Highdeal CTO on cloud computing and new broadband billing models

Fergus O'Reilly

Earlier this month, Highdeal, a telecom billing software vendor spun out of France Telecom’s labs in 2000, was acquired by German software giant SAP, giving even greater global reach to privately held Highdeal, which already boasts more than 200 customers, including three of the top four Canadian cable operators. Highdeal’s chief technology officer, Fergus O’Reilly, spoke with Telephony’s Ed Gubbins about how cloud computing will be billed and how broadband billing models will evolve.

On cloud computing billing: We have a number of customers providing software-as-a-service and network-as-a-service -- operators playing the open-API game and charging developers for access to their APIs, for sending SMS, getting a user’s or subscriber’s identity or location – that type of thing. We already support that for a large European carrier we can’t name.

There’s dynamicity in the pricing model toward the developers simply for the APIs; that tends to be usage-based. Sometimes it’s a prepaid relationship, sometimes it’s a capped, post-paid relationship. There’s a lot of online charging required. Behind that, you get into quite sophisticated models very quickly because developers can initially develop on, say, Google or Amazon’s infrastructure, but then next they want to access the customers of that infrastructure provider. People developing on Force.com (part of Salesforce) are getting [their apps] sold through Salesforce’s App Exchange. If you’re developing on an operator’s infrastructure, you want to get sold through that operator’s app store that they’re all talking about launching. If you’re developing on the iPhone, you want to get sold through Apple’s iTunes store. So that app store gets revenue from consumers and shares it back with developers. So the developers may be paying for using the API on the platform, but they’re also receiving money back in terms of revenue-sharing when apps are sold through the app store and billed through the operator. You already have a bidirectional revenue model happening there, and we’re also seeing people in the platform business talking about becoming a central counter party for many of these financial transactions because developers will also be using services provided by other developers on the same or different platforms. So a level of inter-developer settlement needs to go on, and it’s bilateral in terms of the money flow. That’s a position that a number of actors in the space are looking to see whether they can move into. It’s obviously a key position because you’re sitting in the middle of that money flow and managing these billing and settlement relationships. The infrastructure required for that is quite significant and in many cases exists within the telecom operator community today, but it’s piecemeal, so I have a system for customer charging and pricing, I have a different system for my carrier interconnect, a different system for my hosted MVNE model – all sorts of disjointed systems. Whereas this cloud-computing-type business requires a [single] system where I can configure these business models in a very interlinked way because they are entirely interlinked – it’s a percentage of a percentage of a percentage, with tiering and all that mixed in. You need to manage those relationships very tightly, from a contractual perspective, a settlement perspective and a compliance perspective. With Highdeal from a rating and charging perspective and SAP from a broader collections and compliance perspective, we’ll have a very good end-to-end solution around that.

On moving beyond flat-rate broadband: Operators are realizing that the flat-rate model we had for broadband is no longer tenable. It’s hard to roll out [usage-based models] when subscribers don’t know how much a gigabyte is or what the term bandwidth means. Some [providers] have done better than others. In the Canadian market, for example, it’s getting to be accepted. Rogers has done a good job informing customers about their usage and charging them for overage with cap-and-overage-type schemes. In the US, it’s been a little more difficult. Time Warner Cable let slip that they were doing something and got negative press for it. It became difficult for them to roll that out -- one step forward, two steps back. But overall throughout the market, pretty much everyone is equipping themselves with the policy management systems they need to measure and qualify bandwidth usage. The flat-rate model for broadband will change, and we will pay depending on usage, whether that’s measured in [quality of service], absolute bandwidth or a number of those factors.

On mobile broadband models: It was ironic that all the carriers in most of the developed economies went to a flat-rate 3G style of broadband across the various networks. Pretty much everywhere in Europe, you can get flat-rate wireless broadband for about 10 euros. These plans are contractually capped, but in reality, many operators are not stopping you when you reach a certain cap. We’ve heard some people talking about moving to prepaid models for broadband. It could be the whole model starts to collapse; if one provider decides on a pure prepaid model, that might be a competitive threat to others. It could change rapidly or take a long time.
On getting users comfortable with usage-based pricing: When you have an always-on broadband connection, you don’t think about how big an email is as opposed to a movie file download. When it starts to impact you is when you talk about time and how long [things] take to download. People have a notion of time: You click the streaming movie link and see the bar moving across your YouTube. You can translate that into a pricing model. The turbo-boost-type mechanism [may become] something you turn on or off, some control-panel link to your broadband connection, a button on the device or something that is auto-sensed by the network operator. I’ve seen some examples where [providers] say, ‘We’ve noticed you’re about to try and upload a video file to a video-sharing site or you’re about to upload your holiday snaps to a photo-sharing site. Do you want a bandwidth boost for that? That will cost you $X and reduce your time to do that by such an amount.’ There are various ways to roll that out that directly talk to consumers in terms of the time it takes to get something done. Those are relatively easy to sell. The other good consumer-friendly example is time-shifting. An online backup service where you’re doing a backup not to some local drive on your network or connected to your home PC but across the network to a Mozy.com (part of EMC). People can understand those types of services relatively easily. It’s probably cheaper for you if you schedule that backup to be done in the evening or at night when bandwidth usage is less, so it’s easier for the operator to accept that carriage. In many electricity schemes, you have a cheaper night rate than day rate but a time-based rate in terms of weekend or peak and off-peak periods; it’s something people are familiar with and doesn’t require them to count bits.

It’s tougher to sell cap-and-overage schemes. Unfortunately many of the charging systems operators have in place are relatively simplistic. And moving to these more sophisticated schemes -- time-shifting and proposing a bandwidth boost -- many times the blocking factor is, ‘Well, I don’t know how to do that.’ So we propose a very flexible charging system that makes that easy so you can have these dynamic business models that will make more sense for the consumer.

On quickly launching new services: There’s a lot of frustration among particularly larger telecom operators -- weighed down with legacy [networks] -- with the fact that the cost of billing has been skyrocketing, going from something like 1.5% of revenue back in 2005 to almost doubling over the next few years. It’s a drag on profits. One of the major impediments is the fact that most operators we speak to are dealing with rollout times for new pricing plans that are measured in weeks -- sometimes months, if it’s a moderate or high level of complexity of change. That’s due to the nature of the industry. The Highdeal approach is all about configurable business processes and elements and getting the vendor out of the loop. We have one very dynamic customer in Eastern Europe that has a three-day turnaround from ideation among their marketing teams saying, ‘I want to have this new prepaid offer that does XYZ,’ to implementing that on day two and rolling it out on day three. They do that on a weekly basis. They’re not like an incumbent operator often is, where you can only roll out pricing changes every two months in a very particular time window and you have to prepare four weeks in advance. This is literally having an idea on Monday and putting it out in the market on Wednesday. That type of rapid change is only possible if you have very flexible platforms that don’t require you to get coders or developers involved. It’s a different approach to software architectures. We’ve built our reputation on it.