"but the general consensus is that those 'B' candidates can't survive the break-neck pace of innovation and that they'd generally not be able to match up, let a

"but the general consensus is that those 'B' candidates can't survive the break-neck pace of innovation and that they'd generally not be able to match up, let alone keep up with other performers and do the work that's expected of them exceptionally well. Since promotions are merit-based, and due to the stack-ranking employed by Amazon to rank the enggs, things only get diff for 'B' candidates." This is probably completely wrong , unless all of your B candidates are being hired at very low levels only, and none of them ever get to participate in the promotion process. Additionally, most tech companies are really bad at dealing with low performers in general (they tend to stay around too long). In any case, my sources are completely empirical, having managed large numbers of people at a number of tech companies (and thus watching things like what happens to B candidates during cross-organization score calibrations, etc), as well as mentoring and helping managers with some of these issues. I'd love to be wrong, but I have yet to see a company where the hiring of B candidates did not have significantly more impact that was expected.