

 
Now that the march of technology has enabled sophisticated media production and distribution via the cloud, and the ongoing COVID pandemic has accelerated an already pressing need to produce content remotely, the logical next question the industry needs to address is, what’s the best way to reimagine business models so that cloud-based media production becomes more practical, affordable, and engrained within media institutions? At least, that’s the view of Larissa Görner, director of Cloud Product Management for Grass Valley and co-chair of the SMPTE Media in the Cloud Advisory Board.  

Görner points out that in the last year-plus, there has been “a big run to the cloud” by media companies, “because they had to survive. To stay in business and continue producing content with people at home, the cloud was a natural [tool] to move toward. But the most important thing to remember is, why is it so important to do it [beyond the COVID crisis]? It’s for the same reason good business people do anything—to create value, to grow their business. Therefore, you have to look at cloud-based media from multiple levels. If you don’t look at it from an organizational perspective, you probably won’t succeed with it in the long run.” 
 
As discussed in Newswatch earlier this year, the media industry, even before the pandemic struck, had a wide range of initiatives underway designed to point the industry in the direction of a cloud/software-based production revolution in the near future, most especially the detailed MovieLabs white paper that was published in 2020.

But Görner emphasizes that there are a range of subtler cultural and transitional issues surrounding how media companies large and small can possibly get there. As a consequence, she suggests, media companies need to seriously examine ways to bring what Görner calls “their separate silos, or small islands within bigger media companies” together under the umbrella of a cloud-based media strategy.  
 
“What we are seeing is that big media companies are now looking really hard at what their cloud strategies should be across their entire chain, to bring a consolidated approach so that their entire company can move into the cloud seamlessly,” she explains. “A transition to the cloud is most cost effective when everyone begins to work on the same platform. You have to examine these existing silos—playout, post-production, live production, archive, asset management, and so on—and come up with a single strategy that makes sense for all of them to be in the cloud.” 

 
“After you understand what makes sense financially and technically generally, then look at how to leverage this strategy across multiple productions and see how you can grow it,” she adds. “The big thing that every media company is facing in terms of best practices is getting people from all the different silos together—the people from archive, storage, playout, live production, and so forth. We need to get them all sitting together to understand what efficiencies can actually be created within a common cloud environment, keeping in mind that, in the end, the goal is always a financial one, to create more content with the available funding that they have. 
 

“Then, each silo can accelerate the process and bring more of their assets into the cloud, to make these efficiencies work. And then, instead of silos, they all become a chain. Playout is not a silo anymore—it is integrated into your production flow. Master control becomes a certain step in the process, a feature of your whole platform, instead of a separate department in your company.” 
 
One particular silo, of course, is more complicated than others in terms of converting to cloud-based methodologies—live production. Görner points out that live production in the field for certain kinds of programming like sports, news, and certain shows can be harder to efficiently execute due to challenges like bandwidth limitations, acceptable latency, and more generally calculating when it is worth it, and when it isn’t.  
 
“There has always been this magic triangle for live production—parameters that can have quite an impact,” she explains. “So, the triangle of cost, quality/bandwidth, and latency is a really big thing. Obviously, you can only use the bandwidth that is available, and that is largely where the big cost lies, so you need a workflow that is cost-efficient but might produce a little latency. You can’t beat the light, after all—your signal has to take time to travel wherever it is going. So, you need a production chain that synchronizes the media within the whole chain, no matter where the media is, getting everybody globally working in the same time window. Then, you can make your operation work, and once you have a smooth operation, you can be more creative, and when you are more creative, you can increase your quality.” 
 
This essentially means that, for certain types of live productions, at least, a certain degree of latency has to be considered acceptable on this new landscape. Görner says the problem comes in figuring out what “low latency” or “acceptable latency” is for particular operations and under particular conditions. “For example, five milliseconds would be considered really low latency if you are taking about audio,” she says. “But [picture] might take 50 milliseconds or above. So, everything below 50 milliseconds would be considered low latency in that sense. But if you bring really high-quality media into the count, processing that media might take 500 milliseconds. Everyone on the chain has to get used to managing certain ‘delays’ when working on a global scale. The important thing is to adapt the workflow so that the delays don’t create disruptions for the operator.” 
 
And then, she adds, “bandwidth is where the big cost lies. You might have [the advantage of having] big, dark fiber to move uncompressed media to the cloud, but even then, it still has to be processed. And right now, processing all that data in the cloud is not always financially viable for many companies.” 
 
And so, for all these reasons, Görner continues, productions “will need to compress the signal. You will want to anchor the media, which might add more latency. But that is currently the tradeoff needed to work globally, whether from the field or home or the studio.” 
 
Of course, much content is not live, and is being captured in extremely high resolutions like 4K or even 8K. Görner explains that many creators of high-end content are moving toward a private cloud model, when possible, that permits them to work on media uncompressed locally. 
 
“In those cases, there are certain workflows they would use to leverage their content into the cloud,” she says. “They are seeking connectivity between their facility and the cloud or a cloud provider. For them, we are seeing, over time, the processing capacity of servers really increasing. Just two years ago, you had to compress such media hard. But today, we are starting to see technology for running high-quality content into the cloud. At Grass Valley, that is the point of our flagship K-Frame on AMPP production switcher, which can bring up to 40 streams in without making a big tradeoff on operation, quality or latency for live productions, for example.” 
 
Large, sophisticated media organizations are largely “pivoting their IT infrastructures,” according to Görner, to address these kinds of issues and build, over time, state-of-the-art IP-based platforms for whatever type of content they are producing. However, smaller broadcast entities have financial and logistical limitations on how far in that direction they can travel, and so, she adds, they are relying far more on the so-called vendor model these days. By that, she means that such companies tend to hire specialized vendor companies to take care of particular cloud-related tasks as they need them. In so doing, she points out, such companies are participating in a fundamental shift in their business models from an older customer-vendor model to more of “a partnership,” as she puts it with service providers. 
 
“We see these smaller media companies really adapting,” she says. “But things have changed so that it is no longer a simple vendor and customer relationship, like for render sharing, for example. Now, it becomes actual partnerships. The vendor has to secure the media assets and protect them, and the media company has to get through audits and feel secure about what the vendors are providing them. So now, the business model changes—you pay as you use a service, so you can automatically scale up your facility, in a sense. You can partner with the vendor, select a cloud-native platform, tell them what days you need certain services, and you can spin it up until you are done, and then spin it down, only paying for what you have used. 
 
“These are more business-related issues than technical things, because the technology has become [more vibrant and available] than it was before when moving to the cloud, because workflows are more established. But this model creates opportunities for smaller media companies to properly calculate within their investment cycles whether it makes sense to invest in giant infrastructure or concentrate on methods of uploading assets to the cloud and then paying [vendors] as they go along. For some organizations, it makes a lot of sense to focus more on the business of content creation and distribution, and less on the maintenance of technology.” 
 
Irrespective of a media company’s size or infrastructure and what approach works best for it overall—building a massive internal IT infrastructure, private networks, and all the rest, compared to relying largely on vendor partnerships—certain aspects of the new paradigm are largely about properly wedding the broadcast culture with the IT world’s way of doing things. Görner points to security as a prime example of this.  
 
For the most part, the media industry needs to keep its focus on translating for its own needs existing IT-based security standards that already exist for their own needs—protocols like the EBU’s R 143 cyber-security standard, rather than “trying to re-invent the wheel,” she suggests. “The goal is to make those recommendations easier to understand and more accessible to everyone in the media industry, not to try and invent new ones.” 
 

Because of all these variables, at the end of the day, Görner suggests that the biggest change facing the media industry when it comes to adopting and shifting seamlessly into cloud-based workflows is the business model challenge. Even figuring out how to calculate and evaluate what your costs are, could be, or should be on the Cloudscape can present a huge learning curve, she emphasizes.  
 
“We need lots of education about business models,” she says. “How is total cost of ownership even calculated? Before, I had a traditional calculation for a certain workflow. How do I translate that to the cloud when, after all, things are often not apples-to-apples. With traditional workflows, there are a lot of fixed costs associated with that way of working, like amortization of assets. To keep cameras and computers running requires electric power, cooling, cleaning, and so on. But it’s different once we move into a cloud environment. When you count up cloud costs, you don’t have as much to do with your entire infrastructure environment—all the redundancy and maintenance and management costs. With the cloud, they will be included in your contracted rate with a vendor in many cases. But you will have costs for bandwidth and modular services as you spin them up when you need them, if you transfer data between cloud environments, and so forth.” 
 
Görner hastens to add that there are good-quality financial cost management tools and services available to media companies. Further, there is a rapidly growing program called FinOps (Cloud Financial Operations)—offered through the FinOps Foundation, which is run by a group of partners that includes the Linux Foundation and the Cloud Native Computing Foundation. That program is essentially about spreading tools, techniques, and education about sound financial methodologies for those involved in cloud-based business. The FinOps website and literature emphasizes Görner’s key point about this ongoing transition—that it will take the creation of new and evolving best practices for different kinds of media companies. Indeed, the site explains that “at its core, FinOps is a cultural practice … where everyone takes ownership of their cloud usage supported by a best-practices group.”  
 
But Görner feels such education “needs to be even more accessible to media companies, particularly smaller media companies.” And that goal, she adds, is a major focus of the SMPTE Media in the Cloud Advisory Group, among others.  
 
“In [the Advisory Group], we are working on these multiple angles—not only the technology and standards, but also looking at the supply chain and workflows—all the areas you need to understand to build a sustainable business model,” she says. “We have focus groups working on these things, and people creating outputs that potentially can evolve into standards or adapt existing standards and clearly adapt them for the media world. Also, since we have huge cloud providers like Amazon AWS, Microsoft Azure, Google Cloud, and others, we want to help the media industry learn more about the variety of cloud strategies available to them, and figure out how we can bring in standards for multiple cloud vendors so that multi-cloud strategies can work more efficiently.” 
 
Along those lines, she points to the SMPTE+ Media in the Cloud virtual event that was held recently to address many of these issues, and she expects more upcoming in the near future. And more importantly, Görner says she is eager to see other entities across both the media and IT worlds do likewise and even work together on new initiatives—entities like the EBU, MovieLabs, the SAS Institute, OSA (the Open Security Architecture organization), MISA (the Microsoft Intelligent Security Association), and several others.