The calculations - and opportunities - that come with providing internet

In Summary

• Fortunately, since not all users will be online at the same time, the network operator can deploy a lower-capacity core network, of perhaps 10Gbps instead of 100Gbps.
• A network may have been costed and built for one service, but then it can host and deliver other services at little or no extra cost.
• The same network gets even better when that text that costs close to nothing to transmit has monetary content, otherwise known as mobile money text.

By JOHN WALUBENGO
More by this Author

In a previous blog, we discussed the fact that Internet services are proving to be unaffordable in Kenya. Today we try to understand how operators price their services based on the bottom-up cost allocation method.

In most business enterprises, the price of a product is pegged on the cost of producing it. Basically, you sum up the costs of the raw material, labour and power, among others, and mark this up by, say 10 per cent, to arrive at your selling price.

The same thing happens with telecommunications networks. If the environment is completely new with no existing infrastructure, the operators simply estimate demand for their communication service in order to establish the size or dimension of the network required to deliver the service at a particular level of quality. This is known as the bottom-up costing method.

For example, a company looking to provide Turkana County with Internet services may estimate the county’s market potential as being 100,000 users (10 per cent of the total population as per the 2009 Census) and then proceed to build a network that can support this many users.  

Assuming each user will need a 1Mbps (megabits per second) quality of Internet, the operator will need to deploy a network with a core capacity of 100Gbps (Gigabits per second) or 1Mbps each for the 100,000 users.

This can be a very expensive network to build, considering that 100Gbps is the current capacity of TEAMS, a submarine cable built at a cost of $100 million.

Fortunately, since not all users will be online at the same time, the network operator can deploy a lower-capacity core network, of perhaps 10Gbps instead of 100Gbps. This value is based on the assumption, for simplicity, that only 10 per cent of the users will be online concurrently.

Having dimensioned the network, the operator would then work out the cost of equipment, maintenance, operations and administering the network for a period of 10 to15 years. With a working figure and a business plan, the operator would then approach financial agencies and investors for money in order to roll out the network.

Let us assume the network for Turkana will require $10 million to deploy and run over the next 10 years. Ignoring the time value of money, the network will require $1 million to run and maintain each year.

Dividing this amount over the projected 100,000 users implies that each user should pay at least $10 per year for 1MB per second of Internet — just for the operator to break even. However, just like a good business enterprise with shareholder value in mind, the operator can mark this up by 10 or even 100 per cent depending on disposable incomes, or competitive and regulatory pressure.

But the network is an interesting asset. It may have been costed and built for one service, but then it can host and deliver other services at little or no extra cost. For example, the network built for voice communication can send millions of text messages with close to zero incremental costs.

Yet again, as a good business enterprise, and for the sake of the shareholder, the operator will only charge you between Sh5 and Sh10 for sending an SMS that costs them zero shillings to transmit.

The same network gets even better when that text that costs close to nothing to transmit has monetary content, otherwise known as mobile money text. Suddenly the zero-cost text message can be charged based on the value of money contained in its content.

Finally, as if to turn all the economics upside down, the operator can decide to provide free Internet access for particular websites, the idea being that the free access will increase traffic to the website and allow the operator to obtain revenue based on items purchased rather than the cost of transmission.

Essentially, the network costs and product pricing are decoupled in the new economy. One does not necessarily depend on the other. It’s really about the shareholder and what they feel is a fair amount of return on their investments.

Mr Walubengo is a lecturer at the Multimedia University of Kenya's Faculty of Computing and IT. Twitter:@jwalu email: jwalubengo@mmu.ac.ke