Things are probably going to get worse before they get better, and plausibly quite a bit worse, but in terms of positioning one's business over the medium term, I'd be spending a lot of time thinking "What changes to the status quo are going to be the new normal going forward?"
One answer here: hundreds of thousands of people are getting their first exposure to remote work and will likely be in that environment for months, and hundreds of employers will have X,000 employee-quarters worth of data on how remote workers perform, for the first time.
It seems very plausibly to me that some of those employees say "Actually, I wouldn't mind this being my new normal" and some of their employers go "Actually, we didn't notice the lack of office presence, so maybe we could hire in *does math* 99.5% more of the US, as a start."
There are going to be some less positive versions of this. A lot of companies are going to do the "OK, batten the hatches and pull out the usual playbook: let's cut once and cut deep" thing.
In some cases, that will be eliminating positions which will likely not come back.
In some cases, that will be eliminating positions which will likely not come back.
On a more macro level, expect quite a lot of concern about how long and complicated supply chains have gotten. There will likely be rejiggering of that; there will also be a thousand startups trying to answer questions that there are no answers for right now.
e.g. If you are WalMart I bet you really keenly care about "What deliveries will not arrive on time if we take e.g. Manila out of the map for 6 weeks starting in 3 weeks" and I bet all the king's horses and all the king's men cannot answer that for you right now.
Also broadly I think many of us are going to get an education in non-linearity of systems, including supply chains, logistics, the healthcare system, etc, and the fact that humans are so bad about reasoning about non-linear effects will create some demand for aided reasoning.
Non-linearity of systems example: a Rails app which has 400 ms p99 responses with 0% drop at 1,000 requests per minute, 5000 ms p99 responses with 20% drop at 1,200 requests per minute, and ~100% drop at 1,500 requests per minute.
Sounds implausible... if you haven't done Rails.
Sounds implausible... if you haven't done Rails.
The things that will start showing implausible amounts of strain at striking thresholds are much more important than Rails apps and the impact to their ability to deliver service to the standard SLA is much worse than "a request gets dropped on the floor."
