Q. What’s the best way to market social networks?
A. Other social networks of course.
This past week, Anderson Analytics was beta testing a couple of new applications on Facebook, one of which is “Compared to Me”. Marketing Guru, Guy Kawasaki, was kind enough to offer to tweet about it and this past Friday the 13th at about 1am his tweet went out. I was curious to see what effect it would have on usage of the application. After reviewing some of our usage statistics it looks like we got the Guy Kawasaki Bump! (See Canvas Page Views Below)
Of course it looks like Guy’s initial bump will be followed by additional Twitter bumps as his followers re-tweet his original tweet. Last time I checked, already 6 of his followers had re-tweeted the link to our application. So it looks like this initial bump will be followed by even more smaller bumps (which will look like one large bump in our data).
Web2.0 is truly fascinating. On the one hand the opportunities it provides for tremendous new deep insights for market research and secondly learning how marketing is changing.
For anyone interested in the “Compared to Me” application mentioned above you can access it here:
http://www.facebook.com/apps/application.php?id=43439829009
Thank You Guy for the Tweet!
Tom
![[Post to Twitter]](http://www.tomhcanderson.com/wp-content/plugins/tweet-this/icons/tt-twitter-big1.png){kind=icon}
6 responses so far ↓
1 Susan RoAne // Mar 15, 2009 at 11:58 am
Best thing about the Kawasaki Bump is Guy himself. In a world of self-important divas and divos, he is one smart, savvy, down to earth great Guy. I’ve been his fan for years.
Got a bump from him myself when he blogged about how to schmooze based on my tips from How To Work a Room®. It worked!
2 wwwhitney // Mar 16, 2009 at 10:51 am
Hey Tom, I suppose I’m contributing to the Kawasaki Bump Discussion Bump as I just learned about your app from this post.
Do you have any data on when respondents’ statistics start to stabilize? I answered 15 questions initially and then after reviewing my ratings, I answered about 10-15 more questions and I went from having an above average rating in almost every category to having a below average rating in each category.
As it stands, I think most people won’t answer enough questions to provide real data on their personalities. You may get good data on the aggregate population, but it won’t be as useful for the individual user unless he or she is encouraged to answer many more than 15 questions.
Anyways, I thought it was a very novel idea for a personality test and would be especially great if you’re able to sort the data by demographics.
3 Tom H C Anderson // Mar 16, 2009 at 11:03 am
Hi Whitney,
Yes, we have experimented with question randomization formula. Originally we weighted the first 5 questions as more important requiring each user to answer the same question at least twice. However as we started receiving more data, and due to complaints about getting the same question multiple times in beta we switched to a more fully randomized weighting formula. The model will get more accurate as more data is collected, but as you point out, the aggregate data does provide a good picture already.
There are two output tabs available as of right now in this application. One is a personality overview, the other is personal level question detail. I think the latter will encourage users to answer more questions for more accurate results as you obviously noticed. While early Facebook Applications required as many as 50 answers prior to showing feedback, and also required you to invite several ‘friends’, this type of forced usage is no longer acceptable to Facebook and the average Facebook user. These are some of the things we have learned as we’ve developed applications on Facebook and elsewhere. You can expect the application to become even more interesting as we continue to tweak it.
Tom
4 wwwhitney // Mar 16, 2009 at 4:25 pm
Great stuff Tom. I am glad to see you embracing the trend of not requiring referrals in order to use advanced features.
I would love to see an entry in the future on what sort of insight you are able to glean from this application and how it might differ from data collected from other sources (especially online). While your app is not a game per se, it offers a different incentive for participating (entertainment) than other online research techniques (compensation).
On that note, have you considered any other ways of making the app more engaging? For example, maybe include badges, avatars, or humorous quiz results. For example, you could have avatars which users put on their Facebook profile or blog whose appearance and description is generated by the users’ profile on Compared to Me.
Users would then have more incentive to answer more questions so that their avatar better reflects them and it gives them a more fun visual result than a graph. I think this would particularly resonate with Generation Y and younger users who have grown up taking these sorts of quizzes all the time on Myspace and now Facebook. Of course, at the same time encouraging people to post their results on their pages would increase exposure to the app for potential new users.
And now I’m even getting farther ahead of myself, but you would then have an additional way to filter data between those who post their avatars elsewhere (i.e. those that are interested in this aspect of the app) and those who don’t.
Well that was a bit of stream of conscious brain dumping there, I hope you don’t mind!
Anyways, love the blog, keep up the good work.
5 Tom H C Anderson // Mar 17, 2009 at 11:00 am
Good ideas, thank you. We do have a badge available. Avatar type badge is a great idea.
6 Model Questions ROI of Paid Tweets // Apr 15, 2010 at 5:51 pm
[...] by guy Kawasaki several times. We like to call it the Kawasaki Bump internally. I’ve even blogged about the bump to get Kawasaki to tweet it again to get another bump, a bump on a bump. It [...]
Leave a Comment