Mapping Internal Communications

Patrick_Keogh writes "This article in The Economist discusses some research work from the Helsinki Institute of Physics which confirms what Scott Adams knew, and a lot of us suspected, that nobody talks to their boss. The research uses some novel mapping and visualisation techniques to map the communication interactions within a large engineering organisation." It's an interesting idea, but I would guess that it can't capture verbal communications very well, and that seems like a major flaw.

Comments

tracking e-mail skews results toward islands

[kopriva]

I assert that analysis of a dataset that results from the tracking of communication by e-mail (exclusively) will preferentially give you the "islands" of small engineering groups in the map of communication flow.

When an engineer has a question for someone in his group, the question is usually a short one with a short answer. Furthermore, both engineers will have the necessary context for the question and the answer already, and therefore will not require lots of conversation to establish the context. If the engineers are not in the same room at the time that the question arises, this is the ideal scenario for a quick e-mail exchange (e.g., Q:"Is the default level for GOBHP high or low?" A: "low")

However, if you have a question for someone outside of your group, you probably need to spend a little (or a lot) more time and effort establishing context. Also, you might need to do a little more personal introduction (e.g., "Sorry to bother you, but I work in the XYZ engineering group and we're using your ABC tool to do..."). This is the kind of situation in which people tend to get up from their chairs, walk across the building, and sit down for a ten-minute chat with a fellow employee.

Even beyond missing verbal communication like this, tracking e-mail misses the importance of technical documentation. The study did mention that it tracked what files people downloaded, but I doubt that this really captured the flow of technical documentation, since much information is _still_ exchanged in paper form (gak!).

I am a technical writer for a medium-sized chip manufacturer. I am subscribed to virtually every company-internal technical mailing list. I spend weeks or months in face-to-face (and face-to-whiteboard) conversations with engineers in one group so that I can put as much information as possible into a document that is then made available to the entire company. If I do my job right, then I am acting as a narrow but very-high-singl-to-noise-ratio bridge between the islands of engineering groups. I don't see the Finnish study as capturing this quality of information exchange.

Finally, meetings --- both between boss and subordinate and between larger numbers of engineering peers --- can be opportunities for a huge flow of information between these islands. (Yes, meetings can also be black holes from which no useful work can emerge, but that only seems to be the case around here when one of the non-engineering executives is involved).

In short, I think the Finnish study misrepresents the real situation. At least in my company, we have a pretty open flow of communication between (not just within) engineering groups.

When I was boss, I used ONLY verbal

[OlympicSponsor]

....for my underlings, anyway.

There was no grand scheme, I just liked talking to them (and they to me, I think). Oh sure, every once in a while I'd send an email--if they weren't in that day and I didn't want to forget the question or whatever. But mainly I just stood up and walked over there to find out what was going on.

The informality allows you to learn a LOT more. Facial expressions and body language are often a more accurate report of the actual status. Plus the person (and the people in the nearby area) will bring up other topics that you also need to know about. Furthermore, the conversations were audible to the other programmers, so THEY got an update as well--making us a more close knit team. (the noise/distraction factor wasn't important in this situation).

As a contrasting example, we programmers were almost entirely isolated from the rest of IS/IT, making our interaction with THEM very weak.

So I propose a small refinement to the model: I imagine their "communications" already include email, telephone and scheduled meetings. Now add a factor that measures physical distance to calculate the probability that the boss and employee talk face to face. If the distance is only a few meters, add a bunch of "face time" to the communication list. If the distance is thousands of feet (or more) add little if any "face time".
--
MailOne

Almost all verbal

[ajs]

Only the most formal of communications (usually CC'd to others) go to my boss via email or paper. The spoken conversations with my boss are usually the most productive and communicative.

I think it's different in the software world than in the pure Web world. In software, you're usually doing much less trouble-shooting, so email can be a better mechanism. On the Web, you've got a user having a problem, and you're expected to fix it in real-time, so email is just a little too slow.

I have yet to take the plunge and use IM for everthing. Jabber may change that for me....

More like Twister

[laetus]

I think the researchers got it wrong. Most communications inside of companies are like the game of Twister. Depending on the color spin of the day, you never know who's going to be in your face or who's going to have their foot up your ass.
----------------------------------

Cultural Affect on These Results

[mellifluous]

It would be interesting to see what kind of data this system returned in other cultures. I currently work for a Finnish company, and one of the things they told me when I got here was that "If a Finnish manager isn't talking to you, they are happy; but if an American isn't talking to you, they are probably angry." That is certainly an oversimplification, but I do believe that common management styles can vary drastically from one culture to another. This techinque might help to codify and visualize these trends.

It's the SNAFU principle

[Greyfox]

Robert Anton Wilson wrote about this thing ages ago in "Illumanatus." It's why hiearchical organizations become increasingly detached from reality as time progresses; communication can only happen between peers. When you're talking to a superior, you'll tell them what you think they want to hear. The longer the chain of command is, the more detached from reality the people at the top will be. And while Illumanatus is fiction, Wilson IS a perceptual psychologist and would probably know...