Table of Contents
Assessment of the “Free Internet Access” Project on the
Internet Market in Egypt: System Dynamics View
Fatma H.M.Abdel-Hafez
Core Network Optimization Lead Engineer, Vodafone Egypt
MBA Graduate, Maastricht School of Management, The Netherlands
Fatma.Hany@ vodafone.com
hafez@intouch.com
Khaled Wahba
Assistant Professor, Systems and Biomedical Engineering Department
Faculty of Engineering, Cairo University, Egypt
Academic Advisor, Regional IT Institute, Cairo, Egypt
Tel: +2 02 737 6006, Fax: +2 02 739 1380
Khaled.wahba@ riti.org
ABSTRACT
In January 2002 a major step was taken by the ministry of communication and Information
Technology in Egypt towards increasing the penetration of the Internet through the launch
of the “Free Intemet’ project. New Rules and regulations were imposed on the ISPs of
which; sharing their revenue with TE from the Intemet calls instead of the subscription fee.
The research revealed that telephone and computer penetration has a positive effect on the
penetration of the Internet in Egypt. Local content is another factor attracting the users, but
e-commerce is not yet mature in Egypt, so it does not actually have a high effect. Although
tariff is one of the factors limiting the number of users in Egypt, it is not the main factor.
The research also revealed that competition between ISPs changed from a pricing
competition to a competition over content and differentiation. Data gathered from
interviews as well as secondary data were fed to a system dynamics model, in order to be
able to predict the future of the Internet market in Egypt in the coming years.
INTRODUCTION
In year 2001, the Ministry of Communications and Information Technology (MCIT) in
Egypt anounced a plan to increase the number of Intemet users through transforming the
Intemet access to a “Free” one. Rules and regulations for the Intemet market were
completely changed imposing additional responsibilities on the different ISPs such as
moving their equipment to TE premises, applying for one of the three new ISP licenses; A,
B or C and finally having their revenue shared with TE from the Intemet calls instead of
the subscription fee relying on the concept of retum sharing between Telecom Egypt (TE)
and different Intemet Service Providers (ISPs). The percentage agreed on was 70% to the
ISPs and the other 30% to TE.
From even before its launch, The announcement made a lot of turbulence in the Intemet
market in Egypt both on the ISPs level and the users level; new users refrained from
subscribing to different ISPs waiting for the Free model to be implemented soon, as for
ISPs, they were considering the new model and its implications on their business, and the
new arrangements required from them such as moving their equipments to different TE
premises.
Even after implementation, controversy didn’t end; while the project supporters believe it
did cause an increase in Intemet penetration, those who object to it believe it has caused a
deterioration in both quality and service, and that although the number of users may have
actually increased, the content still revolves around entertainment and does not include any
valuable information that will leverage the capabilities of the Egyptian user. Another
important issue was the effect of the project on the ISPs business, especially after being
denied their subscription fee, which was their only source of revenue as well as the huge
costs they incurred due to moving their equipment to different TE switches in different
parts of Egypt.
The project was actually launched on January 2002. The Intemet Market seemed to be
hysterical; many ISPs closed down, others consolidated with each other and a lot of new
ISPs started their new business. Competition increased; commercials on TV, in the streets,
in magazines dramatically increased but did the number of Egyptian Intemet users really
increase? Did the MCIT achieve its goal?
As an Egyptian user of the Intemet, watching what is going on in the market and facing
quality and service problems in the new Intemet model, the researchers want to analyze
and study the new Intemet model dynamics and its effect on the Intemet market.
PROBLEM DEFINITION
Signs and Symptoms diagnosed at the Internet Market in Egypt after the launch of the
“Free Intemet” Project suggest that the Project is responsible for the deterioration of both
quality and service and the closure of many Egyptian ISPs. There is also a doubt that this
project achieved its main objective, which was to increase the number of Intemet users in
Egypt.
STUDY OBJECTIVE
The restructuring of the Intemet service industry through the launch of the “Free Intemet”
Model in Egypt is a very controversial issue; some of the ISPs are furious and believe that
the ministry deceived them and caused them a lot of losses while others (ISPs) believe that
their business flourished more and more and increased their opportunities in attracting
more users.
The Purpose of this Study is to assess and analyze the impact of the “Free Intemet” Project
on the Intemet Market in Egypt by using quantitative research and support it by the system
dynamics concept.
STATE OF THE ART
INTERNET EVOLUTION IN EGYPT: An Overview
Intemet in Egypt started exclusively for academics and researchers within their academies
and universities starting October 1993 when the Egyptian universities were linked for the
first time to the European Academic and Research Network (EARN). The government,
since then, always had the ambition to increase the number of Intemet users in Egypt and
this was translated through many steps; in December 1995, a decision was taken by the
chairman of Telecom Egypt allowing commercial Intemet services by establishing Internet
gateways to twelve Intemet Service Providers (ISPs) to begin operating. In the early 2001,
the Ministry of Communications and Information Technology announced a plan to
implement the “Free Intemet” model in Egypt in order to increase the penetration of the
Internet into Egyptian homes.
“FREE INTERNET” World-Wide Practice
“Free Internet’ in Europe: the revenue source is not anymore the direct one which is a
subscription fee but the sources of revenue are indirect such as subscription to other
services such as long distance calls with a specific company, permanent advertising, and
minimum buying from ecommerce sites.
The advent of the Intemet without subscription or “Free Internet” caused many ISPs and
long-distance operators to cooperate and sometimes consolidate together. In the majority
of the European countries applying this model, the principle ISPs were bought by long-
distance operators. Little independent ISPs survived except the ones with European
dimension; serving many European countries at once. (Buisson, Sergent & Frattaroli,
2000). In Europe “Free Intemet” has had some recent successes. In the UK, “Free Intemet”
providers receive a portion of the call revenue paid by users to the telecom company. This
revenue sharing has given “Free Intemet” providers a leg up, but the advantage may be
short lived. Analysts say that by 2002 the UK and the fledging European “Free Intemet”
market will have collapsed as call interconnection revenues fall by around 30%. At the
same time, competing paid services have put pressure on the telecom companies, and
industry analysts expect them to soon reclaim their market, (Ensoport Intemetworks,
2001).
“Free Internet’ in Latin America: “Free Intemet’ service in Latin America is
concentrating on advertising revenues and a very small portion of this money is going to
Intemet marketing channels. This is in part because the number of Intemet subscribers is
low compared with more traditional media such as television and print publications. In
Latin America the free providers have not made deals with the local telecom companies, so
they do not have this extra revenue boost. And analysts report that the telecom companies
are unlikely to volunteer a share of the revenue.
In Brazil, when the concept of free ISPs made an impact on Intemet some of the bigger
for-pay ISPs engaged in a sort of restricted, but free, net access. They all failed, and the
companies closed the free access but maintained their paid subscription services. Free ISPs
funded by advertising, have proven to be unrealistic business models. These Free
connections in Brazil are always “dial-up” services that require a modem and a phone line,
but connections are very slow and people always complain about that. In addition to this,
phone costs are the most expensive issue for the Intemet in Brazil; people may have money
to buy a computer and even a subscription with a paid ISP, but they can’t afford to pay the
monthly costs of the phone line if they use the Intemet for many hours every day. That’s
why Free-connection services in Brazil started to disappear last year. But Free ISPs do not
rely solely on ad revenue from their free-access business. Besides minor paid services,
their revenue also comes from different sources such as wireless broad content portal,
online pizza delivery service, electronic boutique and high-speed connection services,
etc... (Balancing Act News Update, 2001). The case is the same for some Free ISPs
working in Mexico and Argentina; they too canceled their free access.
“Free Internet” in the United States of America: The model, since it emerged in USA in
1996, has proven insufficient. Several high profiles free ISPs have failed and survivors in
the market are consolidating and looking for other revenue sources, (Ensoport
Intemetworks, 2001). There are two types of “Free Intemet” service providers in the USA;
those that use advertising to pay for the service and those without advertisements. The
most common free ISPs use advertisements to generate revenue. They work on the same
principle as broadcast television stations. Inteme advertising can take different forms. The
most common advertisement techniques are banner ads and pop-up windows. The second
type of free ISP is rarer. These services will generally be offered by companies as a benefit
to their consumers such as the one offered by American Express Online. Only American
Express customers can use the service. Sometimes customers can’t access the service
except after buying with a certain amount of money, an example of that type is the “Free
Internet” offered by K-Mart the famous retailer, (Sloboda, 2001). Dealing with Free ISPs
in USA is not very safe; some factors must be taken into consideration. Some of these
factors are the quality of Technical Support, which sometimes lacks the same level of
quality of customer service in traditional ISPs. Another factor is the speed, which is
sometimes reduced because of the banner ads, which require more speed because the
amount of data streaming into the user’s computer takes up bandwidth and may slow down
other data flowing into the computer, (Sloboda, 2001).
Conclusion: although the model implemented in each country is some how different from
that implemented in another, there are some commonalities among them, such as:
* The source of Revenue is not anymore the subscription fee.
* Other sources of revenue are exploited by the ISPs such as; subscription to other services
like long distance calls with a specific company, advertising through banner ads and pop-
up screens, minimum buying amounts from e-commerce sites, high speed connection
services and/or revenue sharing with the telecom company. Not all Free ISPs in each
country apply all these services, some may apply all and some may apply a combination
of two or three of them.
¢ Problems associated with the new model are numerous, among which are the lack of
good customer support since ISPs don’t know anymore who are their customers, the
degradation of the quality and speed because of the huge amount of banner ads
associated with the new model and also facing under-capacity because of the difficulty in
assuming the number of users which will use a specific “Free Intemet” number.
¢In most of the countries that previously applied the “Free Intemet” access model, ISPs
either tumed to the fee model after facing a loss or consolidated with other ISPs in order
to be able to face the problems of the free model and that too didn’t save them from
losing.
¢ Analysts in most of the countries applying the model expected the model to collapse, and
by year 2003, almost all ISPs applying the model in different countries abandoned it.
The “Free Internet” Model in Egypt
The “Free Intemet” access model in Egypt is based on revenues from advertisement, in
addition to a share of the revenue collected by Telecom Egypt. As agreed upon with TE,
70% of the revenue goes to the ISP and the other 30% goes to TE.
“Free Intemet” services attract subscribers in large numbers, but it is difficult keeping and
leveraging them. Since its launch in Egypt starting year 2002, some of the Intemet users
complain of the quality and service deterioration while others are happy with the diversity
of choices among different ISPs and they think that the new project gives them more
chance to select the best quality and speed regardless of the number they dial which
represents a certain ISP. While some of the ISPs are furious and believe that the ministry
deceived them and caused them a lot of losses, others (ISPs) believe that their business
flourished more and more and increased their opportunities in attracting more users.
The new project came with new regulations and rules among which was the distinction
among ISPs according to a license class (Class A, B & C)
* Class A ISPs: there are only four Class A ISPs. These ISPs are characterized by the
owning of the Infrastructure (equipment, network and ports) through which the Internet
access is possible. They also have an exclusive contract with TE for intemational long
distance gateways access (Bandwidth). They also have the right to lease their
infrastructure and their intemational long distance gateways to class B & C ISPs plus
their ability to sell their Intemet service to end users via their free numbers.
Internet services (Class A)
Providing cor
ions
with Internet network (excluding voice phone calls se
For the final consumer through the licensed companies selfowned networks
® Nile On Line
® Link Egypt
© TE Data
* Egynet
Figure 1 Class A ISPs
(Source: TRA, 2001)
* Class B ISPs: there are only eight Class B ISPS, they also own the infrastructure but
they can only access the intemational gateways via Class A ISPs. They sell their service
to consumers and enterprises.
@® Data communication & internet class B
The interconnection to the public internet is ta be established via a class A licensed
service provider
* EgyNet
*Yalla
* Global One
© Noor
® Raya Telecom
® Menanet
* Online
* Soficom
Figure 2 Class B ISPs
(Source: TRA, 2001)
* Class C ISPs: there is a large number of this ISP class (approximately 50 to 60);
they don’t own the infrastructure or the intemational gateway access. They can only
lease from class A ISPs and then sell to consumers (Etcheverry & Khalil, 2002).
MB Internet class C:
TRA is in the process of issuing license to companie
The interconnection to the public intemet is to be prov
Class C form
Figure 3 Class C ISPs
(Source: TRA, 2001)
Internet Market Speculation Before/After the Implementation of the “Free Internet”
Model
Competition: With the total cost of connecting now reduced by as much as 59 percent,
“Free Internet” has heightened competition among ISPs, who must compete on quality of
service, speed of connection, marketing strategy and brand recognition, according to Arab
Advisors Group, the new scheme of competition would be tougher (Arab Advisors Group,
2001), (Hranjski, 2002).
This tough competition may cause consolidation and shakeout phase in the country that
will reduce the number of infrastructure-based "real" ISPs to 9 by 2006 (Arab Advisors
Group, 2002).
One result of this market consolidation is the emergence of the virtual ISP (Class C ISP), a
company that takes lines and numbers from the class A and B ISPs and pwmotes them
with their own access code numbers. Privately branded virtual ISPs, promoted by major
brand names in Egypt in association with the big ISPs, are expected to increase to 145 by
2006, predicted Shanin Shanin, an analyst at Arab Advisors Group (Southwell, 2002).
Users: When the MCIT announced its plans to implement the “Free Intemet” model in
Egypt, it hoped to see one million more Egyptians join the community of browsers on the
same year of the implementation (Pyramid Research, 2002). The ministry along with TE
took many steps among which was the decrease of the Intemet call cost by 59% and
reducing the rental fees for primary rate interface (PRI) ISDN services as well. Among the
expectations of the number of Intemet users increase after the implementation of the
project was that of the Arab Advisors Group who concluded in its Research Note that the
introduction of the “Free Internet” model in Egypt will result in a boost in the number of
Intemet users in the country and an increase in the revenues of the Egyptian ISPs (Arab
Advisors Group, 2001).
They project Egypt to have 680,000 Intemet accounts in 2006, a penetration rate of 0.93%,
up from 0.18% in 2001. Their definition of "Intemet Accounts" includes regular users of
the “Free Intemet” model" Shahin Shahin, Arab Advisors Group analyst said. "In 2006,
Intemet users are projected to exceed 2.6 million in Egypt, up from an estimated 540,000
in 2001(Arab Advisors Group, 2002). According to an announcement by the counselor of
the minister of Communication and Information Technology, the number of Internet Users
in Egypt reached 1,500,000 users as of January 2003. The number was assumed according
to statistics of the TE tracking 750,000 different telephone lines dialing the “Free Intemet”
non-geographic numbers, and assuming that two persons are using the same telephone line,
we can reach the above number of Intemet Users (Taha, 2003).
MENTAL MODEL
HIGH LEVEL FRAMEWORK
As shown in the diagram below, the Framework of this current study was divided into
Independent, Dependent and Moderating Variables. This higher-level framework is further
divided into a lower level one represented by a causal loop, which is fed by information
gathered from interviews.
Independent Variable Dependent Variable
Moderating Variables
Independent Variable Dependent Variable
¢ Subscription fee cancellation.
* New License Classes. ¢ Internet Users.
* Movement of Equipment to TE ¢ Internet Usage.
Premises (New Costs). ¢ ISPs Survival.
* Revenue Sharing Model (70-30%).
Moderating Variables
* PC penetration.
Literacy.
* Computer Literacy.
* Culture.
* Censorship.
* Local Content.
+ E-commerce.
+ Intemet Call Tariff.
* Customer Loyalty to ISPs.
* Telephone line penetration.
LOW LEVEL FRAMEWORK —
1)
ecommerce
‘Local Content
Dp
Competition Local Content
Advertisements
Class C ISP:
ere SiS i.
Class A ISP Profit Class B ISP ProfitClass C ISP Profit fon to=)
C ISPS Market Exit,
tf
VO Tine lines
pees “4 penetration
4—) ae Computer
eenT Usage peeinien
Class C ISP *y
Revenue Traffic
\ (
TE Revenue
Class A ISP
Revenue
Needed Ports
+
Needed
Bandwidth
$ Outflow
Figure 4 Internet Market Causal Loop
The above figure (figure 4) was that of the whole Intemet Market Causal loop comprising
all the relations gathered from primary as well as secondary data.
The loop is composed of twenty-three different loops but only the description of the eight
most important ones is to follow.
Loop 1 (Users/Usage): Balancing
Telephone lines
Quality/Speed penetration
+
: Intemet Users
& _ ae Computer
4) penetration
Users/U sage
es Tariff
Traffis
Figure 5 Loop 1: Users/Usage
This loop shows the effect of telephone lines penetration, Computer penetration and Tariff
on the Internet Users. Both telephone lines and computer penetration affect the Internet
users in the same direction, which means that any increase in the telephone lines
penetration and in the computer penetration lead to an increase in the number of Internet
users too. This increase in the number of Intemet users will increase the Traffic of the
Intemet (ie. Intemet Usage in minutes). An increase in traffic will in tum lead to a
deterioration of both quality and speed, which will affect the traffic in the opposite
direction. As for the Intemet tariff, it affects the number of users in the opposite direction,
which means that any decrease in the Intemet tariff attracts more users.
Loop 2 (Users Attraction): Reinforcing
eer Advertisements a
Competition ‘Awereness
ms
+ f+
+
Intemet Users
Class A ISP Profit Class BISP Profit Class C ISP Profit
se so
Class C ISP i)
Revenue Users Attraction
i
‘.
Class A ISP
Revenue
Class B ISP
Revenue
TE Revenue,
Traffic
Figure 6 Loop 2: Users Attraction
This loop shows the effect of Intemet Users increase on both TE Revenue and the ISPs
Revenue. An important clarification must be mentioned here; before the “Free Internet”
project launch, all ISPs were treated equally, there was no classes among them'. Another
important issue is that the revenue of the ISPs was solely coming from the subscription
fees of the users, which means that any increase in the Intemet users signifies an increase
in the ISPs revenue, while an increase in the traffic generated by the users signifies an
increase in TE revenue only because it was used to collect 100% of the Intemet call tariff.
After the launch of the project, the revenue of the ISPs from the increase in the number of
Intemet Users is zero because there are no more subscription fees, while its revenue is
coming now from the traffic; TE collects the Intemet telephone fees and gives the ISPs
70% of the Intemet traffic Revenue while TE takes up the other 30%. After the launch of
the “Free Internet” project, the competition between the ISPs increased so they tended to
create new means of attracting the subscribers to use their “Free” numbers. This is
reflected in the increase in advertisement, which leads to an increase of awareness of the
Internet and hence attracts new users as well as old users to use a specific “Free Internet”
number.
Loop 3 (Local Content): Reinforcing
fo Content
+
ae ica >
Oo +
Internet User:
Class A ISP ProfitClass B ISP Profit Class C ISP Profit
4 Py a
Class A ISP Class B ISP Class C ISP
Revenue Revenue
TE Revenue
Traffic
Figure 7 Loop 3: Local Content
This loop is exactly the same as the one above; the only difference between them is in the
means of attracting the users, which is the increase in local content in this loop. By local
content, the researcher means the WebPages in local language, which increased
tremendously after the launch of the project.
1 This point is taken into consideration in the Stock & Flow model
Loop 4 (e-commerce): Reinforcing
e-commerce
pert
Competitio,
+ eee
* +
Internet Users
Class A ISP Profit Class B ISP Profit Class C ISP _c)
Class AISP Class BISP Class C a SONY
Revenue Revenue + Revenue
+
(*
LE Revenue
Traffic
Figure 8 Loop 4: e-commerce
This loop is exactly the same as the one above; but the means of attracting the users is the
increase in the availability of buying and selling through the e-commerce sites. But
interviewees confirmed that the attraction of new users through e-commerce is almost
negligible.
Loop 5 (Costs): Balancing Advertisements
pe
oe
A
Class A ISP Costs
thet C ISP Costs
Costs
Class B ISP Profit
Class C ISP Profit
Class A ISP Profit
Class B ISP Costs
Figure 9 Loop 5 Costs
This loop represents the costs of advertising on all ISP classes, which in its tum decreases
the profit of each class. The same also goes for the local content increase and the e-
commerce websites increase, but the cost of advertising is much higher than that of the
other two.
Loop 6 (Bandwidth /International Access Lines): Balancing/Reinforcing
gre SS
re Ny
Awereness
Class A Lo Profit \
4) Class B ISP ProfitClass C ISP Profit
Internet Users
lass : a Profit
Class A ISP ¥
Revenue, Class B i. Costs Traffic
Class A ISP Costs
-) Class C ISP Costs
cask = Loss a
Needed
Class 2 ISPs Loss . =
Bandwidth ee
Figure 10 Loop 6: International Access
The increase in Intemet usage, which is represented in this loop by the word “Traffic” will
require at a certain point an increase in Bandwidth. Before the implementation of the “Free
Internet’ project, TE was responsible for the bandwidth and it was monopolizing the
market and rent these lines to different ISPs. After the project, this responsibility was
shifted to Class A ISPs which in their tum lease these lines to ISPs from Classes B & C.
So, any increase in bandwidth will add a cost on Class A ISPs, because it is the class
responsible for that increase, and after increasing those lines, Class A ISPs will lease them
to Class B & C ISPs and hence this item will constitute a cost for them but a source of
revenue for Class A ISPs.
Loop 7 (Ports/ Infrastructure): Balancing
; Advertisements.
5 Awareness
Competition
+ +
*
Internet Users
Class A ISP Profit Class B ISP Profit Class C ISP Profit
fe
ClaséA Isp Class B ISP
Révenue Revenue
Class B ISP Costs i
+ Class C ISP Costs Traffic
ip =
Class C ISPs Loss
Ports <a Quality/Speed
t—) Needed Ports
A ISPs Loss
Class A ISP Costs
Figure 11 Loop 7: Infrastructure
This loop is exactly the same as loop 6, except that the bandwidth is substituted by the
Ports (Infrastructure). Another difference from loop 6 is that the Ports expansion is the
responsibility of both Class A ad Class B ISPs while Class C ISPs lease the needed ports
from class A ISPs.
Loop 8 (Class C ISPs): Balancing
Class C ISP:
ing
Class C ISP Profit
Class C ISPs Market Exit
Class C ISP Costs
4
Ports Leasi
+
Bandwidth Lease
Figure 12 Loop 8: Class C ISPs
This loop called the “class C ISPs Market exit” represents the attraction of new entrants
(Class C ISPs) to the Intemet market due to the increase in the profits of the Intemet
business and the low entry barrier. These ISPs have then to lease some ports and a certain
bandwidth and this will increase their costs, which if not covered by the revenue will lead
to an exit from this market.
STOCK & FLOW
Locd Content Sits,
Computer Penetration
Ports lease fs
Figure 13 Stock & Flow Diagram
Input to Stock and Flow
Table 1 Stock & Flow Input
Variables Status Equation Units Notes & Assumptions
It takes 3 month for a person to
Awareness Conveyor TRANSIT TIME =3 Unitless become a user after being aware
of the Intemet
Advertisements* Potential : Advertisement increases
Awareness Increase Flow Users*0,0000001/12 Uniteless/Month ‘Awareness
Delayed Awareness | Flow CONVEYOR OUTFLOW 3 Months delay
Increase
Bandwidth Stock INITIAL Bandwidth = 150 Mbit/s Mbit/s (TRA, 2001)
F Mbit/s
Bandwidth Increase Flow GRAPH per Month (Taha, 2003)
Class A Profit Stock Revenue - Costs EGP
if TIME<=24 then ISPs Rev*0.55 else ISPs
ClassA Revenue | Flow | Rev*0.65+Bandwidth Lease+0.75*Ports EGP Market stare trom (Etchevery &
L Khalil, 2002)
ease
Advertisements*0.65+(Current Market share from (Etchevery &
Class A Costs Flow Ports*0.75*3000/12)+Bandwidth*5000/12+ EGP Khalil, 2002), Ports distribution
ecommerce*0.65/3+Local Content from TE Interview, else is
Sites*0.65/2 assumed
Class B Profit Stock Revenue - Costs EGP -
CIASB ReVvEnts Flow if TIME<=24 then ISPs Rev*0.3 EGP Market share from (Etchevery &
else ISPs Rev*0.15
Khalil, 2002)
Advertisements*0.1+(Current
Ports* 0.25* 3000/12) +B andwidth
Market share from (Etchevery &
Khalil, 2002), Ports distribution
Class B Costs Flow Lease*0.25 EGP from TE Intervi se i
+e-commerce*0.15/3+Local Content Tom sesumed else Is
Sites*0.15/2
Class C Profit Stock Revenue - Costs EGP -
if TIME<=24 then ISPs Rev*0.15 Market share from (Etchevery &
ClassCRevenue | Flow else ISPs Rev*0.2 EGP Khalil, 2002)
Advertisements*0.25+(Ports
Class C Costs Flow | — Lease*0.3)+Bandwidth Lease*0.25+e EGP Mill 30) davizeeet
commerce* 0).2/3+Local Content Sites*0.2/2 , ,
Class C ISPs Stock Initial Value = 60 Units Gathered from Interviews
F Ateach time the revenue does not
Class C Market Exit Flow if (Class C Revenue - Class C Costs) <=0 Units cover costs
then 1 else 0 é j
one ISP will go out of business
Computer Penetration Stock INIT Computer Penetration = 700,000 Units IDSC Interview
Computer narwese Flow Computer Penetration* 0.24/12 Units/Month IDSC Interview
Current Ports Stock INIT Current Ports = 65,000 Number of Ports IDSC Interview
Tntemet Usage Stock INIT Internet Usage = 0 Minutes/Month Accumulated Minutes
(Internet Users*A verage Access E
‘ Traffic will never
time)>(Current Ports*24*30*60) then. . : —
Intemet Traffic Flow (Current Ports*24*30*60) else (Intemet Minutes/Month soe eevex “ will
Users*A verage Access time*Quality/100) y the quality
Intemet Users Stock INIT Internet Users = 500,000 People (TRA, 2001)
Potential Users Stock INIT Potential users = 32,000,000 Number Assumed to be equal to the
educated population
Educated Population
Growth Flow Potential users*0.0166/12 Person/Month (The World FactBook, 2002)
(min(Computer Penetration, Phone lines
Intemet Users Increase Penetration)*0.03/(Tariff*0.5))
Rate Blow +delayed Awareness increase Local Person/Month Assumed
Content Sites-++e commerce
Local Content Stock Initial Value = 1700 Units (Tucker, a Shalaby,
* Competition over websites
shales mess Flow coment conta UnityMonth intensified
. before the project launch
Phone bines Stock | INIT Phone lines Penetration = 5,820,000 Units (TRA, 2001)
Phones Increase Rate Flow Phone lines Penetration+(500,000/12) Units/Month TE Interview
is Ean Oven Cook Time =4 Unitless IDSC Interview
if Current Ports< (Intemet Users/10) then r ,
Quality Converter (100-(((Intemet Users/10)- Unitless iD oe a ge i a a
Current Ports))/Current Ports*100) else 100 ~
if Quality<100 then ((Intemet
Quality Enhancement Users+(Internet Users Increase
Requirement Flow Rate*4))/10)- Number of Ports Assumed
Current Ports else 0
Needed Ports Flow Content of Oven after Cook Number of Ports
if time <=24 then 0 else if Competition<0
Advertisements Converter then 50000 else if Cotnpention< EGP Assumed
30*1000*1000
then Competition*0.001 else 100,000
if TIME<24 then 450
Average Access else if (TIME>=24 and TIME <28) then Minutes/Month :
Time/user Converter 500 else (if TIME>=28 and TIME <36 then per User Average Value From Interviews
600 else 750)
Competition Converter | ClassC profit+ClassA profit+ClassB profit EGP Assumed
if TIME <20 then 0
e-commerce Converter else Competition*0,000000001 EGP Assumed
Ports Lease Converter Current Ports* Ports lease fees EGP -
Ports Lease Fees Converter 3000/12 EGP/Month TE Prices Brochure from
Interview
Bandwidth Lease Fees | Converter 5000/12 EGP/Month TE Paces B rochure from
Interview
Bandwidth lease Converter Bandwidth* Bandwidth Lease fees EGP -
Intemet Access Fees | Converter IF Time <=24 then 100 else 0 EGP Actual Data
if Time>24 then TE Internet Revenue
ISPs Revenue Converter Delay*0.7/0.3 EGP Actual Data
else Internet Users* Intemet access fees
Phone Lines Fees Converter IF Time <=24 then 1.8 else 1.23 EGP/Hour Actual Data
Tanff Converter Internet A ccess fees+Phone line fees EGP
if TIME <=24 then Intemet Traffic* Phone
TE Intemet Revenue | Converter line fees/60 else EGP/Month Actual Data
Intemet Traffic* Phone line fees*0.3/60
TE Intemet Revenue Telephone bill cycle
Delay Converter DELAY (TE Intermet Revenue,3) EGP/Month ie 3emonths
Model Validation
In order to make sure that the model is representative of the Intemet market in Egypt, it is
essential to regenerate the actual historical data given when running the Stock and Flow
model under the same conditions.
Table 2 Stock & Flow Model Validation
Start of Y ear 2003 Output from Stock & Actual Error
Flow Data
Intemet Users 1,575,708 1,500,000 -5.05%
Ports (Infrastructure) 163,448 170,000 3.85%
Average 6
Usage/Month 437,559,645 520,000,000 15.85%
The above table compares some of the Stock and Flow model outputs with the actual ones.
The number of Intemet users at the start of year 2003 was 1,500,000 and the average usage
per month is around 520,000,000 minute per month as stated by the assistant of the
minister of Communications and Information Technology, (Taha, 2003).
As for the number of ports, it reached 170,000 ports by the start of 2003; this information
was gathered through interviews. The error of the model in both the Intemet Users and the
Ports number is 5% and almost 4% respectively, as for that of the average usage per
month; it’s almost 16%. This high error in the average usage per month is due to the fact
that the researcher was not able to have accurate statistics about that number, and the
information gathered through interviews were so diverse so the researcher had to assume
an average usage per subscriber which differs from period to period’.
FINDINGS
Conceming revenue, the elimination of the subscription fee affected the ISPs to a great
extent, as many of them were solely dependent on this fee. TE also had a revenue decrease
because since the project launch, it gives 70% of the revenue to the ISPs in addition to the
decrease of the call tariff from 1.8 EGP per hour to only 1.23 EGP per hour. The following
graph clarifies this point, although the numbers may not be very accurate but the slope and
shape of the graph is correct to a good extent.
2 See above table of Stock & Flow Input
0.90 10 180 21.60 73.80
Figure 14 TE & ISPs Revenue
Conceming costs; as infrastructure constitutes to all ISPs more than 50% of the total cost
according to ISPs interviews, it increased tremendously for all classes due to the increase
in infrastructure before and throughout the launch of the project. For classes A and B, their
costs increased in order to build a better network capable of supporting the new traffic all
over Egypt and not only in Cairo, and they also increased their bandwidth in order to
enhance speed and to conform with the MCIT/TRA new rules. For class C, they too had to
lease more bandwidth and ports in order to enhance their performance. For all classes and
especially class C ISPs, (who were originally either small portals or small ISPs providing
the Intemet Access service for a fee), they all had to increase their marketing and sales
budget which was almost nothing before the implementation of the “Free Intemet” Project.
The only cost that decreased a little for many of the ISPs is the customer service cost part
responsible for customer's access problems. This service is no longer important for dial-up
users because if the user confronts any problem during access, he/she simply switches to
another number, but some of the respondents said that they are keeping this customer
service as a value added for users of their dial-up free number.
prs
f
Figure 15 ISPs Costs
Concerning rules and regulations, all ISPs suffer from the lack of information about the
new rules and regulations they get from the ministry or from the TRA and they also
complain from not taking part in the decision making of the new nules and regulations.
Most of the respondents also said that the new licenses are not fair; the condition that has
been taken in order to be a class A ISP was that the ISP must be in the market since 1997.
For many of the ISPs this prerequisite was not fair, but anyway this exclusivity that the
class A ISPs have now will end by 2005, but the deterrent then will not be the license as it
will be the high entry barrier.
Market Predictions
The following market predictions are based on both data gathered from interviews and
output from Stock & Flow model.
The number of Intemet Users in Egypt will continue to grow reaching almost 4,000,000
users by the end of 2004 (month 60). The following graph shows the output from the Stock
& Flow model.
209
Figure 16 Internet Users projection
Quality and Infrastructure must always be ameliorated by ISPs in accordance with
the MCIT/TRA mules and regulations and as a response to the increase in the number of
users.
Figure 17 Bandwidth & Infrastructure Projection
The above graph shows the projection of the increase in ports and bandwidth till end of
year 2004. Note that the bandwidth graph is an input gathered from interviews as well as
MCIT announcements in the newspapas, which says that the total bandwidth will reach
1Gbit/s soon (MENA, 2003).
As a result of users increase and hence competition, more local content sites will be
developed by ISPs trying to attract the users to use their free numbers. ecommerce will
increase too but not with the same pace as that of the local content. The increase of the
local content along with advertisements will increase awareness.
le
Figure 18 Awareness, Local C ontent Sites & e-commerce projections
Predictions for the future of the ISPs as gathered from interviews, qualitative and
quantitative analysis is that consolidations and mergers will continue to exist. Big ISPs
with large capabilities and size will buy smaller ones whether ISPs or portals. Class A ISPs
are not to be worried about because of their relations with the ministry and because of their
size. Many of the interviewees stated that class A ISPs survival is a political issue, in
addition that they have quiet steady revenue, if not from the “Free Intemet” dialup, it is
coming from the bandwidth and ports rental business. Class B ISPs future is not as good as
class A but some of them have the advantages that class A has and some of them also
enjoy the big size and diversity which may guarantee a long life for them.
For these two classes, ISPs must go for consolidation which is confirmed by many of the
interviewees who stated that the only advantage of this “Free Intemet” model if there are
any, is that it brought ISPs together; many of the ISPs are now holding meetings and
discussing their problems trying to solve them together and help each other by pushing on
the MCIT.
As for class C ISPs, their future will not be very bright especially for those who don’t have
a good marketing strategy to follow or who are not backed by good brand named
companies or advertising agencies. For this class, either it will fall as an easy target for
competition or it will be bought by bigger successful ISPs from class A or B. In the graph
below, starting with a number of sixty class C ISPs in 2000, the number is 50 by week 36
(end of year 2002), which is almost true. By the end of 2004, there will be only 25 class C
ISPs
Figure 19 Class C ISPs
CONCLUSION
The “Free Intemet” Project achieved its main aim of increasing the penetration of the
Intemet in Egypt to a great extent but there were some side effects that could have been
eliminated or decreased if the ministry had taken some more steps before the launch of the
project, these steps would have increased the penetration of the Intemet even more. It is
true that the Internet tariff and the subscription fee were an obstacle for many to use the
Intemet but another more important deterrent was the computer penetration.
The usage of the Intemet seems to be increasing too, but with the new telephone bill
presenting the Internet usage explicitly and the required amount of money owed for this
usage; this could have a negative effect on usage in the coming months.
The “Free Intemet” project also had a bad effect on the survivability of the ISPs. Some of
them lost their business completely while others still in the market suffer from high costs
and low revenue. Very few interviewees, only two out of ten ISP interviewees, said that
their revenues are covering their costs; they represented two big ISPs only. The project
didn’t only affect ISPs after its launch, but the announcement of the intentions of the
Ministry to launch the project almost a year before the actual implementation had the worst
effect on the market. Even before the cancellation of the subscription fee, people did not
want to pay or renew their subscription because they were confused about the
announcement and thought the “Free Intemet” project already started at that time during
year 2001.
Problems raised before, during and after the project are not only the responsibility of the
MCIT but also the ISPs could have eliminated some of the problems they are facing now.
As for the MCIT, a deeper study of the market and its determinants as well as the study of
the “Free Intemet” model implemented in other countries would have had helped the
ministry avoid the problems faced by others. Also, consulting the ISPs with all the
experience they had in the Egyptian Internet market since the 1990’s would have had a lot
of benefit for both sides especially that ISPs would have been less resistant to the project.
Also, the condition that class A and B ISPs must co-locate their equipment inside TE
premises was totally unfair for the ISPs because it costs them more than originally
budgeted for, causing an irrecoverable loss in revenue in the short run.
As for the ISPs, they also made mistakes since the announcement of the project; first when
they found the subscribers refusing to renew their subscription, they launched a campaign
of “99 EGP for Life” subscription. This campaign affected the loyalty of the users because
they realized then that these ISPs were just trying to collect money and this life
subscription was just a way for doing so. ISPs lost the trust of the customers by doing this.
Second, since the announcement, every ISP was buying and extending its ports and
infrastructure to all govemorates in Egypt as if this ISP would be the only one covering all
of Egypt, having 100% market share. These huge investments added to the trouble of the
ISPs and after the launch of the project, they all realized that they were overestimating the
number of users who would actually use their network and that is why many of them
started to decrease and abandon some of the ports installed, which further added costs. A
study by the ISPs before their extension would have made a difference and decreased their
costs.
The sole and only winner from this project is the end user who has gained a better quality,
a higher speed connection, and a larger content in his own language, all this with a lower
price. However, the content is still mainly for entertainment purposes such as chatting,
music and pictures download, not educational or valuable content which, if it existed,
would have made a huge difference to cultural literacy and would have helped the Intemet
penetration even more, helping achieve the aim of the ministry which was to the benefit of
Egypt in the first place.
REFERENCES
Arab Advisors Group (2001). Free Interna in Egypt by the end of the year. Retrieved
November, 17, 2002 from Web site: http://www.arabadvisors.com/Pressers/presser-
240601.htm
Arab Advisors Group (2002). Egypt's Internet subscribers will grow substantially, but
consolidation will diminish the number of its ISPs. Retrieved November, 17, 2002
from Web site: http://www.arabadvisors.com/Pressers/presser-050602.htm
Balancing Act News Update (2001). Brazil’s Free ISPs Heading For Profitability? May
Be.. Retrieved November, 11, 2002, from Balancing Act News Update Web site:
http://www.balancingact africa.com/news/back/balancing-act_72b.html
Buisson, J., Sergent, J. & Frattaroli, V. (2000). Arcome, Les Competence Telecom:
Panorama européen des conditions d’ interconnexion pour le trafic Internet commute.
Ensoport Intemetworks (2001). Free Internet Service: Vandalism on the Information Super
Highway. Retrieved November, 11, 2002 from Ensoport Intemetworks Web site:
http://www.volja.net/accounttools/eng/presse_ministry_en.pdf
Etchevery, A. & Khalil, R. (2002). Gratuit, mais pas encore pour tous. Retrieved
November, 27, 2002 from Al-Ahram Hebdo Web site:
http://hebdo.ahram.org.eg/arab/ahram/2002/11/27/A ffal .htm
Hranjski, H. (2002). Egypt boasts free Internet service. Retrieved November 16, 2002 from
Web site: http://www.philly.com/mld/philly/business/technology/4031269.htm
MENA (2003). Middle East News Agency. Retrieved March, 09, 2003 from MENA Web
site: http:/Awww.mena.org.eq
Palmgren, M.A. (2002). Free Internet access introduced in Egypt. Retrieved November,
15, 2002 from HURIDOCS-Tech Web site: http://www.hrea.org/lists/huridocs-
tech/markup/msg00812.html
Pyramid Research (2002). Egypt: Free Internet to Provide a Boost to ISPs and Telecom
Egypt. Retrieved November, 16, 2002 from Pyramid Research Online Store Web site:
http://shop.store.yahoo.com/pyrstore/egfreintopro.html
Sloboda, B. (2001). What's Free On The Internet: An Update for Users. Management
Quarterly, Summer 2001, 25-43
Southwell, M. (2002). arabia.com offers free Internet access in Egypt. Retrieved
November, 20, 2002 from The hformation & Technology Publishing Co. Ltd. Web
site: http://www.itp.net/news/102873433342674.htm
Taha, M. (2003). Egyptians use 520 Million Minutes per Month of Internet & 40 million
EGP Revenue of Link.net from the Free Internet Retrieved January, 30, 2003 from
Masrawy News Web site:
http://news.masrawy.com/masrawynews/16012003/116634news.htm
http://news.masrawy.com/masrawynews/16012003/116636news.htm
The World FactBook 2002 (2002). Egypt. Retrieved December 20, 2002, from the CIA
World Fact Book Web site: http://www.cia.gov/cia/publications/factbook/geos/eg.html
TRA. (2001). Licenses. Retrieved Febmary 18, 2003 from Web _ site:
www.tra.gov.eg/f_licensed_data.asp
Tucker, A., Younis, S.F., Shalaby, T. (2002). Cross cultural perceptions of the Internet
and Virtual Reality. Retrieved November 15, 2002 from Web site:
http://www2002.0rg/CDROM/altemate/669
Back to the Top
