Metadata in the Cloud
This blog is one of a series of Media & Entertainment Cloud Ecosystem interviews.
Executive Interview with Rob Tucker, Founder of META
While in-person events have been curtailed by COVID-19, innovation continues to advance in both tech and media and entertainment. I’ve been sitting down (virtually) with leaders of some of the leading AWS Partner Network (APN) Technology Partners involved in the M&E ecosystem—and their insights can offer a lens into the future of the industry.
With over 15 years innovating tech solutions in media and entertainment, Rob Tucker leverages his technical expertise and industry insights to help broadcasters and studios maximize the value of their catalogs. He founded META nearly four years ago to provide comprehensive metadata services to leading media companies, and leads all operational, technology, and design aspects of the business.
In this interview, Rob and I chat about metadata, catalog management, supply chain integration, and what he sees for the future of the industry.
Q: For people who aren’t familiar with your company, tell us about META.
A: META came out of some consultant work that I was doing. I spent a lot of time consulting with film studios and broadcasters across the supply chain on things like supply chain rights, scheduling, asset management systems, all that kind of stuff.
Something that’s become obvious over the last decade or so: the asset side of the business—video, components, etc.—has had a huge amount of investment, but the data side of things was all over the place—and still is all over the place. There’s lots of custom-built legacy systems, complex integrations, transformations, that kind of thing.
I was on a project for 12 months with Warner International, and it’s almost as extreme as it gets. They’ve got 200 channels, 80 languages, delivered to 3,500 VOD partners and platforms—so it was the extreme end of the spectrum. And it was all being served through these custom legacy applications. I worked with James on that to come up with the original project—or application specification—and that’s really how META started.
It initially was quite straightforward: a total catalog management and enrichment platform— with a real focus on user experience, which with their quite extreme international requirements is the best place to start.
As an API-first application, what started to happen was that they would have other systems in their supply chain that were quite horrible to use; poor user experience. So we started to get asked if we could integrate into those systems and bring that data into our system—present it through this lovely user interface. We’ve shifted from being this total catalog management application to more of a supply chain front end.
Q: Yes, we have called it the spaghetti monster or the spaghetti ball. I wouldn’t even say it’s just spaghetti. You’ve got some ravioli in there for sure, and some tortellinis—all kinds of things.
A: Yes. Everybody’s got a different flavor!
Q: So, tell me, as it relates to AWS and the cloud space—was that part of the architecture from day one, or did you come to it later? How did you get involved in the cloud?
A: We were with AWS from the very beginning. How it started was, we applied to be on their Activate scheme, which is this scheme they have for startups. What that gives you is credits to get started—and fundamentally gives you access to all of the principal solutions architects, and that has just been incredibly valuable for us.
You’re trying to build these quite complicated SaaS web applications, and you’re able to quickly get access to the leading people in the field to help you and guide you through it. We’ve had phenomenal support from Amazon from the beginning—and still do to this day.
We have engaged with other—I don’t want to be dismissive—but we’ve engaged with other cloud vendors as well, and we don’t get anywhere near the level of support or even knowledge in this space. We’re huge fans of AWS.
Q: That’s fantastic. And within the ecosystem—obviously with a company called META, you’re dealing with metadata. What are the other places that you plug into with the M&E space?
A: In terms of our ecosystem or the media and entertainment supply chain, you’ve got the standard stuff. It’s different per company. You have rights and your catalog systems, MAMs, localization, scheduling, delivery—all knitted together with some integration layer or service bus or something.
It’s really hard—we see this all the time—it’s really hard within a business, within the supply chain, to know where to go to get certain bits of data, or where the source of truth is for a given attribute. If you want to report on something from across those systems, it’s pretty much impossible.
What we do is start at the very beginning of the supply chain. So the title catalog lives with us. Actually, that’s not quite true. Normally, there’s a rights management system, and a title will begin its life in there. So we’ve got the rights for X title, or we’ve created X title, and that will have a tiny piece of metadata associated with it—like a release year and a title, or an episode number or something.
It then flows into us, and then we become the source of truth for everything downstream. So in the first instance, what we would do is connect into some of the standard enrichment services or matching services like EIDR, IMDB, Common Sense Media, Gracenote, all of that stuff. We can start to aggregate that per the companies’ requirements. We really start a little bit outside of the ecosystem or the supply chain.
We sit on top of it. Once a title gets created in us, we then go into other systems and start performing the necessary tasks. We will go into an asset management system, and we’ll make all the placeholders and folder hierarchies that they need for the content. Same thing with image management—we’re setting all the IDs, so we can keep the supply chain reconciled. We’re integrating into all these systems, so in the first instance, we can keep the supply chain consistent—but what we then do is pull key data in from those systems that the customer wants to see.
And they force multiply—and the effect of that is that now, through our application, you can then report on anything from any system. You’ve got all the rights data and all your assets and that kind of thing, so you can run reports. Like, show me everything that is scheduled to go on this platform, on this date, in this country, that doesn’t have these subs or dubs, or is missing this synopsis. It gives you this incredible visibility over the whole ecosystem.
Q: I’m curious. When you talk about EIDR or IMDB or Gracenote—is there a category for that type of product? I know what they do, but what do you call it?
A: That is so interesting, and it’s almost like I fed you that, so you could tee that up for me—which I didn’t do! It’s absolutely true, we’ve been looking at this today. It’s a thing.
There’s all these different types of enrichment sources. You’ve got the ones like IMDB and Gracenote and TiVo, and they’ve got comprehensive datasets. Then you’ve got more niche ones, like Common Sense Media or EIDR or whatever it is. And it’s a bit of a—it’s not a mess, but it’s just very vague. There’s no real definition around it.
A couple of weeks ago, we actually had this conversation. We were saying, “You know what would be really good, would be a whitepaper about the landscape. Who are these sources? What do they cost? What kind of data do they have? What kind of technical delivery mechanisms do they have?” We’re actually working on that.
So the answer is: yeah, that’s a great question. It’s something that we’re hopefully going to put out a paper on over the next week or so that starts to put in some structure and explains that whole area.
Q: It’s so funny. I’m working on a whitepaper right now that’s going bigger in terms of the ecosystem, but you coined it: enrichment sources. I’m gonna create a little space in the ecosystem for the enrichment sources. So, if it sticks, it sticks. We’ll see.
So talk a little bit about the cloud technologies, other than yours, that you’ve read about or you’ve heard about or you find especially interesting.
A: I try to stay abreast of what’s going on, but we really get into the detail with customers and the types of systems that they’re deploying.
Some of the stuff that looks really interesting at the moment is the integration layers that have always been really complicated. You have all these cloud-based systems, but then how they were integrated was through some potentially old, legacy integration layers or service buses and that kind of thing.
We’re working with MGM, and they’re using a company Workato, and they do this—I think it’s like iPaaS, like integration platform as a service or whatever that new sector is. They’ve got this really lovely approach. It’s a cloud-based application, and they’ve got these recipes that you put together—which is quite a cool way of approaching this new integration space. So that’s one that’s really interesting.
There’s another in that specific space, which is Reshuffle, and it’s the same thing but for developers. Developers can build integration through Reshuffle in incredibly short timeframes. So that’s integrations and how a couple of companies are taking out a lot of the complexity there around integration layers.
Then more broadly, as supply chain metadata in general improves and matures—which it is, and we see that everywhere—I think there’s going to be a huge shift in personalization of content.
You need to do the basic stuff well. You’ve got to have your editorial and technical metadata good, which people are getting to. But then there’s companies now like Canvas, who do emotional and audience sentiment response, which is fascinating.
Then getting into things like personality profiles and character traits and habitual data—that is where stuff is gonna get really, really, interesting and hopefully get to that north star of having a UI-less entertainment experience. With that kind of thing—and it’s actually going back quite a few years now—I always think about the UI in that movie Her by Spike Jonze, where that whole thing is completely data-driven from the application side, but also from the user side. You just get the user entirely out of the way. You know—less is more, and you just deliver the content.
So I think things are definitely moving in that direction. And I also recently started looking at some of the stuff that Melinda Ozel was doing around facial expressions.
Q: Like sentiment analysis?
A: No, so you’ve got Canvas, who are doing audience sentiment by scraping what people are saying about stuff and finding what people feel about it and presenting that to an API—which is super cool in itself.
But what Melinda does is more about the extraction from video content. So, it’s like being able to scan in a movie and pick a scene—and actually go into the faces and pull out the emotions of the characters in the scene. And so this builds up an emotional profile of a piece of content. When you start to pair that—the type of profile of the character of the content with the profiles of the characters of the people that are watching it—then I think that your personalizations are gonna become orders of magnitude better than they are today. So that’s one area.
Then a company that for the last three or four years has been continuing to do super cool stuff, is WIREWAX. It’s a company in the UK. They started out as an agency, but now they do shoppable video stuff. They had to build tools to identify heat maps in videos and things—so you can see that’s a Fendi bag and go and buy it.
But what they’ve done is shifted that technology to do extraction of videos. They’re always coming up with these super cool tools—and then about 18 months later you start seeing them being deployed in some bigger vendors. So, they’re always leading the curve in that space.
So that’s a few different businesses that I think are pretty cool at the moment.
Q: We were just talking about this before we started the camera—but coronavirus, right? You’re in London; I know you guys are on lockdown. What do you think the impact of the coronavirus is going to be on the media and entertainment industry over the next 12 months?
A: Obviously, it’s a quite difficult, crystal-ball kind of question. To put it in context, from our perspective, the impact initially was a—what do they call it, white swan or black swan event. Our customers have channels and platforms and things that are revenue-driven. You’ve got customers in the airline industry as well, so you can imagine that’s a tough place to be.
And that’s gonna trickle down on us, and it has made things more challenging. You know, obviously, there’s this huge shift to working from home, which initially seemed to be quite manageable. But I think companies across the board are starting to feel the lack of community and the absence of this spontaneity that comes up when you’re working with people.
For us, building new features and continually innovating the product is a fun, collaborative process—with your whiteboards, animated discussions, and all that kind of thing. We really struggled early on with the distance between the team members.
A few times I’d be on a call with a product team, and we’d really be struggling to talk through something that was an absolute breeze before, and it would get to the point where I would just say, “Forget it. I’m coming to your house now.” Then we’d just sit down and scribble it out, and you’d smash it out in half an hour or so. But it just got to that point.
Something else that we noticed is, when you’re like us, and you’re a vendor in the space and trying to connect people and get their time—so you can show them what to do—it’s been very difficult, historically. But now, weirdly, it’s been much easier because people aren’t travelling around, and they’re not going to different offices for meetings and things—so that whole portion of people’s calendars have been removed. It’s been replaced with these boring Zoom sessions now—not like this one of course, but everybody’s calendar...
Q: All the other Zoom sessions, not this one!
A: No, no—this is much more interesting than the normal ten people looking through a spreadsheet. But yes, it’s easier to get people’s time, which has been a positive outcome. But I think it’s really starting to reach its limits.
We started to, within the rules and regulations, start getting people back together as much as possible for key things. So that’s more of an answer of our perspective, and I’ll leave the crystal-balling to someone else.
Q: What’s interesting about what you described in terms of that connection—I think that the existing connections, the preexisting connections, were sort of feeding on the connections that we’ve made in the past. Like you and I, we’ve spoken to each other, and we’ve worked together in the past.
I think that for customers, too, they’re not going to conferences. When they can’t go to conferences, they can’t get something that’s going to stick and be stimulating. I think the relationships they’ve had with people in the past become even more important. It’d be really tough to break into the industry right now.
A: Yeah, it’s a fascinating thing—I was thinking the other day that we would frequently be in our customers’ offices. They’re embedded working with the teams and things. And I just realized, we haven’t made any new friends since it started. You don’t realize that, but you’d be in a meeting, “Oh, who’s this new guy, this is so and so,” and then you have a little chat. You don’t get that anymore at all.
And then, weirdly, we were on a call—we had this call scheduled a couple of days ago, and somebody wasn’t able to make it, but we didn’t find out until 15 minutes in. So we just had a chat for 15 minutes. It was the most office-like experience that we’ve had in a long time. “Oh, we’re just here, why don’t we have a chat,” because you would normally do that when you’re working outside a meeting with someone.
Q: Yeah, I’ve started actually going back to college friends and looking to see which people I knew from college are living in town, you know—it’s “Hey, let’s reconnect.”
So, going bigger. Get your bigger crystal ball! That was the little one; this is the 35mm, 45mm one. What do you think the impact of cloud is gonna be on business over the next three years?
A: Well, I can remember the pre-cloud supply chain. I’ve seen that shift from how it used to be to these cloud-based data-driven supply chains. So I can see the impact up until now has been phenomenal. As that matures and we move forward, we’re seeing customers now in their second generation cloud-based value chains.
I think it’s less of a cloud thing but more of an industry thing—that all these systems are getting easier to tie together. The longer they’ve been in the cloud, the more mature they become, the easier they are to plug together. So there’s standards and frameworks emerging that let these supply chains be put together in really short timeframes, and it’s going to continue.
As an example—which is a bit left field—we’ve been working on this slightly guerilla project with a customer that is essentially putting together a quick delivery pipeline in AWS. So it’s taking some metadata, some images, video, multi-language subtitles, and other things—then sucking that in, ingesting it, validating it, transcoding it to DASH and HLS with all the ABR bitrate ladders and things, and then pushing that into a customer’s API. We’ve been able to get a working version of this in, basically, a week.
That is something that a few years back, even the notion of that would’ve been ridiculous—and now it’s something. So, I think moving forward, it’s just going to continue. Obviously, you’ve got the technology in the cloud, but you need the right people in place to be able to leverage it, make the most of it. But I think it’s just going to continue to make companies more agile from a technical perspective, to be able to adapt to market shifts more quickly. And then do a lot more of these things like pop-up channels and these really quick platform launches. So, I think, it’s just going to make businesses a lot more agile.
Q: Indeed. Well, Rob, thank you so much for the time. I want to ask one last question. If people need to get in touch with META or they want to get in touch with you, what’s the best way to do that?
A: So, first go to the website, which is www.meta.how. The idea behind that URL was that people would go there to find out how META can help them. But now all that happened is people think the company’s called meta.how—which was the unintended effect!
But that’s the way—go to the website and all the contact details are there.