Revenue allocation and the free ISP model

1
Revenue allocation and the free ISP model

• Russell Southwood
• Balancing Act
• http//www.balancingact-africa.com
• Info_at_balancingact-africa.com

2
Balancing Act - The commercial

• News Update. Weekly e-letter on African telecoms,
  internet and computing. 6,500 subscribers across
  Africa. Want to subscribe...
• Projects - ICT and Entrepreneurialism series and
  CD-ROM. Ghana, Kenya, Botswana, Tanzania and
  Nigeria. Forthcoming in Mali and Burkina Faso.
• Consultancy - Example - Regional IXP

3
ISPs in Africa - Background

• Dial-up as yardstick. SA and Egypt are cities.
  Kenya and Senegal are large towns and places like
  Mali are small villages
• Stagnation in the growth of dial-up subscriptions
• Consolidation of number of ISPs
• Incumbent often skews the market with own ISP
  (Sonatel in Senegal. 80 of market)

4
African ISPs caught in a price squeeze

• Customers want ever cheaper prices but the volume
  of customers is not increasing. Majority of
  customers are buying on price, not other factors
  like quality of service
• Upstream bandwidth providers (esp int) have
  margins under pressure and are looking to pass on
  costs (Satellite floor price USD2-3000 mbps?)

5
Business model in an unsustainable phase?

• Too many ISPs earning too little income chasing
  too few customers
• Likewise too many cyber-cafes chasingetc (diff
  between top and bottom end rate 12)
• Rush for higher ground of corporate market also
  makes that space crowded.
• Me-too market. No competitive advantage lasts long

6
Too little revenue in the current model

• amount of revenue paid by customer left to ISP
  after paying upstream providers
• Nigeria (20), small ISP - SA (30), Tunisia
  (33), Kenya (45)
• N Africa - Av 43. Morocco (48) 51 going to
  incumbent telco (IDC Study)
• Need change in revenue allocation OR new
  revenue streams. VOIP calling?

7
How revenue allocation and free ISP models cross
paths

• Discussions of Q of revenue allocation often lead
  in direction of revenue sharing (cf UK
  experience)
• Recurring underlying Q In a stagnant market, how
  do we create growth for ISPs?

8
Whats a free ISP model

• Free at the point of delivery
• Egypt (revenue sharing, no monthly charge, only
  access time)
• Freenet, Uganda Telecom (no monthly charge, only
  access time) Put market in jeopardy without
  substantial growth in numbers
• South African examples

9
Egypts free ISP model - Background

• 70m population. Larger volume of users (1.2 m)
  Economies of scale. Incumbent Telecom Egypt
• Policy level Ministry for ICT set up in late 90s
  as dedicated Min. Backing from top. Mubarak at
  library openingI need to see a leased line
  here
• Technology hub for ME. Easier to get system
  engineers. Much greater level of ICT awareness

10
How was it structured

• Toll-based model at a premium rate. National
  number, no password. Collections done through
  telco every 3 months
• All ISPs had agreement with telco so that they
  could co-locate
• Split of traffic income 70 for ISP, 30 for the
  telco. Cost USD0.22 per hr. Launched Jan 2002
• Predicted it would increase subscribers 4-fold
• Accompanied by a computer loan scheme

11
Impact on the market (1)

• Largest ISP (25 of market) 15m minutes up to
  30m minutes but then only marginal growth.
  Slipped back, barely doubled.
• 19 POPs to start. Within 6 months needed 87 POPs.
  Hard work and costly. All ISPs hurting but a
  users paradise

12
Impact on market (2)

• Ministry/telco- Privately, OK we got it wrong.
  Incd cost of local calls by 20. How can we help
  on the rest?
• Cost of E1s and co-location from the ministry.
  Take 30 off next 3 quarters.
• Still not working? Lets talk bandwidth charges
  and import duties

13
Conclusions

• After all that, just barely covering the costs of
  delivering the service. Still not making money.
  Same number of subs for this ISP but selling more
  cheaply. Govt cant afford to back down
• Other barriers to growth Income, education,
  literacy

14
What model of revenue allocations can drive
long-term growth?

• Customer POV
• ISP investment POV
• Is there a new revenue split? Is there a free
  ISP model based on revenue share that could work?

15
Are we reaching the limits of growth?

• If no, where does new growth come from?
• If yes, what does that mean for African ISPs