Friday, October 23, 2015

Securing You Video Service

As telco giant TalkTalk in the UK becomes the latest corporation to suffer from cy ber attacks, I thought it might be useful to share some of the strategies we've used at TV Everywhere to deal with such attacks in the past on our video services.

We've experienced attacks on client systems ranging from political parties to the Red Cross down the years, so have developed a range of strategies for managing such attacks. First of all, let's presume that you've followed best practice in developing your systems, such as the guidelines found at, here are some further tips:

1) If you are able, don't make yourself a target. In other words, try and be nice. Unfortunately, as our Red Cross experience shows, this doesn't deter some of the people out there any more than being a child doesn't protect you from a mugger. There are just bad people in this world... Still, not writing about security as I'm foolishly doing here, or making any claims about the robustness of your systems
2) Monitor - make sure that you have monitoring in place so that you, not your users, are the first to detect an attack. And have a plan to make sure you know who does what in the case of an attack. There are tools that can even help you automate your response - most attacks start big and degrade, so your immediate response is crucial.

2) Control your DNS - your domain management should be separate from your hosting, otherwise you will have no control over your URL if your hosting company or cloud goes down. If you have control, you can repoint the service to a separate network or at least put up a holding page for your users as you put out the fires.

3) In most cases the first front of attack will be a distributed denial of service assault, which is the equivalent of a burglar using a tank to try and get into your house. Once all the walls are down there is little protection. The best way to ameliorate this is to have a large scale Content Delivery Network in place. The CDN can then scale the availability of your website whilst blacklisting the sources of the DDoS.

4) Ensure that you have redundancy. Using either a separate network or cloud hosting, make sure you have somewhere to redirect your traffic if necessary. We have also used degradable services, which mean, for example, that the video stream plays through a flat web page without all the heavy server calls,  JS and HTML that are usually deployed in video interfaces these days. You can also try redirects to apps, which are unlikely to be attacked. Of course, if you don't use a CDN and host your own video, you're in trouble.

5) Get sophisticated, which can use configuring your network to use sacrificial servers where you redirect the attack and bring it down whilst deploying your white hat team to track down the hackers and turn them over to the authorities.

Provided they're not the authorities of course....

Bye bye Sky

I have officially become a cord cutter.

Yeserday saw the termination of all Sky TV services beyond what Freesat offers. So, in a year I have gone from two Sky Sports subscriptions to none. I do not know if this is significant, but anecdotally I have spoken to a number of friends who have taken the same course of action.

When the likes of Amazon Prime and Netflix cost under a tenner a month, justifying forty quid for a few football or rugby games just doesn't cut it. Especially when they can be watched down the local pub.

For me, it was the rugby. The only rights found on Sky now are some Pro12, some European Championship and French Super 14 games. I can find alternatives for all of them on BBC, S4C or BT Sport. So, paying a massive tax which goes almost directly into the pockets of Premiership footballers and the Murdoch family becomes really galling.

Along with cutting out other packages, the result was a drop in our Sky bill from £102 to £33. We got most of the packages back by paying £7 a month for Now TV to add to our Netflix and Amazon Prime subscriptions. 

We also have a couple of Roku dongles and a Smart TV so everything is available on demand on these and our iPads.

Selecting a dongle (or stick) wasn't easy. In the US Roku carries all three on demand services mentioned above. Unfortunately, in the UK Amazon Prime is missing, allegedly due to the investment made by Sky in Roku (whose box is used as a basis for their Now TV box).

As far as I can tell, at the time of writing, the only way to get all the main on demand channels (BBC, ITV, 4, 5) plus the three on demand channels is to buy a Samsung Smart TV. I made the mistake a year ago if buying an LG set, which for some infathomable reason still doesn't have All 4 and ITV Player apps, and before you buy a Chromecast dongle, this suffers from the same shortcoming with a serious lack of services, as does its Amazon Fire TV rival.

So, you must excuse me now. I have a box set or two to watch....

Thursday, October 22, 2015

Netflix Has A Problem

Yes, Netflix has a problem.

It is not acquiring viewers as quickly as it promised shareholders and seems to be suffering from increasing churn, or customers leaving.

The company puts this down to the introduction of chip and pin to the US where credit cards are not automatically renewing.

I think this is a bit of a red herring which hides Netflix' underlying problem.

Once you've watched the 70% of original content, the 30% of box sets and the 2% of movies that are any good on the service, there's nowhere to go.

With more and more viewers watching 'box sets' voraciously, you can work your way through an original series in a weekend, and beyond TV drama and movies, the content pickings are thin to say the least.

Netflix needs to up its acquisition strategy because the competition is coming to get it, with more and more SVOD services being launched every month. HBO Go, especially, contains a lot of must-see TV (or Sky Go/Now TV who have the rights in the UK).

To date, Netflix has played a clever game of balancing the acquisition of C and D level movies with good original programming that it can exploit over the long term. It's clever because it builds a balance sheet rather than creating a P&L black hole.

But Orange Is The New Black and House Of Cards are, arguably,  not must see TV in the way Game Of Thrones, Mad Men or even NCIS are (Netflix themselves have admitted the popularity of awful The Blacklist and the naval cop series over their own programming).

Quite simply, Netflix has to find some big new original hits and needs to buy, buy, buy if it is to maintain its market position.

Friday, October 09, 2015

Is It Art ?

When investors value companies they find all kinds of ways to do it and one investor can place a value of £1m whereas another can place a value of £100m.

Why is such a thing as the value of  a trading company so subjective? 

Well, there are factors such as potential, property, defensibility, uniqueness.

And then you look at the art world. Here, in an artists’ lifetime, dreadful, poster art such as Monet’s water lilies were coveted, whereas van Gough’s sunflowers were derided.

Now they are both worth stupid amounts  of money.

But why is this ?

Well, the reason is quite simple, intrinsically they have little worth or value. Paintings of flowers are easy to do and easy to replicate – exactly if you want. 

All these paintings do is provide a value for hubris. The only thing a £20m painting represents is the arrogance of the person who paid £20m. It is a bet that their arrogance is worth more than that of the nxt person.

And so what price do you place on a song or a movie ? The most valuable song in history – happy birthday – was recently ruled to be out of copyright and is a dirge anyway. Poems, despite, arguably, being more beautiful and valuable than many paintings, are valueless. 

You own a copy of The Wateland ? Try getting £100 let alone £100m for it. Yet it is intellectually  more valuable than any painting, in my view. Erase the Mona Lias or erase The Wasteland ? I know which way I would go.

When I studied film there was always an argument about it as craft or art. The reality is there is no difference. Art is something speculators and hedgies make of craft. They have always been with us, fir centuries, which is why Monet died rich and van Gough died as a pauper.

Thursday, October 01, 2015

Balanced approached to IP protection

They say that it is a recession when your neighbour get fired and a depression when you get fired. Similarly IP protection tends to be front of mind when an issue emerges for an individual or company.

This excellent article from Chris Blackhurst in the Evening Standard illustrates the problem very well.

For a variety of reasons such as resourcing issues in the Police and corporate entities not wanting to end up with liability in an area they don't really grasp as yet  (and therefore not engaging with the problem) there is a perception that cyber crime is much more difficult to resolve than others.

At the corporate level standards such as ISO 27001 have emerged which contains guidance such as;

"Information about technical vulnerabilities of information systems being used shall be obtained in a timely fashion, the organization's exposure to such vulnerabilities evaluated and appropriate measures taken to address the associated risk".

For a medium to large corporate with dedicated IT function this may be helpful but for an SME this guidance probably seems very circular in nature unless there is a pretty high degree of knowledge of the risks to start with.

Reactions to cyber risk from TV companies to one man bands vary from denial to dusting off the typewriter and carrier pigeon and banning the internet and mobile devices. Budgets are possibly squeezed from the legal and IT budgets to deal with an issue that simply did not really exist 10 years ago.

Neither denial nor dashing off to a log cabin makes a lot of sense when cyber risk can be reduced by taking pre-emptive action to ensure that the very obvious vulnerabilities have been addressed. The equivalent of leaving the car keys in the ignition with the doors unlocked.

To discuss this further please contact us at KLipcorp IP.

Monday, September 28, 2015

Hole In The Street

I came out of my house last week to find that neither I nor my car could get out due to a trench being dug in the road right outside. On looking at the contractor sign I was intrigued to see that the chaps in the hard hats were laying cable for Vodafone. On talking a bit more they said they were laying it to the old NTL box at the turn in our road.

Most of Reading is cabled for Virgin Media, but it seemed to end at the box and be unavailable to us. Until now, I hoped. An opportunity to get off BT!

So, the news that Vodafone and Virgin's 'merger' talks have come to nowt is a bit surprising.

Without knowing the intricacies of the discussions it's difficult to see the logic of not pulling these businesses together. But business logic rarely plays a part when disparate valuations are involved, and I suspect that this was the stumbling point.

Vodafone now needs to dig lots of holes in streets, or pay BT Openzone through the nose for passing on a dreadful service to its customers. It also needs to build its own TV service to be a serious quad play operator.

Broadband, landline and mobile contracts are all commoditised despite being underpinned by a virtual monopoly. TV is the differentiator, as even BT has found. And a thin sliver of content rights, largely dominated by sports, a few TV programmes (think HBO or AMC) and perhaps movies (but Netflix has never bothered playing the Hollywood release window game) is the only differentiator.

Of course, this also means that if they're not playing nice, Virgin Media and Vodafone are going to be competing in a marketplace that is increasingly become crowded.

Holes in the streets may soon become holes in balance sheets.

Friday, September 18, 2015

Is This Admageddon ?

Sorry for the dreadful pun, but is this the week when we can say that the online advertising industry truly changed for the better ?

Let's look at the situation. If you want to build and run an online video or television service and want to support it through advertising you have two choices: you can build a YouTube channel and get a paltry revenue share on the ads they sell or you can build your own channel and use a tool like DfP or a platform such as AdX to serve your video and companion ads. As you probably know, YouTube, DfP and AdX are all owned by Google, so this is a virtual monopoly in markets like the UK despite the aggressive moves being made by Facebook, Yahoo!, AOL and others.

Of course other ad networks are available, but they will only engage at volume. And by volume I'm talking about millions of impressions.

So, if you're starting out with a video service, then everything is stacked against you, especially if you're a specialised channel. We're typically seeing revenues of under $4 per thousand viewers for pre-rolls on highly targeted video sites across Google properties. This just isn't commercial.

The reason for these low rates are that Google doesn't care about your content. It just cares about the eyeballs you are delivering. Online advertising is far more interested in blitzing users with the same ads over and over whatever site or channel they visit based on 'profiling', which is, at best, dubious.

So, what changed this week ? Well, Apple has enabled ad blockers on its Safari browser (but not on its own News service nor on the Facebook app on iOS incidentally). Initially this may seem to be very bad news for channel operators and publishers dependant on ad revenue, but I have a different opinion.

First of all, there were sites I actively avoid because the sheer weight of advertising has rendered them unusable, particularly on tablets. It's not just the serving of the ad, but the tens of different calls that are made before an ad is rendered. The appropriately named Peace is the number one iOS app on the Apple Store already for good reason - it works very well. Pages on some news sites are now loading ten times quicker for me and actually fit into the viewing area on the tablet.

And if you wondered why those pages were so slow even on your sooper duper fibre connection, this is why:
Each ad is going through a tortuous process before reaching you, the viewer.

So, getting rid of this jumbled is good news for the viewer or user and this should actually  improve engagement with the content.

A second reason to welcome the changes is that publishers will need to start looking beyond the easy sources of low revenue available from Google and start to become more proactive: consequently, this should have the result of opening up the advertising ecosystem to new and more interesting players.

Thirdly, the implementation of video ad tags using the VAST and VPLAY 'standards' have been technically unimpressive: they are horribly fiddle to implement and they really shouldn't be. So this is an opportunity to implement simpler alternatives.

And we shouldn't forget they guys paying for all of this. There is a mounting body of evidence that ads, and video ads in particular, are billed to advertisers even though they're never played (or played properly). The current system is hardly accountable in any meaningful way.

There is a very basic and very obvious solution to all of this: channels with small audiences should go native and serve up their own ads. It's time for a marketplace where they can set up specific profiles and advertisers can bid directly for very narrow audiences just as they would book traditional magazine advertising. Selling mountain bikes in Manchester, then book mountain bike and local channels and sites for your ads. I suspect the results would be better for the publisher, the advertiser and the viewer.

Wednesday, September 16, 2015

Are TV & Video Prodcos Knowing The Cost Of Everything & The Value Of Nothing ?

As the techies from the TV industry return from Amsterdam and their annual trade show, IBC, their colleagues responsible for trying to commercialise the programmes produced are gearing up to venture down to Cannes and the annual MIPCOM sales fair.

Nothing has really changed very much for decades. Well, apart from one thing: the format of the products that they produce and sell.

What used to reside in film cans, on one inch, BetaSP and digiBeta tapes is now a digital file.
This brings with it both problems and opportunities.

The main problem is how to store this vast amount of data.

The main opportunity is in opening up new markets for content in the form of clips, programmes, formats and channels or feeds.

These are not unrelated. However, the industry cannot see its way past the problem since the cost of storing master content in the cloud is significantly more expensive than that of storing tapes.

Or is it ?

The trouble is that the calculation is made based on the TV world we used to live in, not the one that we will inhabit.

Most production and post production companies employ a lot of people. Almost all other industries have seen a huge decrease in employees due to the introduction of technology, with massive productivity and cost savings. Everything from milk to car production.

Currently, most production companies are using a variety of strategies to deal with the issue of digital file storage:

Using an online/nearline/offline strategy using networks, NAS (network attached storage) and LTO (yes, the irony of it, the content ends up back on tape). But the cost of buying deploying and operating an LTO is considerable.

Buying cheap HDDs (hard disk drives) and hoping for the best: these consumer devices are mechanical and easily break, rendering a six figure shoot or production lost forever. The time to manage, index and duplicate them is considerable.

Here is a table that some might find controversial, but it levels the cost of redundancy and management alongside actual storage costs:

5 TB Video Storage Costs Over 2 Years
Base storage Cost
Redundant Storage Cost
Time Management Cost
Total Cost

Attached HDDs

Nearline Storage

Cloud Storage

But there are other factors at play. Here’s a crude analysis of the different features of these approaches:

5 TB Video Storage Comparison

Attached HDDs

Nearline Storage

Cloud Storage

Obviously, content held on HDDs or nearline has no accessibility to show and sell. On HDDs unless there is a very disciplined (and expensive in terms of manpower) approach to logging the content could be difficult to find, causing additional costs in the edit suite down the line.

You can add other benefits such as automated QA, automated technical metadata and full metadata management to the cloud proposition.

So the industry needs to take a long, hard look at how it operates and what it needs to do moving forward: I’m afraid that IBC and MIPCOM may becoming echoes of the past.