Comment by acdha
20 hours ago
> If their output is as a tech lead then even a 2x can provide an overall 10x increase through second order impact via their team.
This is something the 10x mythology tends to leave out: there are a vanishingly few people who are significantly above the 90th percentile in terms of individual productivity but if the discussion shifted to team dynamics, that’s where you can actually see really big gains by helping a larger group be more productive.
I think its also that few companies have a way to allow a 10x individual productivity engineer to focus on just the things they are 10x as good at. It's almost never everything. Once you add in meetings, politics, proposals, perf ladder requirements, mentoring, code reviews, etc. the result is a regression to the mean.
10x came from actual measurements a few decades ago, getting people to implement the same project and seeing what the result was. The two parts missing from the modern usage of the term: the measurement was within a given team (not overall), and it was a comparison of best and worst, not best and average.
10x came from the 80s, so already fairly different in key ways (internet documentation, CI, platform and tool maturity, etc.), and the methodology is challenging because you don’t have easy comparisons between complex real projects without tons of confounds and trying to measure artificial challenges runs into different but also significant challenges selecting the candidates and ensuring that the work is representative.
There are definitely people who are more or less productive but I think we’re very prone to focus on the individual while ignoring the environment they’re working in, as well as the question about broadly applicable that result is.