The first thing that came to mind when Verimatrix explained its new Verspective Intelligence Center to us this week, was that it was Google-esque. It is a way of merging the data from 800 or so global operators and looking for beneficial patterns inside it. We also wondered whoever thought up such a project.
We talked together with Steve Oetegenn, President and Steve Christian, Senior VP Marketing
Christian gave us a for instance on how Verspective might work, “If we find a Verimatrix watermark on a piracy site, each operator has its own outputs, and at the moment, we have to check the mark against each operator to get a match. Using Verspective we will have all of that data in the cloud in one place and it makes it a lot easier.”
For easier read cheaper, both for the client and for Verimatrix, and saving costs is really important right now in DRM because it is in danger of being simplified out of existence. Browser companies like Google and Apple are getting rid of plug-ins, and with that favoring only their own DRM components – making DRM automatically multi-DRM, and pushing down opportunities for proprietary systems.
Christian continued, “It also gives us the ability to collect analytics data on how keys are used. We anticipate it will lead to the collection of a large amount of authentication and entitlement data which can be manipulated by Big Data tools.”
But he is clear that the data they are playing with, for the most part, doesn’t belong to Verimatrix, but to its clients. “We won’t write Big Data tools ourselves, we will partner for those, and with other Apps and deals like that are perhaps 6 months away. For now we will focus on improving systems deployment and fault management on ours systems.”
“On our new product Ultra (the version of VCAS for supporting UHD video) it can run on physical serves or on a virtual machine or in the cloud, where it can use Verspective. We have to distribute secret keys for chipsets, and we can mirror these and back them up in the cloud. There are a certain set of functions which belong to the operator and these reside in the operator’s domain, and there is data which does not belong to the operator and it’s in our domain.”
What we are supposed to conclude from this is that key usage patterns which are abnormal could help identify piracy, and this can be investigated, or that the cloud can be used to dispatch updates and patches uniformly to customers and that’s all well and good. Verimatrix says this will make its own operation scale better, and lower its costs.
It will give it a “bigger picture” and ultimately let it respond more rapidly to customer piracy concerns – if Verimatrix sees a threat at one operator, it can adjust operations so no other operator becomes susceptible to that threat. It can do that now, but it takes a good deal more work.
But there are many more things that this type of data can tell us. We’re not saying Verimatrix won’t use it to improve security, it will. It wants anything that can differentiate itself from other content security providers.
However it is also sure to use the data for more than that as well. Think about it, every time a home user presses his remote control (or touches his phone or whatever is controlling his video view) there is a change to the license stream. That change in the delivery of entitlement messages, if re-interpreted and stored, gives a blow by blow account of the viewing experience, which could be every bit as good as Nielsen achieves by interviewing 5 million people, or as TiVo gets in its clickstream data in its research and analytics.
Entitlement messages are generally sent in discrete cryptoperiods – if someone asks for some messages, and not all of them, they didn’t watch the program to the end. If an entitlement message is called for after a live or linear program has ended, then it is being called for from the DVR. If two streams are being called for one will be stored for later viewing.
It doesn’t tell you that 3 people were in the same room as the TV, but even this can be worked out. Recommendation systems have methods of divining who is watching the TV, and these are already in the cloud. Combine the two outputs and throw it at a big data machine and you have a hell of an analytics engine. The beauty of that is that it doesn’t involve reaching into a multiplicity of playing devices, and returning with extracted, anonymized data – no, because the data is already out there in your cloud platform and it can be made anonymous at a stroke. An operator can still control who has access to all of this, but what operator wouldn’t let his best content supplier have a peek at some slice of data – in return for some favor or other. So someone like HBO might be able to find out not just what consumers of a single operator thought of the Game of Thrones episode that just aired, but could see what the entire world thought of it, from, a single system.
“This will be the biggest pay TV platform anywhere in the world,” said Oetegenn, “and that’s a lot of anonymized data, by the time it is ready it will be from 1,000 operators, showing us viewer statistics and behavior.”
“That kind of data will be able to point to physical network faults. It can be fed back into systems from Inneoquest and Agama, who’s core business that is, and also show who watched what and when. It could form part of operator reporting tools and analytics. “We’re not planning to compete with the analytics guys, but suddenly there may be lots of potential partnerships we can enter into,” said Oetegenn.
We raised the subject of Nielsen, would it be a potential partner? and Christian deflected us with some praise for its rivals like Rentrak, and it sounded to us that rivals to Nielsen which have gone further towards putting a system into the cloud were more likely to be high on the list of companies Verimatrix might partner with.
We thought more about this. If a customer has just stopped watching a particular set of channels, does that make him more likely to cut the cord. If they are watching more content VoD and OTT, does that make him right for an upgrade or a new service? If they have just stopped watching the TV in favor of their phone, maybe they are at risk of leaving the operator? It might provide information on the UI and how people navigate around it.
It’s not that a lot of operators don’t already have much of this data, but extracting it from a single cloud based system, without any interrogation of the end user device, may be cheaper to do, and easier. And this will get Verimatrix into the data game, get it a seat at the table so to speak, early on in cloud orientation for security.
The first thing that programmers learn when they use a cloud platform is to forget where a resource is, just focus on how little it costs, suddenly. The second thing is that you can build a series of interlinked separate systems, micro-servers if you like, or islands of data, which can talk to one another. You don’t focus on building a system, but a single ingredient for each system. And then the next ingredient. We appear, in pay TV, to be moving to a world with many islands of data, that may well talk to each other about piracy attempts, but can also talk to one another about improving the enjoyment of services or the cost of bringing them to people. And anyone that has a particularly large island of data, will be differentiating themselves by helping customers in multiple ways – at the same time raising the bar on the price of entry into their own market, by upping the value that each operator feels, once they have paid for security.
Now perhaps you can see why we began thinking this is Google-esque, because this is a typically Google type move.
Another story this week (see Kaltura Universal DRM elsewhere in this issue), reminded us that Google is already working towards a very similar position. It may not have announced a cloud data repository like Verimatrix’ Verspective, but right now it only offers a cloud instance of its Widevine DRM system. So instead of OTT companies building their own license server, as they do for Microsoft Playready, they simply apply to a Google run server for Widevine licenses, when they opt for Widevine as a DRM – something they are more and more inclined to do as Google withdraws HTML 5.0 browser support for anything else.
So Google is pushing operators to offer Widevine and at the same time already has access to the same kind of data that Verspective has – it knows whenever a license is served and can work out the rest from there. If you add the power of Google TV and its growing influence in the pay TV world in the set top, then it could establish a lock-in on audience, analytics and device fault reporting, to the detriment of anyone already in that business.
So when we wondered how Verimatrix came up with that idea, the answer must be, because it can see just what it is up against in Google and it is trying to anticipate any move that might put it at a disadvantage.
Maybe one of the things that Verspective will do is check when a user upgrades their Chrome browser to a version that can no longer takes a Playready license, and could perform a patch on how it serves that device licenses. Now there’s a thought.