The key driver of the Zain move is obviously to take away subscribers away from Safaricom. The size of Safaricom's subscriber book is now seen as the biggest entry barrier into this sector and both the measures are aimed at reducing the book. As an aside, Safaricom paid Ksh4.5bn in interconnection rates (13% of its operating expense), you can thus imagine how much the other 3 players pay given most of their subscribers will be calling Safaricom customers.
Safaricom's subsequent response was clumsily presented, but the upshot is that its 8m or so subscribers will only pay Ksh2 per minute to call each other for the next month. Safaricom made Ksh63bn of its Ksh84bn revenue from voice calls last year. Thus Safaricom responded to the threat on its subscriber book (you can tell this is the case by the fact that postpaid customers will still be paying normal Ksh8 rate). The length of the offer period implies that Safaricom's thinking is that Zain can not sustain Ksh3 per minute beyond a month. Wishful thinking?
The issue is this. Can the voicecall providers make money if interconnection rates are reduced and hence they have to reduce the charge per minute? In my mind, the components that make up the cost of a call would be the fixed and variable (staff, commission, marketing) costs incurred by the provider; any costs associated with interconnection; other business-associated costs. A lot of the smaller players can probably sustain a price war because their interconnection cost has been halved. For Safaricom however, such a war would be costly because interconnection rates are a small portion of its business. It is clear that Safaricom is making a very healthy margin from voicecalls.
But the future portends lower margins and I think Safaricom shareholders should not ignore this especially in the medium-term when voice revenue will still be its predominant source of revenue.