The Zen Of StumbleUpon – Tim Nash StumbleUpon Model

I was introduced to Tim Nash StumbleUpon (SU) Model by one of my friend at SU.  So I decided to do my own testing to see how far this model is reliable.

A little background of streamxy.com and secondworld – the test case subject

streamxy.com is a small blog with a decent traffic between 150 – 200 per day.  As a stumbler,  secondworld only have  about about 215 friend with no of pages stumble between 0 to more than 200,000 pages (yes, more than 200,000 page).  At the time of writing, there are only about 78 subcribers.

A little background of Tim Nash StumbleUpon Model

For detail on Tim Nash StumbleUpon Model and their subsquent  question and answer, you can point your browser to the respective article and post.  For simplicity and illustration purpose,  I would like to list all the model introduce by Tim Nash is his post.

Basic Model

Initial stumbler + (number of thumbs up / number of thumbs down) = visitors

The basic model dictate that the no of visitors will be determine by the initial stumbler score (in my own opinion dictate by the number of reviews and favourites, not just a thumb-up) .

Audience Driven Model

Initial stumbler audience + (% of audience of stumbler per thumb up / number of thumbs down) = visitor

In my opinion, the initial stumbler audience could be determine by the number of their subscribers and contacts list.  In this assumption, a stumbler with a  low subscribers and contacts list will not garner a lot of visitors and therefore affecting the next variable (% of audience of stumbler per thumb up / number of thumbs down).   If someone plan to use SU as a marketing tool, the choice of friends and subscribers will dictate their marketing success using SU.

Audience + Domain Driven Model

(Initial stumbler audience/#stumbled domain)+ ((% of audience of stumbler per thumb up/#stumbled domain) / number of thumbs down) = visitor

This model extend the previous model and add one variable, domain name.   I think this is where the rules, never stumble every pages of your own domain only, came to exist.  This could also explain “ghosting” effect on SU where you can’t seen your submitted web page from your own domain.

SU Intergrity Score

Based on the first three model, I think we should pause for a while.  At this junction, I think we should introduce a new concept for SU, The StumbleUpon Integrity Score.  This score will be determine by the following:

1. No of connections. This should translate into number of friends and subscribers, the higher the better. In order to simplified this variable, let call it audience.
2. No of favourites and reviews.  This will determine whether this profile is an active stumbler or not.  The higher the better.  As in the first one, let call this variable as votes.

Based on this variables, we can presume 4 scenario and possible score:

1. Low audience but high votes. Active stumbler but most probably treat SU as link exchange where stumble will only limited to their existence audience. It is expected this kind of stumbler will only stumble your pages if your stumble their pages (either their own domain or others). In my opinion not much traffic will be generated by this profiles since their stumble will only be seen or read by limited audience.
2. High audience but low votes. This group of stumbler treat SU as social networking more than social bookmarking recommendation services (or sites).  Not much traffic will be generated by this profiles since they might not stumble your page. Hence that explain low votes.
3. High audience and high votes. This is what we might call as the gems of SU and also known as “The Power Stumbler”.
4. Low audience and low votes.  I got a few friends under this category and they normally have less friend and less stumble. Their last login could be weeks sometimes months.

I will leave the discussion on SU Integrity Score at other post since this topic requires special attention.  One particular attention we need to consider what happen if a subscriber also a friend or a subscriber just a subscriber but not a friend, which one is better.

And The Ultimate Model

(Initial stumbler audience /# domain) + ((% stumbler audience /# domain)+ organic bonus – nonfriend) – ((% stumbler audience + organic bonus) + N) = visitor*.

*In the actual model, Tim Nash had not put this equal to the possible no of visitor.

In this ultimate model one additional variable added, organic stumble which is consider as an added bonus.  However despite the complexity of the ultimate model, it can be breakdown into three part:

1. (Initial stumbler audience /# domain). This is our SU Integrity Score.
2. ((% stumbler audience /# domain)+ organic bonus – nonfriend). The amount of  thumb-up or down received adjusted only to the stumbler friends and their subscribers who happen to be also their friend (Refer back to our SU Integrity Score)
3. ((% stumbler audience + organic bonus) + N). I think this is what we might call “Viral Effect Of SU” this is the traffic brought by audience own by the initial stumbler audience.  Hence N, donate the random effect.

The Ultimate Model against the effective of SU Mailing List and Exchange Program

Based on this model, the initial effect of this program will be an increase in traffic but eventually the effect will be reduce.  If someone plan to open-up this program, they should do the following:

1. Treat this program like a social networking but close network, strictly by invitation only.  This will weed-out non-active stumbler because your goal should be to create a close network of power stumbler only.
2. Always have a new influx of power stumbler to your mailing list and exchange program
3. Limit the number of pages sent between participants.  This will slow down the dilution of traffic effect against the possible slowdown of new influx of power stumblers.

So what next

So what do you think?  Have we got it right or I make you more confuse? In the next couple of days, we will look into more about this model.  So stay tuned and bring in your question and suggestion for our next discussion.

Popularity: 14% [?]

Related posts

• StumbleUpon: A Thumb-Up VS A Review (13)
• What To Do For Your Blog In 15 Minutes? (0)
• What Is Your Blog (Or Domain Name) Worth? (2)
• The Best Time Of The Day To Post To Your Blog. (0)
• The #1 Feature Of StumbleUpon That Mostly Overlooked (1)