Everyone who ever goes on strike is aware that they are doing so at the cost of their wages for that period. That's why long strikes are rare and gruelling.

However, the cost to Uber is hard to predict - the strike is on the day of the IPO. If they can depress the share price at IPO by a few percent that potentially costs Uber far more than all the lost rides.

Edit: I suppose the entertaining thing about the gig economy is that it's possible to be "on strike" at employer A while working for employer B. Especially as there's no picket line here. (I think a few French anti-Uber strike/protests did have roadblock picket lines)