Report reveals details about Amazon's secret algorithm redacted in FTC complaint.

Last week, the Federal Trade Commission sued Amazon, alleging that the online retailer was illegally maintaining a monopoly. Much of the FTC’s complaint against Amazon was redacted, but The Wall Street Journal yesterday revealed key details obscured in the complaint regarding a secret algorithm. The FTC alleged that Amazon once used the algorithm to raise prices across the most popular online shopping destinations.

People familiar with the FTC’s allegations in the complaint told the Journal that it all started when Amazon developed an algorithm code-named “Project Nessie.” It allegedly works by manipulating rivals’ weaker pricing algorithms and locking competitors into higher prices. The controversial algorithm was allegedly used for years and helped Amazon to “improve its profits on items across shopping categories” and “led competitors to raise their prices and charge customers more,” the WSJ reported.

  • @KoboldCoterieEnglish
    10•9 months ago

    It’s ‘The Formula’ from Fight Club:

    I was a recall coordinator. My job was to apply the formula.

    A new car built by my company leaves somewhere travelling at 60 mph. The rear differential locks up. The car crashes and burns with everyone trapped inside. Now, should we initiate a recall?

    Take the number of vehicles in the field, A. Multiply it by the probable rate of failure, B.
    Multiply the result by the average out-of-court settlement, C.

    A x B x C equals X. If X is less than the cost of a recall, we don’t do one.

    The key to preventing this kind of thing is to make sure that X is always considerably higher than A x B x C. If monetary penalties are all that are being assessed, they need to be fucking astronomical to the point that no reasonable company would even consider taking the risk of getting caught, no matter how minute that risk is. Until we start operating like that, there’s simply no incentive for any of these companies to stop breaking the law.

      • @KoboldCoterieEnglish
        1•9 months ago

        It’s a quote from Fight Club; I know this case has nothing to do with recalls. The point applies to any corporate lawsuit.

        Let me re-frame it, if that helps:

        Take the number of users, A. Multiply it by the expected per-user profit from price-spiking practices, B. Multiply the result by the chance of not getting caught / fined, C.

        A x B x C = X. If X is greater than the expected fine for doing this, we’ll do it even though it’s illegal.

        The point is that the penalty / fine that’s assessed needs to far out-weigh the potential benefit, taking into account the chance of getting caught / fined in the first place. If they stand to make $10 million in profit, and the penalty for getting caught is $100 million, but the chance of getting caught is only 5%, their ROI is positive and it’s simple numbers - they’ll do it every time.

        If in the above example the penalty was $1 billion, suddenly the ROI is negative, and they won’t do it. It’s simply math. By having penalties that are too low to impact their overall profit, they simply don’t work as a deterrant.