Cloud Migration in 5 Steps
Moving a media workflow into the cloud is no longer uncharted territory thanks to early adopters who have already demonstrated the benefits. Several of the largest media operations in the world have moved their content supply chains into the cloud, where there are no capacity barriers, bottlenecks or idle equipment. We know from their experience the model works - it saves money and time, and it’s inherently adaptive.
The prevailing question is now, how do you get there in the least disruptive way? This guide provides a template for making that assessment. No singular route applies equally to all operations, but there are general steps that can be applied universally to create the optimal migration path for each individual operation.
We propose a five-step strategy that can be applied to all or part of the media supply chain, which we’ve broken into common workflow segments for individual evaluation to determine an optimal entry point. The common workflows are enumerated in Step 1 and elaborated on further in separate section.
The five steps are as follows:
Let’s start at the beginning…
Step 1: Map the Media Supply Chain
This step involves:
collecting information about the supply chain
identifying pain points where change is most urgent
Here, we use the term “media supply chain” to refer to the full content lifecycle, from receiving to distributing and archiving and all the processes in between. Existing on-premise media supply chains consist of multiple different systems, from multiple different vendors, often delicately tied together, sometimes with custom software. They are inherently brittle, and frankly, once you get them humming along, you don’t want to touch them. The problem is, change is imminent in the digital realm, and this on-premise conglomerative system is limited. Viewing it now as an evolutionary step toward full digital immersion into the cloud will help us define that migration path.
Start where you are. Determine what the content lifecycle looks like today. What is the track record of the operation—being asked to do more with less? Left-field requests from the non-engineering end of things? Workload feast or famine? A little fight-flight-freeze in every execution? Let’s bear these conditions in mind as we envision a high-level view to map out the media supply chain.
First off, contemplate the breadth of the media supply chain you are trying to manage. Where does it start and end, and what are the individual processes being managed? (We’ll consider this in a linear fashion, bearing in mind the cloud can be linear in any direction. This linear model does not lock constituents into the analog-like linear model as earlier digital workflows configurations did.)
If you look at your supply chain as a broadly left-to-right process, what are the big steps you go through? Divide it accordingly where you find manageable segments or workflows. Once these workflows are defined, you can begin to home in on that optimal entry point for migrating the media supply chain into the cloud.
Let’s use a typical media content operation to determine where the supply chain starts and ends, and what workflows exist. This is a fairly high-level view of a generic media supply chain divided in content workflows:
(What each of these workflows comprise is further hashed out in Appendix A: The Common Workflows.)
Consider the distinct conditions within these workflows in terms of aging out, reaching capacity and other replacement-cycle criteria. Contrast these conditions (x) with the functions within the workflow where demand is escalating most rapidly (y). The resulting intersection is likely an operational terminus or at the very least, a bottleneck. These are the pain points to factor into laying out a migration path.
Step 2: Determine the Entry Point
This step involves:
envisioning an ideal media supply chain
drilling down into the pain points
identifying the primary goal
Once the workflow pain points (or the likelihood of them in the near future) are defined, focus on the circumstances around them. Who does what, in what order? How much effort is duplicated across teams or work units? Which steps are automated and which require people? How long does each step take? How many emails, spreadsheets or manual steps are involved? Identify the root cause of the pain point.
The optimal entry point will vary according to the scope, size and reach of the individual media operation. Does it accommodate 12 languages or two? Is it local, regional or global? Is it a single entry point or on every platform under the sun? Other factors might include simplicity of migration, ROI measurability, urgency of demand and others.
Don’t leave any stone unturned in this process. Talk to everyone involved—from the operators performing the tasks to the people supporting them, including those delivering the content and those waiting for it. Once you have a comprehensive idea of how things work today, consider how you want it to work in an ideal world. What does your ideal supply chain look like in its future state?
Imagine the process in a system with no limitations and draw it out. Don’t think within the constraints of your current system, but according to the needs and demands of the operation—now, and in years to come. Sketch out how you want content to flow, what needs to happen at each step, and what you want the outcome to be. Don’t go to deep; just form an overall idea that can be simply articulated, e.g.:
“This system needs to reduce manual intervention by x, increase capacity by y and decrease costs by z. It needs to accommodate (local, regional and/or global) content receipt parameters, automated metadata tracking, format detection and transcoding for distribution in x number of formats for x platforms, in real-time and on-demand, plus master-format archiving with instantly retrievable files through any of the number of transcoding engines.”
Assessing Optimal Entry Points
If content receipt is the primary bottleneck, it may be the optimal entry point. Perhaps the next logical step would be to migrate archiving into the cloud, which is rapidly becoming less expensive, reliably secure and more easily managed than on-premise storage.
Content distribution is another possibility, particularly if content is delivered to multiple places, with multiple disparate systems dealing with each platform. Managing that in the cloud eliminates idle capacity and provides cost-calculation accuracy.
Let this ideal be the overarching objective, but not a concrete target. Just getting started is the point, and nailing down a quick success will generate momentum. The “ideal” will likely be refined as the process unfolds. Flexible thinking is important. Set an objective that is well-reasoned but not etched in stone.
Next, define individual goals that move the operation toward the objective. Start with the operational pain points. Where are the bottlenecks? What processes are most labor-intensive? Maybe you’re dealing with several far-flung content receipt sites taking in multiple tape formats and file types that must undergo some form of manual validation. Do a process evaluation to see where automation would be most beneficial. (See, “Determining Where to Automate” Appendix B.)
Determining the optimal entry point may be a simple matter of what will have the biggest benefit or be the easiest to reach. Once this entry point is determined, the next step is to develop and convey the strategy for achieving this primary goal.