Hacker News Reader

Comment by joegahona | Hacker News Reader

I would love to know how they are measuring the success of this measure -- and I'm not being cynical. How would one even begin to weigh the tradeoffs of closing business for everyone for an entire week against the increased productivity upon return and something woolly like "improved morale." What if giving everyone a week of twice a year was actually beneficial to the company? Or even quarterly? Many companies already have an unspoken/informal week off between Christmas and New Year's.

Also, this is incredible Marketing for Bumble, as many people are going to share this in Slack channels in an effort to get their companies to do the same.

View on HN

Some things can't really be measured numerically, I think this is one.

I think you do things like talking to people afterwards and see if they think it was a good idea, looking at attrition rates and try and guess if it moved them, looking at whatever productivity metrics you have, and at the end you probably have a gut feeling on whether or not it was a good idea... That's probably the best that you get.

I think the argument would be that this actually could be measured short term (probably all the staff will say they feel better in two weeks) but if in Q3 the CEO realizes they need to roll out some features before a Q4 quarterly call then what could have been a 12 week project will turn into a cramped 10 week project.

Yea, this is what always gets me about taking personal vacation time. You take a week off, but that week isn't yours for free. Your expected work keeps accumulating during that week. You're just taking a week of time-debt which you will need to pay off in the form of much higher workload over the following few weeks. I wouldn't be surprised if these employees all get back to work and find their workload doubles for a time just to catch back up!

I think good management can turn it into a real vacation. Our teams works in scrum and we just commit to fewer or smaller tasks, accounting for the missing person-days.

And one major distinction is that it didn't seem like they got to pick the time?

Maybe I really like working during the summer but have lots of family stuff in the fall. Seems a bit inefficient for this to be so long but I'm all for experimentation.

I hate it when people demand that every decision needs to come with some metric. For one, they fall victim to Goodhart’s Law:

> When a measure becomes a target, it ceases to be a good measure.

Metrics should drive intuition, not the other way around. We don’t use the scientific method here.

You may also want to cite the McNamara Fallacy:

https://en.m.wikipedia.org/wiki/McNamara_fallacy

Just to note, you can regularly ask your employees how they feel, and with enough employees get a significant measurement. So maybe they can measure the “woolly” morale impact of this. Tools like Lattice make it easy to do a “pulse” survey for this purpose.

Also the great thing about doing this twice a year is you get two experiments so you can start to be confident it’s not just “Christmas makes employees 10% happier” if you also measured a similar bump in the summer break. (Obviously numbers made up for the sake of example.)

That assumes that your employees:

1. Know how they feel

2. Can translate that into numbers accurately.

3. Are willing to express how they feel instead of how they think they should bubble themselves in on a web app you pointed them to.

Nothing is perfect, this is better than nothing though. [1]

The way my company does this is they ask a lot of questions and then mostly look at aggregates. E.g. they might ask 6 questions about "how do you feel about the size of your workload", "do you feel you have time to do the important things in your life", "are you satisfied with the amount of vacation you take", "do you feel stressed", ... and then roll those into a single "work life balance" metric (questions and specific metric entirely made up).

And actually we outsource to a survey company that does that for us, which apart from outsourcing the non-core competency, means that there is a reasonably neutral party deciding how to aggregate things (I think, I'm not involved in the survey design).

[1] Therefore this is better than perfect?

You have to consider the alternative, which us not knowing much at all. This is not a precise science

Psychological science uses this technique a lot. It's not perfect, of course there's lots of noise because of both of the issues you raise (in addition to other issues like how to develop questions to measure nebulous things like "self-esteem" or "extraversion").

But as long as the results aren't pure noise, i.e., there's some signal there, then with an adequate sample size you can do something with it. Lots of fields deal with noisy, imprecise data. It's just conditional on it not being only noise.

It's a good question, which has been extensively studied over the last few decades in the medical context. The industry phrase would be "Patient Reported Outcomes" if you'd like to explore further.

After many papers and lots of clinical trials, it's widely accepted in the scientific community that the answers to both of your questions are "yes". E.g. see this FDA paper[1].

Of course, it's possible to design your questionnaire badly and there's an art to this, so it's right to be skeptical/cautious about putting this into practice. But you seem to be getting at a more fundamental epistemic question about whether the approach is possible even in principle, which has been thoroughly investigated.

[1]: https://www.fda.gov/media/77832/download

You're getting it backwards. What if you go to work 7 days a week, and the numbers are still going down? Bumble is taking a week off because they're doing poorly. I have no opinions on whether the factors are in their control or not.

At a (perhaps generous) average of $120K per employee, a week off would cost $1.5MM - given how broadly this will get picked up and circulated in the media, it'll get more eyeballs than 1.5MM spent on mobile or search ads most likely.

> Also, this is incredible Marketing for Bumble, as many people are going to share this in Slack channels in an effort to get their companies to do the same.

Agreed and ppl sharing stuff of slack channels are their target customers too.

If the effects are minor I think it will be difficult to measure the impact on any one company given the number of confounding variables. In that case the only way to determine the correlation between these kinds of things and business success might be to gather enough statistical evidence across industries.

This was my reaction, not to mention, while everybody takes the week off, the work will pile up with nobody working.

Nothing like coming back to the office after a week off and you're already back up to your eyeballs in work, the moment you step back into the office.

Reply on news.ycombinator.com