Interruptions, Interrupters, and the Interrupted
by Dave Uhlir
Mike Gotta’s continuation of a discussion on ‘the cost of distraction’, itself a continuing thread from here and here is very salient at Jabber, Inc.
Attention is our most valuable commodity, what we are attending to at the moment is what gets done. There are all manner and means of distracting me from what I am doing in the hopes of getting me to do what you need or want me to do.
One argument (to which Mike poses a plausible counter) is that interruptions are killing the productivity of the interrupted. Mike’s counter is that perhaps the interrupters are having an equal or greater increase in their own productivity as they are able to get the information they need when they need it from the interrupted.
So, while the cost equation may be in question, the trend is unmistakable: there will be more means of interruption in the future, not fewer. The key then is how to manage those interruptions?
At Jabber, Inc. we see this as a problem of disaggregated presence and attention management (a topic that seems to be popping up more of late.) We think the solution lies in presence models that account for a seemingly limitless number of rapidly changing variables very quickly to present me with just the information I need or want at a specific moment.
In a press briefing this week, Joe Hildebrand (CTO of Jabber, Inc.) got to talking about presence-oriented networking. The salient item being that if you can trust the layers of security, authentication, and transport methods below the application and data, you can begin to make some very interesting policy choices that allow information to route itself based upon an individuals availability, location, organizational role, current capabilities, subscriptions, etc.
It is complicated stuff to be sure, but it is also being deployed today to help first responders and financial traders amongst others attend to the stuff that requires their immediate attention while filtering out the stuff that isn’t helpful in the moment.
November 21, 2006 at 7:48 pm
Accurate representation of presence requires that a significant number of challenges be overcome before the information provided is trusted in the manner that Mr. Hildebrand describes. Since interruptions are the topic of choice for this posting, I’d like to comment briefly on the unique challenges they pose.
Solving the problem of physical presence in IM or other presence aware applications is only part of a complete awareness solution. Perhaps a greater concern is not whether the user is physically capable of communication, but rather whether they are willing to interrupt their current work flow to participate. Combined with physical presence, a measure of a user’s willingness to be interrupted indicates their availability. As nearly 90% of all brief conversations occurring in an office environment are unplanned, successfully managing availability can be essential to maintaining productivity and allowing valuable communication to occur.
Managers in a study of attitudes towards interruption actually reported themselves as being “interruption driven” (Hudson, Christensen, Kellog & Erickson, 2002). All managers participating in the study viewed interruption management as a part of their job and often reported benefits from appropriate interruptions. It is therefore necessary for any presence system to not exclude all interruptive behavior, but rather to “focus on making interruptions more effective (Hudson et al., 2002).
Even within the study performed by Hudson et al., however, users cited the need for interruption free time. Although common, beneficial interruptions often better serve the initiator while recipients have minimal control over the timing of information reception. In fact, after 41% of all interruptions, the recipient does not return to the task previously being performed (Oconaill and Frohlich, 1995). Hudson et al. suggest that “by making information such as client activity, location, and historical patterns of activity, etc., visible in a way that does not require vigilance and active maintenance on the part of the recipient, potential interrupters can make better informed decisions about whether to interrupt” This balance presence awareness and user availability is essential to the success of any presence system.
Many studies, including some of my own, have attempted to link contextual cues about the user’s interaction with their environment with presence information provided to other users. While promising, it has proven difficult in many instances to provide an entirely positive interruption experience. As this is running long, I will save comment on those experiences for another post.
November 22, 2006 at 3:06 pm
Thanks for the comment Brian, you raise some interesting points.
Joe Hildebrand, Dave Uhlir and I were talking the other day about how various clients handle “do not disturb”. Some blocked messages while in “busy” mode, others did nothing. Joe said he considers blocking messages a bug. If you’re busy it is the sender’s responsibility to self-prioritize the message’s importance. If you want to block all messages go offline. Your presence should have enough hints in it to cue the people who want might want to interrupt
In essence, it is a matter of corporate culture, and I agree that this is part of the solution.
Well understood organizational protocols that allow for the individual quirks associated with productive trusted relationships are important. But as you point out, so is the need for better information from presence publishers that requires very little active maintenance on the publisher’s part.
Better information does indeed allow for better decisions by the interrupters, but again as you said, it is a significant challenge to get it right all the time, which brings us back to Joe and corporate culture.
If the interruptions are not productive and not well prioritized, we have more of an HR problem than an IT problem, and no amount of well architected solution is going to entirely fix the problem until the HR and corporate culture issues are addressed.
November 23, 2006 at 10:05 am
As regards “do not disturb” I wholeheartedly agree with Joe. Automatically blocking messages does not account for recipient discretion as to what is important or not, again bringing us back to the value of interruptions. One interruption may be senseless, while the next conveys critical information. Who is capable of making the determination?
Well, the recipient should definately be able to determine whether the interruption was worthwhile or not. Unfortunately it is too late once the message has been received. The presence model is supposed to portray enough information for the sender to make a deterimination regarding the urgency of their message content. Herein we find another problem.
Senders, those ingesting the information from the presence system and determining whether or not to interrupt, are inherently selfish. Their need to convey or obtain information to/from the recipient often overwhelms the recipients perception of their own availability.
How have users come to overcome these interruptions? Well, I would contend that we usurpt the presence system and introduce a level of “plausible deniability”. We either leave the system set to “online” continuously, change our status to “away” while we are busy to remove any expectation of a reply, or we use “do not disturb” to maintain some level of ambiguity. Does this cheating of the system undermine the trust which Joe Hildebrand spoke of, I think so.
Ultimately, though, the interruptions which come in the form of IMs can easily be prioritized by the recipient and responded to in an appropriate manner. My concern is when the presence systems begin to infiltrate other technologies and interruptions begin to occur on a more physical level (such as a telephone call or drop-in visitor to the office). What will we do then?
March 5, 2007 at 7:03 am
I personally believe that interruptions are necessary sometimes in order to increase productivity. It is difficult and unreasonalbe to keep doing something for a long time. It will be better to change your activity and then come back to the main job.