Saturday, September 11, 2004

An email crisis?

How much spam lands in your inbox? A lot? A little? How important to you is email as a form of communication? What is the fraction of important email (email you must read) compared to the rest of email you receive? What strategies, if any, do you employ to increase the signal-to-noise in this environment?

In qualitative research Judd, Peter Lyman and I did over the summer, it became obvious that an email crisis is enveloping average computer users. Email as a resource is becoming more and more "inefficient" for many users while also becoming more important as a communication medium; people rely on email at home and at work to facilitate important, sometimes even critical, communication. The amount of spam and simply unwanted mail is also increasing, as is the sophistication of spammers; often fake emails appear to be genuine. The tension between these two - the growing importance of email as a communication medium and the increasing amount of "rotten" email - has combined to frustrate both sophisticated and non-sophisticated computer users. Participants in our study made it clear that email is starting to become inefficient, to different degrees, amongst all types of computer users.

We noted a variety of behaviors people used to filter email. The most simple was to use a piece of software - like SpamAssassin, etc. or bayesian filtering tools in a mail client - to filter spam. Other tactics arose like managing many different email addresses used for specific purposes or roles. Naturally, some of these many email addresses were valued moreso than others: people only give some email addresses out to certain people or via certain channels. Some people still read each email and have their own means of deciding which emails are junk. Some people even advocate not using email at all; which eliminates spam for sure, but also excludes some from communication via email entirely or requires building a parallel, seperate electronic mail network.

What do you do to filter email? Not just software, but behaviors you employ to actively increase the signal-to-noise ratio of the contents of your inbox. Do you see evidence of this email crisis in technologically less-sophisticated family and friends? Has a valuable email address of yours from the past become essentially useless due to spam? Do you check it anyway to ensure you don't miss old friends?

What things do you do to filter mail that might not meet a formal definition of "spam"? For example, I have set up a filter that routes every email sent from a certain person directly to the trash. This is because this person sends completely worthless (to me) email.

If you want to know more, check out "Unsolicited Bulk Email: Mechanisms for Control" by Hoffman and Crocker of the Internet Mail Consortium.

2 Comments:

Blogger Calvert Jones said...

I find the research that you and Judd were doing over the summer very interesting, and I was especially intrigued during the last class by Peter Lyman’s comment regarding the ensuing death of email, as a result of its growing inefficiency. Did you two find, during the course of your research, that people were seriously dissatisfied with the filtering tools they had? I know you said “average users” were overwhelmed by their inboxes, but was that also true for the people who were actively filtering? Are average users developing strategies like filtering to deal with the crisis, or are average users largely unaware of these tools? How did you define an “average user”? Who did you interview?

Another interesting dimension to this email crisis is the perception, especially in universities and the NGO where I worked, that email is a kind of infallible communication medium. People might not get your letter, your memo, even your voicemail, but how could they not get your email? Does anyone else think that there is this perception in certain circles? If Peter is right, that email as we know it is dying, then which communication instruments will be the next most favored ones? How will our over-reliance on email affect our use of other traditional communication channels?

2:29 AM  
Blogger Joseph Lorenzo Hall said...

Damn, Cally... that's a mouthful... I'll try and answer things one at a time.

People did tend to use filtering tools that were built-in to their email service... but they didn't do much more (like blacklisting or using Bayesian tools... at least not that they knew of). Many of the participants used multiple email addresses coupled with keeping one or two sort of secret to limit the amount of crap they get. This is a real behavioral adaptation. There's also the other forms of coping... scanning the emails and deleting ones that don't have something to do with them... mass deletions of emails regardless of their content... etc.

Some of our pilot-study participants (SIMS students) did have elaborate, well-maintained and well-understood filtering mechanisms. I don't know if we can say anything meaningful about whether or not this part of our population found email frustrating or not.

Users are aware of the tools but don't really understand them more than superficially... and get annoyed when email they wanted is tagged as spam (they tend to think that it "doesn't work" when this happens). So, they invent their own ways to filter and adapt... we'll need to expand our sample considerably (from 11 to, say, 50) and focus on this in part to get more deep information.

Filtering can actually be an excuse for losing an email... so can mass or accidental deletion.

I'll leave the speculation about the future of email-mediated communication for others.We didn't define the "average user"... since our study was just to "test the waters" so to speak with this methodology and subject focus, we tried our best to get a good assortment of different demographics. Here's that part of our current working paper (I should mention that the participants were of various diverse ethnic and racial backgrounds):

Sample

A sample of eleven participants was recruited for the study
from within the social networks of the researchers.
Participants were chosen based on their willingness to
participate in the study and their being:
• between the ages of 18 and 35;
• not an ‘information professional’; and
• not affiliated with our institution.

[...]

The participant group had a mean age of 26 and was split
evenly between men and women. Ninety percent of
participants reported some college education – sixty percent
some graduate education. Participants were of diverse
incomes, and greater than seventy percent were employed
in a variety of careers ranging from schoolteacher, to city
planner, to line cook.


By and large, the participant group was familiar with
computers and the Internet and reported involvement with
computers on a regular basis. All but one participant owned
a computer, and participants had, on average, owned a
computer for 13 years. Daily computer use varied between
one hour and nine hours with a mean of four hours. All
participants described feeling either ‘somewhat
experienced’ or ‘experienced’ with computers.


All participants also reported having a personal e-mail
address, and nearly 90% reported having more than one.
The participant group consisted mostly of active e-mail
users, with two-thirds of e-mail users indicating they check
their e-mail several times a day, and one-third reporting
hourly e-mail checking. 78% percent of participants
indicated that on average they send more than ten e-mails
each week.

10:58 PM  

Post a Comment

<< Home