> it's not on a per user basis because that would make no sense Unless I have misunderstood, you are suggesting that fixed pricing provides higher expected ret

> it's not on a per user basis because that would make no sense Unless I have misunderstood, you are suggesting that fixed pricing provides higher expected returns than tailored pricing, which I think is incorrect. Imagine people's private value for a particular good is uniformly distributed between 0 and 1. In the scheme you propose the expected per-customer return for a fixed price x is x * (1-x), because everyone the x fraction of people whose private value falls below x will not pay at all, returning 0, and the return for all the (1-x) fraction of people whose private value is above x is just x, because the price is fixed. In a tailored pricing scheme, still requiring a minimum price of x, the expected per-customer return is 0.5 * (1+x) * (1-x), because again the x fraction of people whose private value is below x will not pay at all, returning 0, but for the (1-x) fraction whose private value is above x you expect to receive their average private value, or 0.5 * (1+x) (obviously assuming you could determine this perfectly). The maximum return under these two schemes is using the tailored pricing scheme with x set as 0, meaning that tailored pricing for all customers provides the highest expected return. You can see a plot of the two curves here [0]. This is a simple model, and perhaps someone more skilled in mathematics can provide a more general answer, but it does show that tailored pricing _can_ be beneficial (suggesting firms would probably try to do it if they found it to make sense, and were able). I would also note that the following quote seems like it is probably not correct as well, given that a web server has full control over what it returns to each individual, as well as the means for obtaining fairly reliable distinguishing information. The prevalence of these behaviours is difficult to assess because it is somewhat difficult to perform comparisons between individual responses at scale. > The public nature of most consumer prices makes it so that this kind of scheme is not all that tenable [0] http://www.wolframalpha.com/input/?i=plot+0.5(1%2Bx)(1-x),+x(1-x)+for+x+%3D+0+to+1 http://www.wolframalpha.com/input/?i=plot+0.5(1%2Bx)(1-x),+x...