Hacker News Reader

Comment by vosper | Hacker News Reader

I think one of the most important points in the article is that forcing people to take vacation, and encouraging them to be completely unavailable, means that the company must learn how to survive the absence of a team member. It forces you to reduce the bus-factor, and this is critical.

I ask my team members to take at least 20 days off per year (we also have an unlimited vacation policy) and I'll be using this argument to encourage them to switch off and enjoy their breaks.

View on HN

Just to be that guy ... it actually _increases_ the bus-factor (which is good). The _higher_ the bus-factor the more people you can loose before your company goes under.

Pedantic niggling aside, I completely agree this is a huge benefit to the organization.

How do you do this without creating a bloated organization?

For example if you have 20 team members and expect for 2 to be gone for some kind of vacation at any point. When all are at work you have a 10% inefficiency.

I wonder if it would be more effective to have have stretches of intense work (4-6 months) with long rest periods (1-3 months) between. This would minimize the operational switching frequency, though it would take longer probably.

I manage a software engineering team, and there's always more work to do - we have a big icebox full of things we wish we could get done. I imagine this is pretty normal in my field, and I can't imagine that we'll reach a point where we're overstaffed when people aren't on vacation. For other teams it may be the case that you end up overstaffed; I don't have much experience outside of my field.

I, personally, like your suggestion of stretches of intense work followed by long breaks, but I think that the reality of family life, hobbies, interests, and other commitments make that pretty unrealistic to impose on a team. You'd also have to be very committed to making sure that the work periods are strictly bounded, and it could be very tempting to slip and say "just another few days" / "just another week" and end up working to breaking point.

Edit: Grammar

If a company only has 10% inefficiencies, they are doing a fabulous job.

Most companies probably have 50% of their workforce surfing the web at any one time.

Well the former is structural, the latter is behavioral, so there is difference in kind. Said another way, you can't get work from a ghost, but you can find incentives/realign people so they are more productive.

> but you can find incentives/realign people so they are more productive.

By giving them more vacation maybe?

I was thinking more along the lines of having them do work they are interested in/care about.

Or quit thinking of them as resources or machines that were advertised to you as 100% efficient for 8 hours 5 days a week and realize that the 2 hours in a given 8 hour day spent reading HN or browsing YouTube _are_ part of the 100%. No one can walk into an office and be at 100% efficiency within microseconds. Windup and spin down time is part of the cost of running an organization. Not only is there windup/spindown at 9AM and 5PM, but at 12PM and 1PM, and probably at around 3 or 4PM in addition.

If someone is only putting in 2 hours of efficient work a day, then maybe you should start thinking about shuffling them within or out of the organization depending on the position and level they're filling, but asking about 10% inefficiency due to "missing bodies" is equivalent to complaining about getting shorted a penny from the barista. It's literally pocket change, get over it.

If you ask them, most developers will have something about the current codebase they would like to improve, or a 20%-like project that might improve the products. You can have them do this during times of over-availability.

Inefficiency in the short term, yes, but it still leaves you in a position where a single surprise (an accident, a firing) can erase _all_ the efficiency gains.

It's the same way that buying extra hard drives for backups is "inefficient," since you rarely use those drives, but still far more efficient than recreating all of your lost data from scratch.

I like the idea of personnel redundancy, but my guess is employees don't like that and it's not really best for tackling tough problems because you increase your workforce 2x.

Why would you need 2x redundancy?

Surely in a team of 10 developers working on the same thing, if one quits the others can take over the work, just with a schedule slip?

You are conflating hours with productivity.

Let's assume that no engineer is getting less than 2 weeks of vacation off a year on paper. Most of these articles are written by people giving 4 weeks of vacation. So we're talking about the difference between a 'stingy' vacation policy and a generous vacation policy being 2/52 weeks in the year.

If people gain more than 4% productivity by avoiding burnout, you net win by giving people more vacation. I would argue that net productivity gain from letting people de-stress is way more than 4%, though I have no numbers to back that up.

Agree 100%. I am talking about giving people 1 month off per year. In those cases, how do you maintain efficiency without bloat?

How are you defining efficiency? Do you think it smart to have all of your staff 100% working at all times? What happens when new work comes?

What's wrong with bloat? Isn't profitability the goal, not efficiency? If it's best achieved by having more resources so you continue to be profitable when one resource is unavailable, then have more resources. This is as true of hard drives and fire exits as it is of employees.

Expanding on a point by brational, see Tom Demarco's book Slack: Getting Past Burnout, Busywork, and the Myth of Total Efficiency. It makes the argument that if your goal is long-term profitability, you must have some excess capacity in human resources at all times. http://amzn.com/B004SOVC2Y

Slack is a fantastic book. If you're even the least bit unclear about why/how TFA thinks that minimum vacation is a good idea, then you'll find this a fantastic read.

I've also learned to appreciate slack when learning about Kanban:

http://www.everydaykanban.com/2012/07/27/slack-is-not-a-dirt...

I was expecting some kind of subgenius material. Please tell me JR "Bob" Dobbs is at least mentioned.

The only way to be able to quantify "10%" is if you have the ability to predict the future perfectly. If you don't have any stockpiled bandwidth you will have big issues if anyone leaves the company for any reason, and you will also not be able to easily take advantage of any unexpected opportunities that come your way. I wouldn't call it a bloated organization, more like a flexible organization.

That mean's it's NOT bloated. If you have a bus factor of 1 anywhere, that means your organization is in pretty significant danger. People die, get sick, quit, or otherwise get unable-to-work'd ALL THE TIME.

No mechanical system is ever exploited at 100% capacity. The elevator in your building never has to carry more than half of what it's absolutely capable of. Any kind of building is built with materials that can withstand a lot more than the average load. Your computer never has to work at full capacity - I doubt the CPU's temperature ever reaches past 70C, far from the ~100C at which it can still perform reliable calculations, but also well clear of the ~105C at which the magic smoke escapes.

Smart people don't exploit systems operating at their absolute limits.

In general I would agree, but I think you've gotta be careful that doesn't come off as "we need to make sure you're not a vital member of the team".

Reply on news.ycombinator.com