First impressions count.
Users judge a site just within a fraction of a second. Many hit the back button before reading the first line of text. And of the few that decide to stick around for a minute and give you a chance, there's no guarantee they'll ever return.
HCI gives us a good understanding of how users make snap decisions about site design. But how do users decide whether or not to join a community?
According to social psychologists, users judge communities along four dimensions:
1) What is the current benefit of this community to me?
2) What is the expected future benefit of this community?
3) How much do I like the individuals within the group? (Common Bond Theory)
4) How much do I like the group as a whole? (Common Identity Theory)
Because newcomers by definition have no prior experience with the community, their decision to join is based primarily on the expected future benefits of joining. Increasing the expected future benefit is just selling. Good copy, good screencast, good screenshots, etc.
But let's assume you've been reading your Seth Godin so you're already up on the latest marketing techniques. What else can we do?
I recently saw Bob Kraut present a paper on the mediating variables for Usenet participation. The study looked at how getting a reply affected the chance that a poster would return and post again.
For oldtimers who received no replies, 84% posted again. For oldtimers who did receive a reply, 86% posted again. For newcomers who received no replies, 16% posted again.
What's startling though is the effect getting a reply had on newcomers posting their first time. When looking only at newcomers, getting a reply increased their likelihood of posting again from 16% to 26%. That's a 62% increase!
Apparently, getting a reply increases satisfaction in all four dimensions. It increases current benefit, it increases expected future benefit, it creates a common bond with the individuals who posted replies, and it increases identification with the group as a whole.
Now, translating patterns in Usenet posts into practical design advice isn't an exact science. But if I were launching a new website, here's what I'd do. Instead of hiding the Feedback link in the upper right hand corner, I'd place a form right on the main page. A big form. And I'd bend over backwards to get people to use it.
Bugs, ideas, comments, observations, advice, etc. It doesn't matter. Why? Simple.
Because by emailing you, your visitors are giving you permission to send a reply. A reply that, if crafted correctly, could dramatically increase that person's chances of becoming a full-fledged member of the community.
Now, I can't guarantee this will actually double your traffic. But all the research says that people will be more likely to continue to participate in online communities if their early interactions are successful. And what could make one feel more successful than getting a friendly personal letter from the CEO?
And by personal I mean personal. No form letters allowed. Yeah, you might have to stay awake for a week straight. But so what. If it works, it's worth it.
After all, the only cost of trying is two lines of code and a box of Modafinil.
Sources:
Kraut, R., Wang, X., Butler, B., Joyce, E., & Burke, M. (Under review). Building commitment and contribution in online groups through social interaction. Unpublished manuscript.
Ren, Y., Kraut, R. E., & Kiesler, S. (In press). Applying common identity and bond theory to the design of online communities. Organizational Studies. [PDF]
"When looking only at newcomers, getting a reply increased their likelihood of posting again from 16% to 26%. That's a 62% increase!"
what? that's a 10% increase.
Posted by: hup | April 29, 2007 at 09:35 PM
>what? that's a 10% increase.
He calculated the percentage out of the 16%. That is, the increase is from 16->26 = 10, and 10/16 = 0.625 ~= 62%
Posted by: Muhammad Haggag | April 29, 2007 at 09:42 PM
interesting although likely if you got a response as a first time poster, then your initial post was likely of high quality, and therefore you were probably already predisposed to posting again (e.g. you were more knowledgeable in the subject at hand than first timers that posted more shallow posts and didn't get responses). so, although completely agree with the conclusions, i'd guess that the 62% is a bit skewed.
Posted by: gzino | April 29, 2007 at 11:48 PM
Great post. I'll definitely keep that in mind as we roll out the beta for Avanoo(the startup I work for).
I was kind of expecting a free lunch from this headline, but, just like with getting blog readers or anything else, the more effort you put in, the more you're rewarded. Of course writing thousands of personal replies might be too much for anyone... so efficiency becomes increasingly important.
Posted by: Ilya Lichtenstein | April 30, 2007 at 12:18 AM
@gzino: Great point. The research shows that the wordcount of a post is very highly correlated to whether or not it gets a reply.
@Ilya Lichtenstein: I checked out your site. That is basically exactly what I had in mind when I wrote this post. :-)
(That is, pre-beta sites with an "Add Me to the Invite List" box on the front page)
Posted by: Alex Krupp | April 30, 2007 at 12:31 AM
It's modafinil.
Posted by: Jake Spoon | April 30, 2007 at 12:36 AM
@Jake Spoon: Good catch, fixed.
Posted by: Alex Krupp | April 30, 2007 at 12:44 AM
Very interesting Alex. Makes sense in an age of transparency. Definitely food for thought.
Rich
Posted by: Richard Giles | April 30, 2007 at 04:05 AM
I have implemented a plugin in my blog that automatically subscribe commentors to receive new comment alerts (evil, eh?) and this has increased visitor return frequency.
Posted by: Kashif | April 30, 2007 at 04:44 AM
I use redbull, but yes, I try to reply to all new users on Terapad.com. Feedback is really good so far, and I've even made a few new good friends :-D
Posted by: Stephan Tual | April 30, 2007 at 11:59 AM
One problem that blog comments have is that they limit the posibility of, and the mechanics of checking for, a personal reply.
Since replys on blogs are generally simply linear in their display it is difficult to reply to a specific post or poster (with users resorting to workarounds like mentioning the name of the poster ex. "@Alex Webb - blah, blah") and it is difficult to monitor potential replies to your comments as they could be buried far down the linear list.
If comments on blogs were "stickier" in a number of ways (see how reddit.com handles replys and notification... very effective) then the useful points made in this post could be applied even to the "lowly" blog.
Posted by: ST | March 15, 2009 at 02:32 PM
As a quick addition, I just tried to post the above reply and could not do so without creating a Typepad account. This is an unnecessary hurdle that makes the "community" of this particular blog less appealing to me. In my case I created a quick, one-time use junk account simply so I could contribute what I think is a useful comment on a useful post but next time I am here (assuming I am) I will need to create another one just to participate... makes it less appealing to me (and it may be that the mechanism that allows non-typepad people to post was not working). Either way it limits the community building possibilities.
Posted by: ST | March 15, 2009 at 02:36 PM
You did it with only one line of text:
"Double your userbase with two lines of code and a box of Modafinil"
I guess quite a few druggies were drawn to your blog...
Posted by: Roger Josek | April 18, 2009 at 05:59 PM