I attended the Telco 2.0 event last week so I took the title of this article from a quote by Hossein Moiin of T-Mobile and the MIT Communications Futures Programme. He accredits the quote to one of his peers at MIT, Dave Clark by the way, but whoever said it initially, it captures the two different parts of the Broadband Incentive Problem that is at the heart of the “How to make money from video” question.

The first article in this series showed how content is hot and how social networking is actually accelerating the speed with which something gets hot and concentrating the intensity of that heat. In network terms, this means bigger peaks and article 2 discussed how much these peaks could cost to build networks for. Article 3 discussed Traffic Management as an alternative to anarchic congestion and the all round poor user experience that these peaks could cause.

The internet will break if parasitic peer to peer applications are allowed to dominate the video content space. Recognising this, network service providers have introduced traffic controls – building their fortresses and preparing for war. Honest traders from other places in the galaxy can’t get through the Traffic Management force-fields that protect the ISPs World.

At the end of last week’s article I made a rash promise. I said that this week would “look at how to fix the problem”. I said I would answer these questions:

• What are the fundamental problems and what are the building blocks that would allow us to fix them?
• How can costs and revenues be aligned?
• How can networks be built to minimise the distance that large objects need to travel to reach their destination?
• How can the incentive to solve the problem be shared so that all parties stand to gain by acting positively?

I am going to use Hossein’s quote to break down what I write as I believe there are structural problems on both the technical and commercial side. In summary, I agree with half of what Hossein said: the problem is how to route the money, but I think the problem is also how to route the packets.

I’m also going to have to break the answer down into two posts because in each area there are options that deserve some consideration and this post is already way too big. This article deals with the Money: the range of commercial solutions that will help you make money from video. It is largely written for network service providers but it is vital background for anyone trying to deliver their content over those networks because they need to give operators an incentive to grow.

I will follow up tomorrow with a fifth post in the series on the Technical Solutions that minimise the distance that packets travel.

The Money
Access providers in a fixed price world suffer as usage grows. Network costs increase while access pricing declines because of savage price competition. “The saturation of markets with fixed rate pricing leaves no incentive to invest”, Moiin again.

The theory is to “sell value added services” on top of the basic access services and blah blah blah. It is true that there has been a lot of success cross selling internet security, voice and line rental, but ISPs have spectacularly failed to generate revenues from content services and these are now driving bandwidth growth. Access providers have found themselves with growing costs and declining revenues as the market saturates.

This is the first fundamental problem: there is a missing link between activity on the network and the amount people pay their network provider at the end of each month. Fixed prices made perfect sense to drive adoption but it cuts the service provider out of the value chain when users start to do more with the connections.

Capped pricing
We saw a shift in this direction in 2004 when BT introduced Capacity Based Charging at the wholesale level. BT’s move allowed them to align usage and revenue (before it was really an issue) and gave the ISPs the opportunity to introduce capped services at £17 where all you can eat was priced at £25. £17 was the tipping point and the market doubled in less than 2 years. There are some great highlights showing how much things have moved on as a result of these developments, published today by Ofcom.

Packages with bandwidth caps can be made cheaper, enabling you to offer attractive headline prices. Incremental tiers enable you to charge more for heavier users so the financials work but this is where I want to bring in the second fundamental problem: customers do not understand usage in Megabits per Second or even in Gigabytes and I don’t believe that they can be made to learn.

As humans we take in so much more information from moving images than we do from the same amount of time of spent listening to audio or reading text. In network terms, this fact can be quantified. 1 second of HD is 65 times more information than 1 second of video, but people can’t be expected to use this to work out how much bandwidth they are using. The problem is made significantly more complicated when you throw in evolving standards that cause changes in bandwidth required by applications to go both up (eg. higher resolution) and down (eg. better codecs).

A case study – me:

• I have a capped package but even as an industry insider I couldn’t tell you whether I was using 100 Megabytes or 100 Gigabytes a month without looking at my bandwidth stats.
• It turns out my usage has been 3.7, 1.7 and 1.3 GB in the last three months.
• My usage is declining because I am worried about the bandwidth bill shock from new applications. I stopped me playing around with Joost when I realised it was using 320 MB per hour, even when the application was minimised.

It takes me back to the days when you had to be careful how much dial up you used…

So how does capped pricing do addressing the fundamental problems? It does create a link between usage and revenues, but it does not deal with the second problem because it is all still based on Gigabyte usage.

Capped pricing by itself is almost certain to lead to the ISP playing the role of bandwidth utility, forever under price pressure and always fighting customer expectations that they should be able to do more.

Capping is however a cheap, low risk strategy to execute so it is likely that providers offering simple bandwidth only services will be part of the competitive landscape. This is a serious problem for those that want to be more ambitious as there is a risk that these low cost operators will mop up the significant numbers of low cost users who just want a bandwidth utility for occasional email and web surfing.

Who wins in this scenario? The ISP… perhaps for a while as floods of new users fall for the cheap price, until something cheaper comes along. The content owner… not likely: bill shock will see to that. The consumer…

Become a content provider / aggregator
The polar opposite of becoming a bandwidth utility, but I think we’ve seen this movie somewhere before: walled gardens or portals anyone? Alexa.com has a list of the top websites and you can see that the ISPs have not done very well – AOL, once the shining example of how an ISP could become a content company is way down in 56th place now. Orange are the highest service provider today at 16, Virgin Media at 25 and Tiscali at 27. Hardly a good start in the battle to take on Google...

There is a glimmer of hope for Sky, BT and one or two others they can play in this space by buying rights to premium content. Telenor’s VP Football, Pearse Connolly gave a great talk at Telco 2.0 about how they have built their IP TV service on live Premiership and Norwegian Football rights. The guy clearly loves his job, and why not? He told how they doubled prices to dampen demand for the network’s sake, only to find that customers didn’t care and they all still coughed up the new fees. Telenor became the first Telco history to discover that their customer was not price sensitive by putting their prices up instead of down.

The other thing that Pease described was how they are developing user generated content to build on the already very strong social networks around the clubs. 10% of the rights money is spent building the clubs’ new media capabilities around the core content available from Telenor. This cleverly creates a symbiotic relationship where the interests of the clubs, the supporters and the access providers are all aligned.

For sure this strategy addresses the fundamental problems but causes a whole new set of challenges. The lesson is that you need premium content and you need to spend hard cash to get it. Telenor spent 1 billion Norwegian knoner over 4 years (front loaded by the way 300, 300, 300, 100) and BT an almost identical sum of £84.3m. Both have broadcast partners with whom they are sharing the asset. TV2 for Telenor and BSkyB for BT.

The Telenor execution shows what you can do with internet technologies to make the experience better than is available from the plain broadcast platform. Buying rights is a reasonably successful bet but only for the very hot content. The ITV Digital case shows what happens when you get it wrong. Not one for the faint hearted…

Whitelist pricing
I am using this categorisation to capture the opportunity for service providers to exclude certain “certified applications” from a user’s bandwidth cap, charging it at a special rate or offering it “free” in an inclusive price.

An example of this would be H3G’s X-Series where you can get 1 GB of unlimited data, plus additional quotas for IM, Skype, Slingbox etc. These make the effective cost of using the whitelisted services significantly lower than they would be under the data plan.

The X-Series pricing plan has launched with two variables that allow them to align revenues and costs: (1) the price and (2) the usage limit that this price buys you. Not rocket science really (I hope you are not expecting that from me, if you are, try here instead). The important point is that it insulates H3G from the power user problems that there are on the fixed line internet.

Whitelisting known applications allows you to bring time back into your price plans. You can rate whitelisted services at a price per minute (instead of per GB) to make it easier for consumers to understand how much they are using. Because you know what the application is capable of, you can take into account bandwidth differences and factor these into prices on an application by application basis. Note how the X-Series does this with the Skype and Sling / Orb functionality – you can make up to 5,000 minutes of calls on Skype and watch up to 80 hours of Sling or Orb. These are deliberately big numbers so as not to discourage use of the applications at this early stage but critically, the numbers are much easier to understand than 1.8 GB of Skype and 9 GB of Sling content…

Of course the current £40-£45 is expensive and out of the budget of most, but then so are the usage limits. The current package is priced to catch today’s early adopter who is tomorrow’s power user and H3G will no doubt launch cheaper packages after it feels that it has had a fair run at skimming this early market. Whatever happens they can play with the packages until they find the demand sweet spot, retaining the essential link between more use, more cost and more revenue.

In the “How to make money from video context”, whitelisting requires partnerships with internet content providers allowing them premium access to your customer and allowing you an element of price and bandwidth control. What you get is an access pricing model that addresses the two fundamental problems: there is a link between usage and revenue in a price plan that customers understand because it is in a context that makes sense to them.

Blacklist pricing
Blacklist pricing says: we will give you a good price but you are not allowed to do X, Y or Z (because we know that if you don’t use that stuff, you won’t be able to use more than you pay for). This solves the fundamental problems too because usage is controlled to line up with revenues and people understand “thou shalt not” even if the purists don’t like it.

Consider T-Mobile’s Web’n’Walk service giving you 1 GB of unlimited data for £7.50 a month. Sounds much better than H3Gs doesn’t it, but you can’t use the service for VoIP, IM or Peer to Peer. This protects T-Mobile’s legacy revenues and its internet costbase by blacklisting applications that cannibalise existing business or that introduce a heavy load on the network.

Last July, we did some research into Web’n’Walk speculating that its launch might be the beginning of the end of the iPod because sideloading straight to the mobile phone was a damn sight easier than having the PC in the way. Another rash claim, you might think, but one that might be worth reading more about if this is of interest.

In the context of this article Blacklisting is a more open version of the Traffic Shaping that is becoming commonplace today. Blacklists can even be made time of day specific if you want flexibility. The positive side of blacklisting is that once you have established that certain use is banned on a cheap service, you can launch a premium one which allows an amount of the activity that you wish to control.

Advertising & Data Mining
Advertising already generates large revenues to fund broadcast video and the same will be true for online video. There is hope and fear in equal mix within the telecoms industry about the potential for advertising. The hope is that these advertising revenues will flow through to the service providers, providing them with a rich source of new revenues. The fear is that Google and the like will again grab the lion’s share of the money and that little if anything will find its way to the service provider.

For an in depth study, you could try Chris Barraclough’s report on the subject, of which the highlight for me was that the total global ad market is around one quarter of the total global telecoms market. Ie. even if it all goes online, it is not going to replace today’s access revenues.

The consensus in telecoms seems to be that this will be an incremental market but it will not pay for new networks to be built. Ask yourself, am I going to be able to generate enough ad revenue within a 2 hour HD DVD to cover the £2.10 delivery cost? If you think yes, the please enlighten me because I can’t see it being close.

That is not to say that Advertising revenues should be ignored. I believe that the intelligent use of the ISPs unique position in the value chain can enable them to generate reasonable sums. ISPs have a few key assets that they can exploit: they have the monthly billing relationship and they know the addresses and complete surfing habits of their users.

Marty Algire, VP of Radialpoint at Telco 2.0 highlighted that Google see the risk from Telcos, even if the telcos themselves do not see the opportunity: “because the access provider gathers information from the user in connection with the establishment of a billing relationship, [they] may be more effective than [Google] are in tailoring services and advertisements to the specific tastes of the user”.

Fat chance of that the way things are today, but it is not too late to change the way that service providers use the wealth of data that they collect from their customers. The problem is that the amount of data is daunting and processing it into something useful is an almighty pain. It is easier to ignore it as one of those problems that falls into the “too hard” bucket.

[WARNING: sales pitch] Before you do consign this to the “too hard bucket”, consider giving me a call as one of The IP Development Network’s speciality skills is the ability to clean and thread together vast quantities of data. We have typically done this to highlight cost leakages for our clients, but it is the same skill that would allow us to show you what data you have and how you could use that to help you exploit the advertising opportunity. [end WARNING]

Advertising does not address the fundamental problems, but it is an opportunity to leverage the service provider’s position and generate additional revenues. It may also be that the threat to the content industry can be made serious enough to gain concessions in other areas too.

Revenue Share for Premium Delivery
I believe that it is inevitable that consumer will look first to the content providers and aggregators when choosing their media. The value perception is stacked against the service provider – here on one side is a company offering a movie, on the other is a company offering a gigabyte. The result will be that people buy their media from that industry and they will increasingly expect that such charges include the distribution cost.

Imagine the postman knocks on your door every time you get sent a letter, a magazine or a DVD and demands that you pay him a surcharge. He won’t be a very popular chap will he?

No. People are used to sender pays and crucially, they are used to companies charging postage and packing on top of their goods. Face it, it’s a good time to get the money – when the customer has their wallet out and is paying for the bit they want. As long as you are not trying to charge too much on top of that (the content price has to be a significant multiple of the delivery price), people will pay for delivery.

Another feature of offline delivery that should be learned from is the way the market is segmented. Offline, you can choose how quickly you get goods, and based on how desperate you are, you can pay more to get them quicker. I would speculate that it probably harder for FedEx to have multiple service delivery tiers (how do you make sure that a second class letter takes two days, where a first class letter takes one? Do you have to leave it sitting in the corner for a day…?) But this inefficiency is worth it because it means that the top of the market will pay a premium price for what is almost exactly the same product. The alternative is the lowest common denominator again where everyone pays second class rates, even those that would happily pay for first class delivery.

The conclusion is that the internet delivery networks need to start learning from the way in which offline delivery networks have evolved and come up with a FedEx like differentiation for customers who are willing to pay more. Networks can continue to offer the slower, cheaper experience, backed up by the online equivalent of leaving a package sitting in the corner for a day: traffic shaping.

For this to work though, payment needs to be taken at the content’s point of sale, meaning that ISPs give up this element of the billing relationship – the charging for premium delivery coming under the content owner’s domain just as it does with offline delivery. Sceptical? Go back and ask yourself how different it would feel paying the postie when he arrives with your new DVD.

Also consider then that many online retailers offer a fully loaded “delivered” price, hiding entirely the cost of delivery in with the rest of the content service. How good would it be to have your £2.10 cost hidden in the price of the goods the customer values?!?

If you are looking for ways in which to address the fundamental problems and stop the service provider being seen as the bad guy, this could just about fit the bill. The link between usage and revenue comes from the premium fees and these are clearly priced as part of a package that can be understood by consumers.

The leap of faith is giving up the premium billing relationship (you still bill for access) and there are technical challenges integrating systems and assuring that premium delivery is fast and worth paying for. I suggest that all of these are worthwhile in order to make a clear association in the customer’s mind between buying the content and paying for the delivery.

Managed Wholesale Services model
This has the same end result as the revenue share approach, if successful. It may even deliver a greater slice of the pie, but the associated risk is that content owners will seek a more flexible offer elsewhere because the service provider is exerting too much control.

The model looks a bit like this: build a tiered network delivery capability with guaranteed delivery for content providers who buy your managed service. The new service includes elements that currently sit the other side of demarcation points.

At the network end, this means content servers, program scheduling applications, data analytics tools and perhaps even advert brokering and managed pay per view billing.
On the user end, you add additional elements that integrate with the network to offer Electronic Programme Guides (EPGs) and Digital Rights Management (DRM).
This is likely to require a standard set top box and the service provider truly needs to address the home networking issues to make the whole thing easy to set up and use. Not impossible with Homeplug technology, but if you expect your end users to work out how to configure wireless bridges all over the place, please, please think again…

The service provider in this model continues to sell retail access services to consumers, but adds around that the capability for managed wholesale customers. They buy a connection onto the network that allows them to relatively quickly and easily start offering services. This then starts to look like the transit providers in the early days of the internet where they sold both ends of the same pipe: once as access and once as hosting.

Telecom Italia described a model much like this at Telco 2.0 and it does seem balanced. It maintains the control over the users and exploits the strengths of the network provider while staying clear of markets in which today’s service providers are poorly placed like content production.

It could be a slam dunk for a content owner to buy this managed service – they would just need to make the content and leave the heavy lifting to Telecom Italia – and both parties could concentrate on scaling up very quickly. Vitally it has the potential to set an industry standard for content delivery in that market – anyone who wants to produce IP TV content can go to TI and plug in.

The difficulties lie in the execution. Every element has to be as close to best of breed as makes no difference: it is no good having a great network service if the content hosting and scheduling applications aren’t up to the job.

It may also be that content owners play a political game, which would be understandable given the potential for long term control that such a service would offer network providers. The content owners can play the service providers off against each other so it may be important to include the capability in the wholesale services offered to smaller ISPs too (in order to get close to that industry standard), diluting significantly the retail service providers competitive propositions.

Wholesale Managed Services, I believe, can address the fundamental financial problems facing content delivery today but you need to be in a monopoly position to make it work for you. The link between usage and revenue comes in this example from the managed service fees paid by the content companies rather than direct from consumers. The consumer still pays, but as part of the overall content or pay per view fees that initially go to the content house and then flow back to the network provider in the managed service fees.

Whether any monopoly is brave enough to state its position so clearly to regulators will be interesting to see. I’m not sure where TIs competitors fit in, but then we are back to the IP as a Natural Monopoly question again and whether it is worth sacrificing competition in a low value market to stimulate innovation in higher value services. There is a strong risk that misplaced intervention at the access level could enforce inefficiencies by insisting on multiple parallel investments by competitors in a market.

Summary
There is a lot to consider on the commercial side but all of them will only work alongside the technical suggestions that will be put forward in the next piece.

There will always be bandwidth utilities to compete with. If you want to humanise your restrictions better than “Gigabyte Cap”, you may consider packages that Blacklist certain applications to keep your costs under control.

Alongside these Blacklists, you can offer Whitelisted services from partners and there are ways to generate revenue shares and advertising from these deals that can help you fund the connection from sources other than access fees. This could all allow for low headline pricing of the basic utility with the difference coming in the back-door so to speak.

Is this all really going to allow you to pay for the £2.10 HD movie? No, but it will bring in additional income to complement the structural technical changes that I will go into tomorrow.

Other Articles in the Series

Part 1: The Online Video Market
Part 2: The cost of Online Video
Part 3: Traffic Management
Part 4: Routing the Money
Part 5: Routing the Packets
Summary Slides