3 ms·
Something I've come to believe is that in large tech companies the Pareto principle is not true because only 20% of SWEs are good and the rest suck. More likely
by ZephyrBlu 4y ago
Something I've come to believe is that in large tech companies the Pareto principle is not true because only 20% of SWEs are good and the rest suck. More likely, the majority of engineers are blocked in some way from delivering.
Dependencies on other teams, management constantly changing their mind, having to put out fires, helping other teams, etc. There are very few teams where this kind of thing doesn't happen.
Also, something to keep in mind when saying things like "you can fire 80% of your SWEs and survive" is slack. You might survive, but have no slack which means any sort of tail event is going to destroy your company.
- 88913527 4y agoThere's smart, but then there's creative. Smart engineers might be wizards of mastering the technology, but creative engineers will re-define the problem domain such that their team can deliver without being blocked by other teams, which is a frequent problem in large organizations. Solving business problems effectively is a different skillset from pure mastery of technology (though, you really do need both).
- ZephyrBlu 4y agoI understand this. I've been told a few times I can unblock myself very quickly, but at some point it becomes unreasonable. As a non-Staff IC, I probably can't unblock myself if my problem spans across multiple teams. For a manager/Staff eng, unblocking across a different org is probably infeasible. It feels like Director+ is when it becomes your responsibility to unblock things no matter what. Ex: the "Being a Leader Means Never Saying “That’s Not My Problem”" section in this article https://staysaasy.com/leadership/2022/08/15/how-to-break-into-senior-management.html https://staysaasy.com/leadership/2022/08/15/how-to-break-int...
- lazide 4y agoLarge companies build up armies of folks to stop SWEs from doing this. It’s literally what auditing, compliance, etc. do. Because short term optimizations teams do to get things out almost always cause larger (global) risks, from everything from new attack surfaces to yet-another-copy-of-user-data that now can get leaked because it isn’t actually protected the way it is supposed to be. Near as I can tell from stories, Twitter was already maxing out the auditing and compliance bypassing to get things done, so yikes on the long term fallout here.
- gorbachev 4y agoTwitter most certainly didn't keep the 20% that produce 80% of the output, though.
- gilbetron 4y agoQuite often, the 80% are supporting the 20%. I know I've spent many days trying to review PRs for someone in "20% mode", and likewise, when I'm in "20% mode", I can deluge my coworkers with PRs and RFCs and such. It is very true that reading code can often take more time than writing code. Additionally, the "20% mode" person can often miss little details because they are moving fast, and that detail capture work can take a lot of time for very little code change.