[kictanet] Day 2 Internet Study- International & Domestic Bandwidth Usage

Michuki Mwangi michuki at kenic.or.ke
Wed Apr 25 13:44:39 EAT 2007


Hi Walu,


John Walubengo wrote:
> The Internet Study says that of the about 1G International
> International Bandwidth only 10% is outbound while 90% is
> inbound.  What this mean is that kenya is a net IMPORTER of
> internet content.  In internet economic terms we are
> therefore consumers rather than generators of internet
> content - which is a bad thing.  Value is always retained
> by those who generate rather than just consume internet
> content i.e. the exporters of internet content.
>

Less than 50% of the International content is relevant to us yet it
makes up 90% of our inbound traffic. That means that there is a
potential of more than 50% percent of local content yet to be realised
and monetized.


>What can be done to get that to happen? 

IMHO, i see the following as key catalysts to the growth of local
content in order to change the status quo. Before we think of exporting
content, we need to have sufficient & relevant local content to the users.

1) ISPs & Service providers need to introduce creative products that
focus more on local content. For instance an unlimited Internet
connection for kshs 1,000 giving one unlimited local access per month.
By local access means - can access any locally hosted website and send
emails off a local web-based mail system. With affordability but lack of
sufficient local content will spur;

	o creation of local user generated content
	o Better information on local corporate websites
	o Introduction of local online Gaming & entertainment
	o Introduction of online Services i.e banking, shopping, e-gov

2) Looking at the graphs availed (KIXP stats). The stats (picks and
drops) are replayed across the International circuits at night and
weekends for all operators. Thus the service providers need to monetize
their services in Megabit per Second. This will mean that they can offer
lower traffic rates i.e kshs 1,000 per month for unlimited Internet from
7pm to 7am and full weekends and national holidays.

Looking at points 1 & 2 above it would be feasible (please note the
values are just hypothetical) to have unlimited local and international
night access for Kshs 2,000 per month for x Kbps of capacity.

As the enterprising citizens of Kenya have found ways of monetizing
SMS's - the same would apply to web content, email etc. Of course this
would spawn off issues of net-neutrality but then we would be more
conversant with the issues.


3) As a first step towards e-learning, if the academic sector would
(initially the colleges and universities) have all student assignments,
essays, reports etc submited online with the internally graded results
being posted & accessed online. In essence it will trigger a demand for
both computing and Internet resources. Being local content the above
products 1 & 2 can serve the day-scholars and the boarders probably
paying a flat rate of Kshs 500 per month for unlimited access to the
university online resources and local Internet. As a result this will spur;
	o Academic/Scientific research content & traffic
	o enhance e-learning products and services
	o User generated content
	o Local social online networks.


In summary, the infrastructure providers will keep investing as long as
there's a business case. So far, there is immense Internet
infrastructure capacity on the ground and more being laid out. But quite
honestly its highly underutilized. The situation may not change
significantly if special focus and emphasis is given to availability of
local and RELEVANT content. Kenya is amongst the only two countries in
Africa that have the core elements of the Internet locally namely the
DNS Root Servers (F and J), the main TLD Root Servers (.KE, .COM and
.NET) and a Gigabit capacity Exchange point. We need to take advantage
of these resources and infrastructure sooner rather than later.


> 1 day discussion on this theme, the floor is open...  
> 

Pole this comes more later than the set time.

my 2 cents,

-- 
Michuki Mwangi
KENIC




More information about the KICTANet mailing list