5 ms·
I'm not sure the solution makes sense for a lot of startup people. A mental model for repeatable luck building, especially when in a "yes mode", where just 1 b
by lmeyerov 4y ago
I'm not sure the solution makes sense for a lot of startup people.
A mental model for repeatable luck building, especially when in a "yes mode", where just 1 big fast-moving item is unlikely to be the job (networking, partnerships, devrel, customers, investing, ...) , is learning to prioritize bandwidth over latency. This can be true for engineering as well: solutions, bigco, ... .
A lot of the above things are unpredictable in when & how they strike. A latency mindset means pick 1-2 things at a time and work through them... but that's stressful when they are big hurry-up-and-wait items that would require heroic efforts to make it faster. Think David vs. Goliath moments in fighting the slow pace of whatever organizations you're working with. Instead, prioritize bandwidth by multiplexing to another item.
There's no free lunch even with a bandwidth mentality. Now the issue is being randomly oversubscribed when multiple priorities land at the same time.