I made some thoughts about that topic the last days and want to share them with the community. They are depending on building Linkpyramids. One example: Let´s say you have two profile backlinks that have a chance of 50% each to get indexed. Also that if google find the first one it follows the Link and find the next one. Also there is no lost of linkjuice on the way to Link A to Link B: So assuming that both have Linkvalue of "1" you get on average 1 Linkvalue (50%*1+50%*1) If you link with Link A on Link B and Link B to your moneysite: 25% Chance that no Links get found (0 Linkvalue for us) 50% Chance that both Links get found (2 Linkvalue for us) 25% Chance that just Link B gets found (1 Linkvalue for us) 25%*0+50%*2+25%*1= 1,25 Linkvalue on Average So we win 0,25 Linkvalue if we point them in a row instead directly! If we have a Chance of 75% to get indexed we win 0,1875 Linkvalue. If we have a Chance of 25% to get indexed we win 0,1875, too (but that are way more percentages of our Linkjuice). => So Building various Tiers is more important if the Chance is smaller that the links are getting indexed! The other Reasons. If we have 2 Tiers of Backlinks, Tier2 Backlinks can link to more than one Tier1 Backlink and increasing the Linkjuice. Nobody can tell me that the linkjuice is just half each if we link to two sites. Otherwise there wouldn´t be ********ing blogs existing. => Building various Tiers is more important if Tier 2 can link to more than one Tier1 Backlink So the question is just, how much Linkjuice get lost each Tier. Its impossible that the Linkjuice flows one to one because after a few days the whole internet would be a big swimmingpool of Linkjuice. Actionplan for the Praxis: Web 2.0 properties have a higher chance to get indexed. So they should be the first Tier. Backlinks that can Link to more than one Tier 1 Backlinks are better for various Tiers How many tiers you should use depends of how likely your links get indexed And now the Theory based on that: Building more than one Tier Web 2.0 properties with just one backlink is stupid, because the Chance that all are getting indexed is already very high. So there is more Linkjuice got lost on his way than it brings more. Scrapeboxcomments with multiple domains in the commenttext on autoaproval blogs could be very good as last tier! What do you think about the following pyramid? Moneysite <- Web 2.0 <- Profile links (with just one link on the page) <- scrapeboxcomments with various domains. Thoughts?