Week 1 or Comparing Disruptiveness of Communication Methods
So my first trainee week at Openismus is over, and it has been a pleasant surprise.
Day 1 started with a nice team-building exercise where David and I could show our understanding of real Office Origami: IKEA(TM) furniture assembly. Once that was done we were ready to set up our newly arrived PCs. I appreciate that we were offered a choice over the keyboards and such - those are the details that count! Since I could not figure out how to make it boot from a DVD first (press ENTER, d'oh) I lost 5 coolness points. I lost at least 7 points more for initially refusing to use a 64bit Linux Distro. But that's OK, really - I plan to make my personal coolness points wrap around at -2^31 some day.
The remaining 2 hours of the day I spend with customizing my desktop on that box, and I think I see a pattern here: usually the first three things I do on Ubuntu are
- installing Epiphany (together with customizing its toolbars to only use one line, as it was in ancient versions),
- deleting GNOME's bottom panel (the workspace selector fits well in the top panel),
- reconfiguring the GNOME shortcuts. Honestly, who wants to SHIFT+CTRL+ALT+ARROW_KEY just for moving windows around? Once the Windows(TM) key is properly configured we can get from 4 to 3 keys here. Using the Windows(TM) key for managing windows - I'd like to see that becoming a standard in GNOME.
Day 2-4 was all about compiling GNOME from scratch. For me it felt like using Gentoo again, only that building from trunk versions breaks far more often. But the rest - missing dependencies, poorly written Makefiles, outdated version requirements for build tools - it's all the same (Disclaimer: I don't blame jhbuild, it does a good job, and jhbuild shell is awesome).
Over all these days I enjoyed the friendly atmosphere in the office itself. The common breakfast, the chit-chat at the coffee machine, and - most importantly - no phone calls. Which brings me to what I originally wanted to blog about:
Comparing Disruptiveness of Communication Methods
It goes without saying that among IRC, Instant Messaging, E-mail and phone calls, the latter is the most disrupting mean of modern office communication. Internet Relay Chat (IRC), if used mostly in a passive way (e.g., only answer after being pinged, and only in defined intervals), could then be the other end, being the least disruptive. If we were to include face-to-face communication here (which we don't since I feel I have not much to say about that), it would probably top phone calls even. A study from an IT support organization (a citation from [this paper] (http://jcmc.indiana.edu/vol13/issue1/garrett.html)) "revealed that workers spent an average of just 11 minutes on a task before being interrupted or moving on to a new task, and more than half the interruptions (57%) were unrelated to the task at hand" [1].
So if the frequency of interruptions is given, is there at least a choice in disruptiveness? Let us define disruptiveness as the costs to switch back to the previous task after an interruption, and let's say the Loss of Productivity caused by interruptions equals to
#Interruptions x (Usage_IRC x Costs_IRC + Usage_IM x Costs_IM + Usage_E-mail x Costs_E-mail + Usage_Phone x Costs_Phone)
It then becomes obvious that we want to reduce disruptiveness of each communication type but also the usage of the more disrupting types.
Given the nature of phone calls there is little one could do. If the phone is an irremovable part of the workplace (helldesks, for example) one would probably want to choose a pleasant ring tone, or make it use visual alarms only. Another solution could be to filter every call with an answering machine, and then checking for voice mails only once or twice per day. While this obviously defeats the purpose of phone calls (high priority, synced) it may still be effective under some conditions. Another not-so-obvious part is that the recipient of a phone call has no estimate about the length and the type of information (in-depth with note-taking, simple notification, ...) of the call once she decides to answer. One could postulate that the recipient is almost always unprepared, hence resulting in high disruptiveness.
E-mail immediately solves 2 problems, compared to phone calls: async by default and note-taking (and because this topic is so complex that some make a profession out of it, you can haz your Wikipedia page). However, reading E-mails can be costly. Once the mail comes with attachments it requires additional tasks to be performed by the recipient.
IRC can probably do everything E-mail could do for you, though you first need to find the right networks and channels. It has often been said (I am glad this is only my blog and not Wikipedia, else you would read the dreaded [C... n...] here) how useful this approach is for giving support, since those seeking support would naturally find those that can give support by themselves. Also, the matching that happens between the two groups is random, meaning that the support-seeker is usually not addressing one person directly. Support-givers can then decide whether to interrupt themselves (by answering) or not (by waiting for someone else to answer). In that regard, the disruptive costs for support-givers would be nearly zero, amortized.
Now how about IM? One could think IM is just as disruptive as phone calls: A new message arrives, notification alert pops up in the center of your screen, perhaps even a sound is played. Imagine that happens for each of the gazillion conversations you are currently in. On top of that, you probably only chat with your friends and family anyways.
Luckily we can choose to only use sanely configured IMs, so notification alerts will all be combined in one icon in the systray, without any bubble boxes or sounds. What we get over IRC is improved tracking of the people we want to talk to. We can check their status before sending a message (online, busy, chatting), and that is an important part! If we care to not disrupt our co-workers overly we would naturally send messages only if their status indicates they are online and not busy. Also, the signal-to-noise ratio is dramatically improved, compared to IRC. The paper - "IM = Interruption Management" - which I already mentioned above - concludes that 1) IM users will report lower levels of disruptive interruption than non-users, 2) IM users will have the same overall level of work communication as non-users.
So if you see your sys-admin at lunch tomorrow ask her why IM (or IRC) is filtered at your company, but phone calls are not and whether she pays for the resulting productivity loss in your company from her own salary.
[1] González, V. M., & Mark, G. (2004). "Constant, constant, multi-tasking craziness:" Managing multiple working spheres. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (pp. 113-120). Vienna, Austria: ACM Press.
Comments
No comments yet.
Post a Comment
This site's webmaster failed misserably in upgrading the underlaying web framework.
The comment system is entirely broken right now. Sorry!
