|Avoiding penalties against "thin affiliate" pages|
is this the way to go?
| 5:04 pm on Oct 23, 2006 (gmt 0)|
I've saw what Google can do to thin affiliate sites when I lost all rankings for about 300 pages of book sales pages that I ran in conjunction with a content/directory site on a sports subject.
I'd like to try again, and avoid these troubles, if its possible. Here's my question: If I build good top-level pages (with my own reviews, comments about products, etc.) and get them indexed, can I avoid troubles by excluding G's bot from my product pages with their affiliate link?
I'd use a robots.txt or NOINDEX meta, or both to achieve this. I suppose I'm still taking the risk of a human review.
Thanks in advance to an excellent group for thier help.
| 11:25 pm on Nov 3, 2006 (gmt 0)|
Plenty of sites that have affiliate links rank consistently well in Google, and rightly so.
If you are, indeed, planning on adding value and making your site a useful resource (e.g., giving users a reason to *bookmark* and regularly visit your site), then there's no reason to hide in shame ;).
If you're worried about how your site would look in a "human review," however, that doesn't bode well.
| 3:40 am on Oct 19, 2006 (gmt 0)|
<The next 7 messages were spliced on to this thread from another location>
There's a lot of terminology flying around about this phrase "thin affiliates". What does this mean in Google's interpretation and how are they monitoring it with filters?
I've also read somewhere it's time to get fat.
What does one need to get fat [ except for sitting at the computer too long ].
[edited by: tedster at 5:07 am (utc) on Nov. 4, 2006]
| 4:51 am on Oct 19, 2006 (gmt 0)|
To my memory, this term began flying around with intensity around June 2005, at the time that the training guidelines document Google was giving to human evaluators was leaked.
-- CONJECTURE --
Certain elements of being "thin" can be evaluated by an algo -- things like percentage of text not appearing anywhere else; content that is obviously an affiliate program feed; content that is republished content of other kinds, not from the affiliate program, but still not unique.
But the final judgement about whether a site adds real value to their "dupe" content, and enough value that Google would want to send visitors to this specific url through their SERPs, seems to me to require a human evaluation. If human evaluation sets a "thin affiliate flag" for a domain, then going forward the algo might measure the degree of change Google sees and at some point the algo might trigger a re-evalution.
In the physical world, any outlet for identical products or services needs to add some value or other just to survive -- better display, solid product information from their sales people, whatever. There's every reason to do the same kind of thing on a website.
The "make good money with an instant website" approach is dying. But if you fatten up a previously thin affiliate site with enough real value, that content can even become natural link bait and accomplish a second purpose.
| 8:27 pm on Oct 19, 2006 (gmt 0)|
> with enough real value
And how to assess what is enough real value? Some sites that are performing great as of recent times are nothing but links to dozens if not hundreds of sponsors giving the impression of being authorities. While a visitor may find this one-stop shop convenient the underlying question is where is the unique content?
| 11:32 pm on Nov 3, 2006 (gmt 0)|
Be careful about assumed causation. "Such and such site does [x] and today they rank well!" You cannot know if they rank well because of that particular tactic, nor do you know how long they will continue to rank well.
Some general truisms when it comes to Google:
- If a site provides compelling information, entertainment, or tools and is then sprinkled (reasonably) with affiliate links, then it's likely to do well over time.
- If a site *starts* with affiliate links as the foundation, the chances for success are more slim *unless* it provides compelling info, entertainment, tools, etc.
In other words, if you're building an Amazon.com store, are there any strong reasons why someone would visit -- or even bookmark -- your site instead of just going straight to Amazon.com? If not, why would our users want us to prominently list your site?
Getting "fat" means adding value... and, more specifically, a value that is distinct or greater than the value derived from the actual ecommerce site itself.
| 1:22 am on Nov 4, 2006 (gmt 0)|
Interesting - assuming the same principles as those applied to affiliate sites - I've got some examples, which might create some thought :
- I have become aware of a site which mines snippets of data, combines and matches them into a theme or subject, presenting the data in a currently unique and [ IMO ] compelling manner, which is not done elsewhere. Sources of this data are typically places like Wikipedia, CIA factbook. It's heavily linked to by authorities [ .edu's ] as a great source of organised info and secures income through Adsense.
This is the core of the site. However, there is a thin spattering of comments - so the content is predominantly re arranged, non original content.
[ The site used to rank well, but currently has canonical problems which i spoke to the owner about ]
My score [ IMO ] on this site which acts as a tool is probably 9/10 .... i worry about how the Google algo will score it. [ hmmm :) ]
- Then there's an e-commerce affiliate site which takes data feeds on products and arranges them into sets of themes, let's say red ones , white ones etc. [ Nothing new here. ] It then arranges this product data, sometimes introducing snippets, and compares the data to other products by price, language and key features [ such as comparisons to other services in the locality ] to provide users with an output which makes comparisons and choices easier.
This site was built from primitive beginings and is still evolving.
|- If a site *starts* with affiliate links as the foundation, the chances for success are more slim *unless* it provides compelling info, entertainment, tools, etc. |
I wonder, will this site have problems because it did not *start* in a compelling way? Or am I to believe that *unless* means the site is capable of redemption in the eyes of the Google algo, provided the work is put in.
I give the site [ IMO ] a 7/10 in the regions where it operates since it is unique and offers a compelling service that others currently do not. However, it will have to work hard to be better than the rest over time.
How will Google determin it's value?
IMO , affiliates generally have trouble competing on Adsense because their margins are often 50-60% below their partner providors or direct sellers, negating them from sensible bids.
But, if they are providing quality results with "features" and "value addedness" in the organic search which their partners or direct sales competitors do not, or cannot, it makes for a dynamic and competitive environment on the overall search page, in which an innovative affiliate can play a great part in. [ Just my thoughts. ]
Writing useless blurb could be just as useless as a site packed with no functional benefits and tools.
So i wonder conceptually how Google counts the votes and activity of appreciation that leads to it's assessment and placement in the overall value chain . Very complex indeed [ IMO ].
[edited by: Whitey at 1:33 am (utc) on Nov. 4, 2006]
| 2:26 am on Nov 4, 2006 (gmt 0)|
Thank you, Adam.
| 2:46 am on Nov 16, 2006 (gmt 0)|
My main site is just book reviews, and I have links to Amazon.com and Amazon.co.uk on pretty much every one; that's never hurt my Google rankings that I can tell. I do NOFOLLOW one of the two Amazon links, however, since the Amazon.com and Amazon.co.uk pages are often pretty similar. (That's probably being unnecessarily paranoid.)
But my reviews are original, usually substantial, and in many cases (with more obscure titles) the only commentary on the book available online.
| 7:33 am on Nov 18, 2006 (gmt 0)|
You posed some rather complicated examples.
What I learned about THIN sites comes from the very basic definitions in the Google training document. When you think about it, teaching a human reviewer to make decisions about reviewing affiliate sites as THIN or NOT THIN could lead to unmanageable complexity --everyone has a different idea of what's a *worthy* site. So the instructions kept it very simple, and that makes sense to me.
An example of a THIN affiliate site was given as your typical Amazon store page: no original content, all (or many) links pointing to one affiliate.
One example of a NOT THIN site was the same sort of heavy-affiliate page, but with comparison shopping between two or three suppliers. The comparison feature gave the site enough *added value* to escape the penalty.
I was surprised by how low the bar was set for NOT THIN, since simple price comparison is hardly what you'd call original content.
That being said, its been a long time since those definitions above were leaked. Things could have become much more complex since then. I got burned through my own fault, and learned a lesson from it.