We look at the future of mobile wholesale charging: What alternatives exist to traditional usage-based models? Could capacity-based charging improve return on network investment?
Wholesale Charging: Meeting the Needs of Mobile Data
By Michael Dargue, Eric Holzhauer and William Wadsworth
To date, inter-operator settlements in the mobile industry have typically been based on usage. However, market developments and industry changes are placing these traditional charging models under stress. In this viewpoint, we outline a vision for an alternative mobile wholesale charging model – one based not on usage but on capacity, proposing that it may offer a more efficient, flexible and cost-effective way to manage wholesale billing relationships.
Charging Structure Landscape
Inter-operator settlement is long-established in the telecoms sector. Initially driven by users calling off-net, and later through the arrival of mobile roaming, operators have needed to bill each other for network usage. By the late 1990s, some operators were seeing 30-50% of their revenues from voice interconnect fees. With the rise of MVNO in the 2000s, many operators and vendors had to expand their wholesale billing capabilities yet further, but for the most part chose to stay with traditional usage-based billing (UBB).
While it has its advantages, usage-based billing is only one of a range of charging models, which include:
- Flat: uses contractually negotiated payments regardless of usage – possibly adjusted at intervals to track usage levels.
- Usage-based: billing based directly on volume of units used – e.g. minutes of use, messaging volume, MB of data.
- Capacity-based: billing based on the peak demand for network access in terms of numbers of concurrent calls or data throughput.
- Hybrid: usage-based pricing, but offering tiered discounts for buying larger capacities.
- Bill and keep: based on an agreement between operators at relative parity to “call it even” and recover termination costs only from their own subscribers. Used for call termination in some countries such as the USA, but not applicable for MVNO or other wholesale agreements.
Historically, wholesale charging models have been only loosely related to actual network capacity, and thus only weakly linked to the capital investment needed in the network (see Figure 1).
Figure 1: Charging Models Overview
Limitations of Usage-Based Billing
Traditional usage-based billing presents a number of challenges for operators and MVNOs:
Offering unlimited retail plans is a dangerous gamble for a service provider with usage-based costs. MVNOs, for example, face a variable cost structure under UBB which leaves them financially exposed to heavy users; the risks are exacerbated with mobile data, with users able to quickly rack-up huge volumes. In part this also explains MNO reluctance to offer unlimited data plans for international roaming.
Unlimited plans carry risks for MNOs as well, but MNOs have greater control over their costs on-net: MNOs can choose when and where to invest as they balance service quality with cash flow. MNOs also typically benefit from longer experience, more extensive subscriber usage information and better understanding of long-term network economics.
Incentivising Efficient Network Use
MNOs have an economic incentive to optimise their own retail rates to encourage efficient utilisation of the network, spreading demand as evenly as possible through the day. Under usage-based wholesale pricing, MVNOs do not share these incentives. This is detrimental to both sides of the MVNO-MNO relationship: the MVNO does not have an ability to influence and support network capacity management, while the MNO’s network capacity can become needlessly strained if it has to bear MVNO traffic surges at peak times or in capacity-constrained cells.
Aligning Cost and Price through Capacity-Based Charging
Moving to a capacity-based charging system (see Figure 2) could offer a number of benefits for network operators. The economics of networks are such that marginal costs are not based directly on usage but rather on capacity: capital investment in new infrastructure is driven by the need to support peak demand. A capacity-based system would better tie revenues to costs, could encourage buyers (e.g. MVNOs) to provide network owners with more visibility into future capacity requirements, and will incentivise resellers to use existing network capacity more efficiently, by spreading demand more evenly over time.
Figure 2: Relationship between Throughput (Capacity Demand) and Usage
The MVNOs themselves could reap rewards from the new system as well, benefitting from lower effective rates for off-peak periods or less-busy cells. MVNOs will be empowered to become more innovative in proposition design, potentially emulating operators such as MTN, whose “MTN Zone” offers discounts of up to 100% based on time of day and location. M2M service providers are also well-placed to work with operators to avoid sharp peaks in demand. In particular, M2M could take advantage of dynamic alerts and charging systems which respond to fluctuations in network utilisation.
Economists have put forward strong arguments for capacity-based charging. For example, Kennet and Ralph (2007) present a detailed analysis of how, by closely tracking incremental costs, capacity-based charging offers: greater efficiency (in terms of long run incremental cost) than usage-based systems; greater efficiency than bill and keep; and an effective means of pricing wholesale interconnection services.
There are two primary variations of capacity-based pricing:
- Unmetered: the wholesale service is provided with a fixed capacity (e.g., a maximum number of concurrent calls or bandwidth) at a fixed price; it is up to the buyer to decide on the ratio of capacity to subscriber volumes, thus trading off cost-per-subscriber and quality-of-experience against each other.
- Metered: the wholesale service is provided on a variable capacity basis and charged based on peak capacity demand in the billing period; actual demand is metered and billed in arrears, e.g., monthly; typically contracting parties agree charges based on the 95th percentile of maximum throughput.
The merits of variations on capacity-based charging models are borne out by a variety of well-established use cases in other industries, and by the variations of the model which are seen in other areas of telecommunications (see Figure 3).
Figure 3: Charging Models Designed to Maximise Efficient Use of Capacity
There are industries where charging for usage makes sense – for example in the taxi industry, per-mile fees tie revenue closely to costs which are dominated by fuel expenditure. By contrast, the incremental cost of off-peak usage in a telecoms network is negligible.
2013: The Year of Changing Charging?
With the demand for mobile data set to rise further (13-fold over the next 4 years, accordingly to some analysts), network usage patterns will continue to evolve, becoming even more susceptible to large capacity fluctuations. The volumes and frequency of data sessions make usage-based settlement cumbersome and expensive, and a drain on OSS/BSS resources which could be better purposed to greater innovation in retail tariffs.
As the global mobile industry matures, and as worldwide mobile penetration nears saturation (currently 6.4 billion subscribers, 91% penetration), operators are shifting their strategic focus to cost management in their core products, with innovation in pricing strategies and value-added services. Capacity-based charging will help to facilitate both of these goals, while simplifying settlement processes (compared to usage-based) could lead to fewer errors and disputes.
The benefits of capacity-based billing will only become more important as interactions between operators develop: whether through the continued rise of MVNO (particularly in emerging markets) or the forecast growth in roaming. On a more long-term basis, capacity-based charging may have a role to play in facilitating a shift in network ownership mind-set: in a more interconnected and global world that expects pervasive network access, it will be prohibitively expensive for an MNO to do everything itself, while end users may be largely ambivalent as to which network is carrying their data.
Key Implementation Considerations
Despite the advantages of capacity-based charging systems, operators face a number of challenges in adoption and implementation. Updates may be needed both to billing and charging systems, and also to network components such as the HLR, VLR, and interconnect gateways.
Operators may face legal or regulatory barriers: most current regulations governing prices for roaming, interconnection, and wholesale are written in usage-based terms. Restructuring these to accommodate capacity-based charging would require careful consideration.
Cartesian has supported clients in a wide range of projects concerning interconnection and wholesale services. From strategy formulation to negotiation support, we assist operators, service providers, vendors and investors achieve their strategic objectives. Example focus areas include:
- Wholesale strategy formulation
- Wholesale pricing and proposition design
- Network economics and cost modelling
- Access and interconnection negotiations
 Kennet, M.D., Ralph, E.K. (2007), “Efficient Interconnection Charges and Capacity-Based Pricing”, International Economics and Economics Policy, Vol. 4, 135-158.
 Cisco Visual Networking Index Global Mobile Data Traffic Forecast , 2012-2013
 Ericsson Mobility Report, November 2012