[ntkachov]

They had 16 companies in this survey that were remote, 2 that were "office", And the rest were some form of hybrid. They considered anything other than full remote "some office". This could easily get skewed if one of these office companies had 10x'd their revenue.

And even though they grew faster, the attrition for Office was higher.

[sjducb]

In other words: "Study" was not statistically significant, all results should be ignored.

[rhaway84773]

In other words, as the article says, the study is “suggestive and not conclusive” and others should also be doing similar research.

[wokwokwok]

>>> others should also be doing similar research.

>> BTW, Nick Bloom at Stanford and others have done extensive research with thousands of people on remote and hybrid work here, and here…

^ from the article

> Home working led to a 13% performance increase, of which about 9% was from working more minutes per shift (fewer breaks and sick-days) and 4% from more calls per minute (attributed to a quieter working environment). Home workers also reported improved work satisfaction and experienced less turnover, but their promotion rate conditional on performance fell.

^ from the linked paper.

…but hey, go the statistically irrelevant one to share with people if you want a cool link to prove whatever point you want to make with it.

[sjducb]

You only get to say that when there are no other studies. If other good studies exist (which they do) then your study should add something. That means that your study should either replicate an existing study with similar or greater statistical power, or investigate something that the existing studies are not testing. This study does neither. It re-asks questions that have been answered with such low statistical power that the answers are meaningless.

[ekianjo]

more cherry picked than anything else, seeing the minuscule sample size and the obvious categorical bias

[colpabar]

All the experts agree!

[Zurrrrr]

Or at the least 7 out of 10 dentists agree

[teeray]

Expert tested, landlord approved.

[btown]

From the article:

> The survey sample size was 37 companies from the Reach Capital portfolio. That’s large enough to see patterns, but not large enough to generalize across all startups. Next, Reach Capital’s portfolio of companies are in education and the future of work. The revenue results by workplace configuration may be different in other markets.

But the subtitle of the article is:

> Data shows that pre-seed and seed startups with employees showing up in a physical office have 3½ times higher revenue growth than those that are solely remote.

There is no world where "data shows" or "have" is the correct presentation of this size of a sample. Hopefully the HN mods caveat the title of the post here.

[freitzkriesler2]

Yup a classic case of correlation != Causation

And too small a data set.