stumble upon

I was just thinking about stumble upon, when I stumble a new site it gets a lot of traffic. Here's something that might be worth a test (or maybe someone already knows...)
Would the total amount of traffic be different if:
- 5 people stumble a given site at the same day, or
- 5 people stumble the site in 5 weeks (one stuble each week)

I think its more about the "value of audience" in your account, my value is above 400 that will say people surfing my account/suggested websites frequently. If people with a high "value of audience" surf your suggested site you got more visits then from stumblers with a low value...

I think I will agree with Olaf, but I am not really sure how this works (actually none is :) )

Also the traffic depends on the tags that you list a site. General sites - eg. humour, sports - get much more traffic than niche sites but in anyway stuble is a great tool to get some free traffic.

yes right use a general category from the tag cloud while submitting a new stumble

I love stumbleupon as a tool for a quiet moment but we receive very little traffic from it, mind you we don't go out of our way to promote ourselves on it either, the problem with stumbleupon users are they are normally just browsing to pass the time and are hard to convert to a sale, hence we tend to only suggest Blogs or sites that are article based.
Like nearly all user driven sites most stumbleupon users ignore adverts as well.
Does the category make much of a difference, common sense says a more generic category will get you more users though you will be part of the noise. While a smaller group less matches but less noise.

tim, from the paymentblogger article you wrote in april07:


so i guess things have changed since this last post? does stumbleupon bring plenty of traffic to your sites already?

I must say new stumbles having less SU traffic than before but I have still a lot of visits from older stumbles I guess they changed something for the new stumbles

My theory of stumble upon is this,

a user stumbles a site, stumbleupon shows it to 10 of his friends,

1) if all of them like it then stumbleupon shows it to 20 of their friends.
2) If 5 people like it then stumbleupon shows it to only 5 of their friends 

So every step they evaluate if its worth to show it to more people.

The number are imaginary but I think you get the picture of my "great theory on stumbleupon"

makes sense, maybe you need also stumble by yourself (I stumbled that last time several weeks before).

I will stumble some sites the next days (via the SU bar and suggest than some new sites)

In my opinion their algorithm is very complex. I think they use a value for the site that you stumble (sites with many good reviews get more traffic for each new page they submit), they see if a user submits many pages from the same domain (so if you are going to stumble your whole site it wont work) and they somehow know how those votes are spreaded.

The good thing is that they don't give any penalty to groups so if the same 10 people continue to stumble each other's sites it wont be a problem. For instance if you do this on Digg your stories are automatically get burried.

so first you should promote you main site a lot to get a lot of stumbles for the sub pages

I'm afraid I think your almost completely wrong ;)
I beilive it works more like this...
The original stumbler tags the article and something happens, depending on their audience score depends how many waves occurs.
the first wave is always unrelated persons, who have ticked the first tag (as in their likes) this wave is normally between 2-10 people
The second wave is friends and people who like the first and second tag between 5-50

Now if the original stumbler had a high audience the waves will continue with friends and all tags.
When someone thumbs it up, they cause a wave (again based on audience) of between 2-100

However Stumbleupon has a domain overload feature, a domain that is repeatedly stumbled by the same people will see a rapid decrease in stumbles (I still get a couple of hundred every day) but new stumbles on ventureskills.wordpress.com for example are a mear trickle so.

1st stumble wave is = Audience(of stumbler) - domain saturation = #visitors of primary tag
2nd wave = Visitors of primary tag and friends
3rd = All tags and friends
and on

Friends stumbles do seem less effective also people stumbling a discovered site without stumbling to it seem less effective.

OK its time to stumble :)

I know their algorithm is very complex, that is why I tried to explain it in simple terms.

I am glad that everyone has different theories about it (it could not be otherwise, we all have differrent theories about simple things in life)

I think the truth lies somewhere between All these "Great theories" maybe its time to try to UNITE them, maybe one day the Nature magazine will have a story on US trying to get the GREAT UNIFIED THEORY of the STUMBLING EFFECT ;) and then the debate will start with other members of other forums that have yet their own theories about stumbleupon and they will send missionaries over on our little forum (big depending on how you look at it) and they will try to convert us to their beliefs and they will be long debates that might actually have casualties  ..... ;)

lol :D The big theory of stumbleupon,

Well its easyish to test our theories

Setup a new stumble account with 5 friends (which we can track) and then use it to stumble a new, and prior stumbled domain and see what happens

or start stumbling each other out...

stumbleupon has a saturation point where if the same stumblers thumb up a site without a proportion of new blood the site receives less and less new stumbles and over a great time.

removed...