Build vs. Buy at Sinclair Digital
Ben Miller, VP of digital product development at Sinclair Digital Ventures, spoke at Streaming Media West on the panel "How to Succeed with OTT: Tackling Business Strategy and Unlocking Revenue," and later spoke with Streaming Media on about his thoughts on the "build vs. buy" conundrum. It's not a clear either/or proposition, and Sinclair does both—in fact, Sinclair has patents pending for technology it is building.
Sinclair Broadcast Group is one of the largest broadcasting companies in the U.S., with 173 stations in 81 markets providing local news, weather, sports, and lifestyle programming. Today most content Sinclair serves is clips produced by its own affiliates. "It's a strong, strong asset. That local engagement is incredibly difficult to reproduce," says Miller. They also produce national syndicated content and have OTT properties including ASN (American Sports Network), Circa, and Comet TV, and the company recently acquired the Tennis Channel. With this variety of content available on multiple devices, standardization and efficiencies are vital. Miller sums up Sinclair's approach in a number of areas.
Moving Server Side
"We have a whole initiative at Sinclair called Shift Left," says Miller. "We want to know how far server-side we can push every decision, so we don't have to rely on client-side players, client-side adoption and client-side technologies, because one of those is scalable and the other is not." Part one is using server-side ad insertion to counter one of the industry’s biggest problems, ensuring ad blockers are not removing ads.
Part two is replacing broadcast ads with digital ones, so Sinclair can sell the digital inventory. "When they go to ad break in the (broadcast) booth and they hit a button, we automatically encode the ad break on the digital side also. We don't have to do anything extra. Many companies can't do that, particularly live. For the live scenario, the ad-break lengths are variable. We have IP that we've developed," says Miller. "It allows us to take advantage of the cues happening already in the production flow."
"Rights are insane," says Miller. "Our digital rights for distribution for sports content, for instance—even highlights—are incredibly narrow. If we own the operation we can do live streaming to first-party properties, but not second-party properties," says Miller. Included in digital rights are which resolutions can go where. "They're worrying about the content getting out there (from HD files) so it has to be hyper encrypted," he says.
"It's something we are actively engaged with, and it's not one size fits all by a long shot. We have deep integrations going on all the way into the production side of the house where we're flagging things as being owned by this rights or that right and where they can go out on the channels," says Miller. "To say DRM is a challenge for us would be way underestimating the impact that it is having on our day-to-day operations. I think there is a sea change happening where a lot of this will relax, but DRM encryption and people worrying about people stealing their content, that's never going away."
If you can buy off-the-shelf technology, Miller recommends doing so. "If you can outsource whole massive sets of your infrastructure, do it. Write up requirements, figure out what it is you need to happen in your work flow, evaluate the tools, get an expert involved that you trust that's objective and isn't baked into any of the vendors," he says. "We have a deliberate philosophy: If it's commoditized or something that isn't going to be differentiating to the business, we want to lease it. The industry moves so quickly that leasing services in the clouds, leasing computing power in the cloud, even leasing encoding in the cloud makes sense."
"We are looking hard at ways to differentiate the business and realize if we don't build some things, we become just like everyone else, so you have to reserve those key innovative points that differentiate the business," he says. "Otherwise you become an also-ran, and that’s death to a large media company."
Know Your Audience
Different users exhibit different behavior, and prefer different experiences and require different methods of content discovery, says Miller. "How do you engage your creative production staff to take a look at the analytics to understand who their audience is?," he asks. "For us it starts with audience segmentation. Who is your audience? Can you react to trends in their discovery and consumption habits? Can you differentiate someone who is an avid sports fan with someone who loves talk shows or first-run movies?"
How does Sinclair figure out what content is relevant or what's going to work for its audiences? The company uses something it calls deep segment semantics analyses, another area where Sinclair has developed its own IP. "We are working really hard on analyzing the content," Miller says. "What is this content about, what's its format, and what's the response? (Knowing these answers) allows us to do predictive things, and we can take that data, feed it back, and look at a new piece of content and to an extent predict what its reception will be."
Producers get real-time feedback about how effective the content they’re creating will be, for which audience segments, and on which devices. "We can give (data insight) to the producer and say 'This thing that you've done is not going to do well and here's why. This other thing that you've done is going to do great and here's why'."
Sinclair also tracks what a viewer searches for versus what they watch. "Reading the tea leaves in part one means giving them what they want; the second part is the opportunity to offer them something new, which they may also want," says Miller.
"We Can’t Build an SDK for 200 Devices"
Sinclair is focused on HLS delivery. "How many different formats do you really need? You have to pick the top 20 - 25 devices, "says Miller. "We've innovated and put things in place to rewrite the manifest headers for different devices to check the device, work with the CDN provider so the CDN is detecting that device and giving it different cache response. We have two or three patents pending on this and the ad stitching technology. This is where we decided, it's not commoditized, there are serious problems in the industry that need to be solved and if we pick the right mix of technologies on the server side and we hone in on the 98 percent of the audience we can provide a really good experience that we can manage and monitor. What we can't do is build an SDK for 200 devices."
Is Online a Viable Business for Local Media?
The cost of creating online content is a lot less than for broadcast production, so Miller sees it as a way to generate additional revenue without generating a lot of additional cost. "When you look at it, it's just additive revenue and an additive audience and possibly promoting and bolstering the existing core business. It pays for itself right there," he says. "In the grand scheme of things, it's not a massive investment for the company, really, compared to the opportunity," says Miller. "I think to look at an OTT investment in a niche or in too targeted a way in conjunction with a traditional media business is a mistake. The shift is gradual. The shift is real. You're going to need to be there."
Can millennials be lured back to standard television? In 2017, TBD will deliver videos from popular web studios to broadcast viewers.