What does a daily human productivity curve look like?

February 4, 2009 9:51 PM

What does a daily human productivity curve look like?

I'm interested in trying to create some mathematical models for productivity at work. I'm looking for a basic formula with parameters that can be varied. I'm thinking the basic shape should be the same for most people: relatively high start with the peak near the beginning then steadily decreasing as more hours are worked. I imagine this is generally true, but the parameters will vary according to the individual and situation.

I'm not trying to justify working less, but rather to find an optimal way to organize effort to meet output requirements, or determine if they are infeasible. I'm mostly interested in the shape of the daily curve, but information about the cumulative effects of hours worked on productivity over weeks or months would be useful too.

I feel like there must be a lot of hard research about this out there, but I'm having trouble finding it.

posted by Cogito to work & money (13 answers total)

I think it varies wildly. For me it basically increases and increases until I go stop.
posted by aubilenon at 10:05 PM on February 4, 2009

Mine is a very slow upslope, starting with first cup of coffee and increasing painfully slowly. The first half of the day is pretty useless, as most energy is expended on staying upright and not drooling. 10-12 hours later I am a functioning human adult capable of productive work.

Then I'm peak productive for five or six hours... then a very sharp crash over the next few. And then I'm done. Lather, rinse, repeat.
posted by rokusan at 10:22 PM on February 4, 2009 [1 favorite]

It would be really hard to quantify productivity if the work requires any creativity, out of the box thinking, or critical reasoning skills. How can you measure how productive I've been if I spend two hours learning a new concept, but then afterwards since I know it already, I'm just applying the same thing over? In the first two hours there is no output and zero productivity; but the following output is impossible if those initial two hours aren't invested.

If I was a bricklayer or assembly line worker, it's considerably different. But I am a computer-type developer person. If I could make any generalizations about my productivity, I would guess that I have constant short bursts of high-brain activity, interrupted by huge drops when I either get frustrated, bored, or interrupted.
posted by meowzilla at 10:53 PM on February 4, 2009

I'm sure it's different for different people, different for different tasks, and different for the Cartesian product of those two sets. I mean I can play video games probably all day but I can only do pull ups for like 30 seconds.
posted by delmoi at 10:53 PM on February 4, 2009

You might want to look into the Yerkes-Dodson law, which relates arousal levels to performance. Muzak was developed to manage arousal levels to optimize worker productivity and safety based on this idea.

"Muzak observed that natural levels of arousal are never static or consistently varied, but rise and fall throughout the day as well as over fifteen minute cyclical periods. In response, Muzak arranged programs according to a “Stimulus Progression,” varying musical energy levels over fifteen-minute segments that would be followed by either a thirty-second or fifteen-minute long period of silence, depending on the subscriber’s desire. The length of the Stimulus Progression enhanced productivity by creating more distinctly delimited breaks between spurts of work activity."
posted by imposster at 10:54 PM on February 4, 2009

Like Aubilenon, I snowball. Starting is difficult/slow, but assuming it's not physically tiring work, I just go faster and faster until I am done. Something that might take me an hour to write when I start working is likely to be finished in 15 minutes at the end of the day.
posted by jellywerker at 11:17 PM on February 4, 2009

I remember reading somewhere that people generally take an hour to get going, do their most productive work after that and slowly tail off after lunch. But I can't find anything to support that. My google-fu is failing.
posted by BigCalm at 4:51 AM on February 5, 2009

While I believe that people take a while to get going, I can't really believe that anybody could really have a steady increase throughout the day before a precipitous crash at the end. One may perceive that, but I'm betting the reality is different. Fatigue has to start playing a role by 6 to 8 hours, and that would only increase with continued work. I'm especially interested in cognitive work (i.e., programming).

Are there really no measurement studies about this sort of thing?
posted by Cogito at 6:15 AM on February 5, 2009

There's quite a few measurement studies, but the days when they were en vogue was the early part of the 20th century, and as search, not on google. I can find a few papers/books referenced that would probably tell you what you want to know, but most of these involve repetitive factory work. (Cognitive work, like programming, is extremely difficult to measure regarding productivity. Lines of code doesn't really work as a measure when you're comparing say, perl and cobol, and doesn't take into account complexity of task in any way).

This Day of the week productivity paper isn't really what you're after, but will probably contain some references to other papers that might be what you're after. The research done by Vernon looks like what you're after (Vernon, H. M. (1917b) ‘Statistical Information Concerning Output in Relation to Hours of Work’, Ministry of Munitions, Memorandum Number 12`). BUT, well, it's a 1917 study. You won't find it anywhere!!

The problem with this sort of question is cutting through the million personal anecdotes and motivational tips that pervade the internet to get to the academic information. Good luck, but I have a feeling that you'd be better off looking in a textbook rather than google.
posted by BigCalm at 10:18 AM on February 5, 2009

Oh, and this might be good - The mythical man month (programming related)
posted by BigCalm at 10:29 AM on February 5, 2009

Sorry, that's the Japanese version, here - http://www.amazon.com/gp/reader/0201835959/ref=sib_dp_pt#
posted by BigCalm at 10:32 AM on February 5, 2009

I can't really believe that anybody could really have a steady increase throughout the day before a precipitous crash at the end. One may perceive that, but I'm betting the reality is different. Fatigue has to start playing a role by 6 to 8 hours.

Well, if you start with your theory and seek only data that supports it, you can prove anything.

I'm quite sure that six to eight hours into my work day, MY brain is just barely getting warmed up, in that the morning is barely wearing off. All my coworkers, who have learned to come back and ask me later if they want a better answer, will support this. There's no point asking me anything before 2pm. I'll try, but that big lumpy old brain just won't work right until I've been awake for 8hrs or so.

I'm also only one person. There are most definitely real 'morning people' and 'night people'. That's not some social construct.
posted by rokusan at 3:46 PM on February 5, 2009

Oh, anecdotes, you're so much better than data... here's another one.

When I played organized baseball, I was useless for those horrible, horrible 11am starts. I was an awful player. I would have benched myself. But in night games, I hit over .400 and felt like I had triple the energy.

That fits into the same grown-up work-life pattern I mentioned way above. And that was long before I discovered the deep dark wonder of coffee, even.

So the "everyone is most productive right before lunch" theory takes another hit.
posted by rokusan at 3:49 PM on February 5, 2009