Guest post by Graham Finnie, chief analyst, Heavy Reading

All Heavy Reading’s research around policy management is pointing in one clear direction for 2012: there will be a greater need for policy to handle chargeable events.

We’ve been predicting the policy/charging trend for several years, because our extensive survey work with operators has been telling us that using policy to monetize more service elements is now a key objective of telcos—mobile operators in particular. But there are some major pitfalls to be avoided if policy is to realize its undoubted potential here.

Our major global surveys in 2010 and 2011 identified two contrasting trends. First, there’s a broad expectation that policy will be used to create more complex, personalized, value-driven service packages, and those packages will use policy servers to determine when to charge, and for what. Second, operators see this as a challenging and difficult area: in both 2010 and 2011, the number one obstacle to deploying policy management was—you guessed it— “integration of policy and charging”.

In order to probe these issues more deeply, we followed up our broad surveys of policy with a specific survey among the major Tier 1 network operators on policy and charging, conducted in November 2011, the detailed results of which can be found here.

The new survey turned up some fascinating new data. In the first place it showed—as we suspected—that the challenges in integrating policy servers with online and offline charging or billing systems often lie with the charging systems themselves, rather than with the policy servers.  Charging systems are often inflexible: it’s time-consuming and costly to change things or to add new pricing elements, and operators are extremely wary about making changes that might result in revenue leakage or over-charging. The charging industry is also quite fragmented; and many of its key players earn much of their revenue from professional services work, adapting their systems to customer requirements.

Hence just over two-thirds of our Tier 1 respondents reported difficulties adding  new use cases to charging systems, and almost as many said that high costs due to customization and professional services was a challenge. Moreover, when we asked about the key barriers to closer integration of policy and charging, the number one problem, they told us, was “complexity of legacy charging and billing systems.”

One consequence of that is that many operators intend to use the PCRF itself to handle some tasks here, especially simpler tasks like quota management. That clearly makes sense if the rule is, for instance, “Don’t count Facebook traffic against this customer’s monthly data allowance.”  Nearly 70% of survey respondents said they would use the PCRF to handle some charging, rating or quota-related functions.

At the same time, most telcos recognize that the PCRF can’t handle everything, and deeper integration with charging systems is inevitable: over 90% of respondents are planning greater integration, we found.

That doesn’t mean, however, that telcos are looking for pre-integrated solutions. When we asked what options their companies would consider for linking policy with charging and billing in future, over half said they planned deeper integration of their chosen policy platform with existing charging and billing systems, whereas only one in five intended to use a pre-integrated solution of some kind.

At the same time, there was plenty of evidence of uncertainty about the way ahead. Many respondents, for instance, said they didn’t have a clear view yet on how to handle integration. One specific area of uncertainty is the new Sy interface for integration of PCRF and OCS or OFCS:  respondents gave it a cautious thumb-up, but many said it was too early to say what its impact would be.

For policy suppliers, there are some clear lessons here for the future. First they need to demonstrate that they can interoperate policy servers with third-party charging systems—even if they sell their own charging systems. More often than not, the charging and policy systems will be from different vendors—and experience in connecting the two will count for a lot in these circumstances. Second, they need to assuage fears that changing or adding policies that refer to charging systems will be too costly and take too long. Flexibility and the demonstrable ability to meet this challenge will count for a lot with customers.

Third, they need to work together with customers inside standards organizations—especially 3GPP—to fill in any remaining holes in the standards that may be retarding progress. The new Sy interface is a step forward, but its value needs to be demonstrated in practice, on the ground, and more work may yet be needed.

Most of all, perhaps, our research suggests that, if these issues are resolved, the future, for both policy suppliers and their customers—could be bright indeed, with policy management moving right to the heart of network operator strategy, and playing a key role in meeting the biggest challenge of all: how to build better service packages for customers.

About Graham

Finnie has been researching telecommunications for more than 20 years, formerly as a journalist and latterly as an analyst and consultant. Since joining Heavy Reading in September 2004, following a ten-year tenure at the Yankee Group, Finnie has been responsible for a wide range of research, focusing primarily on next-generation broadband services and IMS. He became Chief Analyst of Heavy Reading in February 2007.

Finnie is based in the U.K. and can be reached at Finnie@heavyreading.com.

By Matt Hatton, Director of Machina Research

Mobile telephony is probably the world’s most successful technology: there are in excess of 5 billion mobile connections worldwide, representing an unprecedented level of technology penetration. To date, MNOs worldwide have built successful businesses based on selling voice, SMS and data connectivity to individual handset users. Recently, business dynamics have changed slightly with the introduction of mobile broadband and mobile content data services. On the whole, MNOs have coped reasonably well with the arrival of these services although it’s not all been plain sailing as illustrated by continuing fears about exponential data traffic growth and concerns about being relegated to a bit-pipe role. One thing that has helped MNOs is that the underlying business logic is broadly the same: sell a device to a person which they use to access services for which they pay. MNOs may have needed to do some work on the network and create a few product management teams for the new services but it hasn’t required a fundamental shift in how they do business. It is a moot point whether they will cope as successfully with the arrival of machine-to-machine (M2M).

As we set out in the table below, in almost every conceivable way, M2M is different from the services MNOs provide today. The most obvious change, from the perspective of industry-watchers is that expectations for traffic, ARPU and revenue are completely different. This has some implications for how MNOs manage the cost of serving those customers. They must keep it as low as possible if they are to be able to cope with ARPU of less than EUR0.5/month. Also, a new consideration within the M2M market is that data volumes per connection are typically sufficiently low that installation costs are often a more significant financial consideration than on-going data transmission costs. MNOs must adapt their tariffing to reflect this dynamic.
Other immediate differences stem from the fact that M2M communication is typically a component of a wider offering, rather than a service in itself. As a result there is often no active end-user. This has implications for swapping providers, complaint handling and device management. M2M is an enabler, and the more transparent the M2M component of an overall service is, the better. In many cases the end user may not even be aware that the device is connected at all.

Furthermore, M2M connectivity is often mission-critical. In many cases customers are entrusting a key part of their business to telcos. Examples include smart metering, insurance tracking devices for cars, a range of fleet management telemetry services and, of course, mobile connected medical devices. As a result customers will have very different expectations over quality-of-service and service level agreements compared to voice and data services where best effort was often enough. Conversely, latency is often not an issue with M2M connections: devices are often connected via M2M with a view to maintaining a certain level of timeliness of information, but without a requirement for real-time information. For instance, smart meters may take meter readings at quarter-hourly intervals, but there may be no urgency in when they are delivered to the utility.

MNOs must also revise their channels and sales strategies. The sale of M2M connectivity by MNOs is often B2B2C: an MNO’s M2M connectivity solution must be integrated into a product which is then provided to a consumer. As a result MNOs must build completely new channel arrangements including identifying sectors they should address via direct and indirect channels. They must also secure sales in an aggressively competitive B2B environment while at the same time delivering a solution that is sufficiently polished and intuitive for a consumer market.

Matt Hatton is a Director of Machina Research, the world’s leading advisors on M2M and mobile broadband strategy. This blogpost is based on an article that Machina Research wrote for the the Machine-to-Machine Insight Report in the June/July issue of Mobile Europe magazine. Matt also maintains his own blog: The Wireless Noodle