Diminishing Effect Of Amount Of Links on PBN

by nik0 Banned

Posted: 11 years ago 3 replies

Lately I started to dedicate domains per client but it seems an inefficient way of spending money.

Quick example:

- Money site A receives links from 5 domains, links to 1 client.

- Money site B receives links from 10 domains, which links to 2 clients in total.

- Money site C receives links from 20 domains, which links to 4 clients in total.

Which site ranks best?

C --> B --> A

While the exact same amount of link juice is being passed.

For the ease of link juice calculation, let's say each domain has 10 juice points

5 domains * 10 = 50 points / 1 obl = 50
10 domains * 10 = 100 points / 2 obl = 50
20 domains * 10 = 200 points / 4 obl = 50

See same,

Why do you think money site C ranks much better? Is it because:

- Amount of referring domains carries so much weight?

- Amount of different anchors play a crucial role?

- There's something in Googles algorithm that doesn't look as straight forward at OBL's as I/we do from a theoretical view?

This is not based on guesses but on a real life example of turning a 160 sites large network (5 obl) and thus 800 link spots for 80 clients, so 10 domain links per client, into a dedicated network with max 1 obl and thus links from only 2 domains per client.

Everyone's rankings are worst then before, and significant enough, for example sites with multipe top 3 rankings now rank at #5-10.

It feels like:

- domain with 1 OBL passes 70% of juice
- domain with 2 OBL passes 80% of juice (40% per link)
- domain with 3 OBL passes 90% of juice (30% per link)
- domain with 4 OBL passes 100% of juice (25% per link)

In terms of dollars that means we would lose $30 out of every $100 spend on domains.

And to use my previous table again

5 domains * 7 = 35 points / 1 obl = 35
10 domains * 8 = 80 points / 2 obl = 40
20 domains * 10 = 200 points / 4 obl = 50

What do you think? Some say add links to authority sites but that's nonsense of course as then we only get 40% of juice vs 70% when not linking anywhere else (assuming those made up numbers are correct).

#amount #diminishing #effect #links #pbn

kindsvater

11 years ago

You are assuming juice points are strictly additive. I do not recall anything in Google's patents suggesting that simplistic approach has ever been appropriate. Way back to the beginning algorithms for evaluating link juice, or PageRank, have been mind-numbing when looking at mathematical equations.

Admittedly, I don't have any secret access to Google's algorithm, so call it educated speculation, but what you may be seeing is 1 vote = 1 point of juice, 2 votes = 2.2 points of juice.

In your example Domain C is passing more link juice because it has more incoming juice due to having more domains link to it.

--------

Alternatively, you could be seeing a spam filter where websites that only link to 1 website have a greater tendency to be gaming SEO, versus a site with multiple links out and arguably trying to assist visitors in finding useful information. A one link site could be receiving a dampener on its link juice because it is trying to channel link juice.

When you say adding authority site external links is "nonsense" that is from the webmaster's perspective in trying to maximize juice flowing to one site. From Google's perspective you get a different result..

.

[ 1 ] Thanks
1 reply

{{ DiscussionBoard.errors[9807594].message }}

nik0

Banned 11 years ago

Originally Posted by kindsvater

I'm just using the logical approach of a water can, if you put 2 tubes in it both glasses of water under it remain the same amount of water, and if you put 1 tube then all the water flows in the single bowl but yeah after large scale testing, instead of guessing, it doesn't work that way.

I've thought about the spam part as well indeed, eg domain linking to only one domain.

Dividing the juice by two by linking to an authority site seems nonsense to me, as I do believe there is a dampening effect but I don't believe that will be as large as 50% more likely 2,2 as you used in your example or 0,7 - 0,8 like I used in my example in other words 10-20% so if that rule of 10-20% would apply it would obvious be nonsense to give away 50% to an authority site, or 40-60% for that matter.

Now if I took the 4 OBL approach it would mind a lot less to place a 5th link to an authority site and perhaps that might be on the border of beneficial.

Either way, I think it's time to start combining some clients and my own money sites on domains again as it seems I'm losing a lot of value doing it in the current way (linking to max 1 client from each domain).

Thanks

{{ DiscussionBoard.errors[9807608].message }}

jinx1221

11 years ago

Originally Posted by nik0

Why do you think money site C ranks much better?

What is your definition of "much better?"

Originally Posted by nik0

Is it because:

- Amount of referring domains carries so much weight?

- Amount of different anchors play a crucial role?

- There's something in Googles algorithm that doesn't look as straight forward at OBL's as I/we do from a theoretical view?

Probably a mix of all three, and then some. All things being equal, a few of the absolute different factors are anchor text variation, amount of referring domains (weight being equal), and those mysterious algorithm matrix thingys.

Originally Posted by nik0

What do you think? Some say add links to authority sites but that's nonsense of course as then we only get 40% of juice vs 70% when not linking anywhere else.

That's not nonsense.. that's one of the mysterious algorithm matrix thingys.. link juice can flow backwards as well as forwards, in certain circumstances. It's called 'co-citation'. Also, there is no spoon.