Why Your MAM is Choking
By Sarah Semlear
Transcoding comes last in a lot of enterprise video operations. You spend months comparing storage vendors and agonizing over a Media Asset Management (MAM) system's UI and metadata tagging, then you map out your orchestration layer, get the budget signed off and roll it out. Somewhere along the way you assume the file processing will take care of itself in the background.
Until it doesn't.
I sat down with Jeff Sengpiehl and Eivind Sandstrand on their Cutting Through the BS of Corporate Video series to talk about the two most critical and most ignored points of the enterprise video content factory: the front door (ingest) and the loading dock (delivery).
I ran video operations at Netflix, BuzzFeed and Comedy Central before I took over go-to-market at Hiscale, and I've watched this scenario play out more times than I can count. A team builds what it thinks is a state-of-the-art video factory, then watches it seize up because they ignored the engine powering the whole thing.
Here are the hard realities your corporate video infrastructure needs to face.
The Scapegoat: Stop Blaming Your MAM
Let me paint a picture that might sound familiar.
You buy the shiny new media asset management (MAM) solution you've wanted for years, and your team is thrilled. To get it up and running, you dump years of town halls and marketing shoots into it, terabytes of raw footage in one go. Then the uploads stall, files won't play back, and your frustrated team starts pointing fingers at the MAM.
But here is the ugly truth: in my experience, the culprit is your transcoder.
Transcoding is the math equation running behind your entire system, and many MAMs rely on built-in open-source tools like FFmpeg to do it. FFmpeg handles basic day-to-day jobs, but drop an enormous backlog of company history on its head and it burns out fast, because it lacks the horsepower for enterprise volume. You need a transcoder built for heavy lifting, so your MAM can get back to what it does best: organizing and tracking your assets.
The Front Door: Fixing Ingest Before It Kills Your Workflow
Before you transcode anything for delivery, you have to get it into your system intact. Jeff and Eivind call this the "front door," and in most organizations the front door is broken.
Your camera crew should be able to offload footage after a long shoot without an engineering degree or a multi-page manual. Plenty of brilliant DPs aren't computer people, and each hoop you make them jump through is a chance for a mistake. Set up automated hot folders, give the operator one single, simple action, and you remove the friction.
You also have to verify your ingest, because an unverified upload is a landmine waiting to go off. Imagine this: your videographer offloads the card from a high-stakes CEO interview, watches the progress bar hit 100% and wipes the card for the next shoot. They had no way of knowing that the transfer corrupted the file, or that it only reached a local cache and never made it to the central server.
The footage? Gone for-ev-er.
Verified ingest checks each file byte for byte before anyone gets the green light to format a drive, and it costs a lot less than reshooting your CEO.
The Need for Speed: Growing Files and Live Events
As Jeff pointed out, a lot of corporate video comes from live internal events like town halls and all-hands meetings. If you wait for an hour-long recording to finish, then offload, ingest and transcode it before an editor can pull a 30-second clip for LinkedIn, the moment has passed by the time the clip goes out.
A proper processing engine supports growing files, so your editors get the frames while the event is still recording. They can clip and publish highlights of the CEO's keynote while the CEO is still on stage, where a legacy system would make them wait for the file to close.
Stop Bleeding Cash: Elasticity and the Hybrid Cloud
Each camera and each social platform wants video in its own format, and as Eivind put it, "this carousel of video formats or codecs just never stops." A lot of companies moved all that crunching to the cloud, where you pay for compute every hour you leave it running.
You need big compute for the spikes, like an archive migration or a global live event, and a fraction of it the rest of the week, which is the job of a hybrid architecture. Hybrid done right, as Eivind called it, means you run a stable, cost-effective on-premise baseline for the predictable daily work, and your system bursts into the cloud on its own when demand spikes, then scales back down as soon as the queue clears. You pay for cloud compute during the peaks and nothing while it sits idle, so your CFO stays happy and your editors keep moving.
The "Magic Box" Myth
Finally, we need to talk about the myth of the "all-in-one" solution. Plenty of vendors will sell you a single magic box that claims to handle ingest, transcoding, asset management and delivery in one place.
Don't fall for it.
A transcoder's job is to transform files, and moving and sharing those files belongs to your MAM and your orchestration layer. Eivind made the point in the video that one box claiming to do everything won't do any of it as well as a dedicated tool. You're better off with a stack of specialist tools that connect to each other, because each one is excellent at its job and together they make something freaking cool.
Transcoding Should Do Its Job Quietly
At Hiscale, we know that transcoding is boring, so we build tools that let you think about it less. You have better things to do with your week than sweat over hidden codec licensing fees or explain to your boss why the MAM is crawling. Your processing engine is the math behind your content factory, and it should sit in the background, scale up when you need it and stay out of your way while you get back to working with your content.
Where does video get stuck in your organization? On the way in, or on the way out? Watch the full episode of Cutting Through the BS of Corporate Video for the whole conversation and a live demo of how we solve this at Hiscale.
