4 ms·

Comparing the fine to total revenue and total profit is completely meaningless. What's relevant is how much they were saving in infrastructure costs by throttli

Comparing the fine to total revenue and total profit is completely meaningless. What's relevant is how much they were saving in infrastructure costs by throttling those users. E.g. If they saved less than $24 million and had a >50% chance of getting caught, this works perfectly well as a deterrent no matter how large there overall profits were.

davesque 10y ago
Very good point. I just hope that the penalties considerably outweigh the risks.
roymurdock 10y ago
Cost is only one side of the equation. You also have to factor in the approximate value of the revenue generated by subscribers who were misled into believing that they had unlimited, unthrottled data, and who might have selected a competitor's service had the secret quota been honestly advertised.
- MichaelBurge 10y ago
  Maybe, but the revenue has to have the price of its competitor subtracted, and the remainder weighted by the percentage of the service that they would use the 'unlimited' data for. So if choice 1 is $70/month for 25GB and $10 for 10GB afterward, and choice 2 is $90/month for 'unlimited' with an expectation to use 45GB, then both choices are equivalent. If the choice 2 then reduces the utility of the service by 50% past 25GB, then the customer has paid $20 for $10 of utility, and so the loss should be marked as $10. And if you buy choice 2 because it says 'unlimited', but you only use 5GB, then you never actually lost anything due to their deceit. I'd expect a similar calculation to hold up here: For each customer, sum up the number of throttled hours over the number of total hours to get the "throttling percentage". Also, assign each zip code a 'statistical competitor' by combining all competing services using some model. Then multiply the throttling percentage by the difference in price multiplied by the percentage of utility lost.

davesque 10y ago

Very good point. I just hope that the penalties considerably outweigh the risks.

roymurdock 10y ago

Cost is only one side of the equation. You also have to factor in the approximate value of the revenue generated by subscribers who were misled into believing that they had unlimited, unthrottled data, and who might have selected a competitor's service had the secret quota been honestly advertised.

MichaelBurge 10y ago
Maybe, but the revenue has to have the price of its competitor subtracted, and the remainder weighted by the percentage of the service that they would use the 'unlimited' data for. So if choice 1 is $70/month for 25GB and $10 for 10GB afterward, and choice 2 is $90/month for 'unlimited' with an expectation to use 45GB, then both choices are equivalent. If the choice 2 then reduces the utility of the service by 50% past 25GB, then the customer has paid $20 for $10 of utility, and so the loss should be marked as $10. And if you buy choice 2 because it says 'unlimited', but you only use 5GB, then you never actually lost anything due to their deceit. I'd expect a similar calculation to hold up here: For each customer, sum up the number of throttled hours over the number of total hours to get the "throttling percentage". Also, assign each zip code a 'statistical competitor' by combining all competing services using some model. Then multiply the throttling percentage by the difference in price multiplied by the percentage of utility lost.