"I was thinking about you all day today and what a great
person you are."
"I wanted to be #1...After 2 months I reached the top position for my most popular keywords."
Above The Fold!
Doing SEO in a Vacuum
25 February 2005
Google tweaked their algorithm and many sites stopped ranking for their own site name. I do work on a limited number of sites but have many friends who do a good amount of SEO work and have seen many sites disappear from the search results.
Change as a Part of SEO
Some sites take risks knowing that they will eventually get hit. Other sites which lose rankings are just the baby thrown out with the bathwater.
You take what risks you can afford and avoid those which you can't.
If your business model is reliant on any single network or partner you are setting your business up for failure. Even the search providers know this.
Ranking for "SEO"
Single word queries, especially ones in competitive fields, can be hard to rank for. The phrase "SEO" itself really is not all that targeted, but many SEOs want to rank for that term as a mantle piece. It is like having a trophy.
Most the traffic for "SEO" is automated bots. Even when my site was top 10 in Yahoo!, Google, & MSN I was still getting far more traffic from "SEO Book" than from "SEO."
Even if there is little direct value, ranking in a top "SEO" listing it is still a fun challenge.
No Longer Ranking for "SEO"
A couple of the sites that were hit in the Google update were ranking well in most major search engines for the term "SEO."
Recently SEO Guy ranked #1 in Google. My SEO Book blog also ranked in the top 10 - 20 sites since June of last year.
I just checked Yahoo! and MSN and SEO Guy was #1 for "SEO" on both sites. My site was #4 and #6. Teoma is a topical clustering based search engine and even their algorithm still ranks my site at #11.
After the recent Google update both SEO Guy and SEO Book no longer rank in Google for their official site names.
Unnatural Linkage Data
Competitive search term relevancy in Google is mostly due to link reputation.
On his forums SEO Guy has a signature message asking members of his forums to link to him with SEO as the link text.
Doing this of course helps control the anchor text, but you can overdo it. If too much anchor text is similar then a site may be filtered out of the search results.
My site also could have done a much better job of mixing anchor text as well. Many of the links pointing at my site used "SEO Book" as the anchor text. Because of the keyword rich URL many unrequested links had the phrase "SEO Book" in them.
Too many sitewide inbound links could also look unnatural.
My blog is on a good number of blogrolls without me knowing about it. I also get links from the owners of many forums that I moderate. I also placed about 4 or 5 sitewide ads on related sites.
SEO Guy also has a good number of sitewide links from his forum members and from getting links in exchange for doing mod rewrites.
What Google Did
- Google could have lowered the % of similar anchor text to trip that filter.
- They could have looked closer at how many sitewide links you have.
- They could have looked at the deep link ratio.
- They also may have rolled in latent semantic indexing or another similar technology.
- Any combination of the listed items.
SEO fundamentally has not changed. Many sites that were ranking still rank fine. Mixing anchor text is something many people have been doing for a long time.
Why Our Sites Were Hit
Most likely because we had rather unnatural linkage profiles. The number proposed is just an idea still under development, but natural link development and people have some normal or predictable patterns. If almost all of your inbound links have the same words often that is not a natural pattern.
Either automated or manually Google decided to filter out the home pages. I think that this new algorithm shift is automated and is a bit on the aggressive side (since it filtered out thousands and thousands of websites).
Google created an e-mail at email@example.com for feedback associated with this update. They do not normally do that, which signalled to me that they know they are in error or aggressive with their new algorithm.
Does My Site Deserve to Rank Well for "SEO"?
In many search algorithms it does, but relevancy is in the eye of the beholder.
For most optimists the genericness of the term means it does not have much value other than as a mantle piece.
Does My Site Deserve to Rank for "SEO Book"?
I think so. But again, relevancy is in the eyes of the beholder. In many search algorithms it does, but if Google decided they do not want to rank my site for its name that's fine by me.
It is their job to control the relevancy of their search results.
When I Started on the Web...
About 2 years ago I created my first website knowing nothing about HTML, marketing, search engines, or the web.
I focused a bit much on learning SEO because I knew nothing else. I knew I had to focus on that topic so I could do it well enough to establish a brand.
In the long run I feel I need to learn significantly more about the social aspects of the web if I am going to keep doing well.
Probably By Accident
I just happened to stumble on the hot idea of SEO. I think that is how many people get into it.
Because of the desire to learn I probably have read about a million forum posts and have wrote many thousands of them.
Not knowingly at first I helped establish myself in some social circles just by helping a few people and reading many forum posts. I know people that know way more than I will ever know and those people are a large part of any success I may have.
After you read enough posts eventually you learn who you can trust and you learn the answers to many questions. Some of the posts I have made over the past few years have likely acted as mini sales pages which help sell my ebook and build link popularity.
Economics of SEO
One of the hardest parts of selling SEO services is helping clients see the work which goes into ranking a website. Often people will spend $10,000 a month on AdWords and bawk at the idea of paying even a few thousand dollars for SEO services.
If a business does trust you to do their SEO then algorithm changes like these may sour those relationships.
One person I spoke with had a client who had just bought $200,000 of supplies and just had their site disappear. After they just unloaded $200,000 worth of inventory how do you tell a client that their distribution channel just dried up?
Search engines would love to have hard to game and unpredictable commercial search results. Sometimes in doing this they sacrifice relevancy.
People create nearly 100% automated systems which game the holes in the search relevancy algorithms. It's no big deal if their sites get penalizes as they can build hundreds or thousands of them quickly.
As time passes and this gets more competitive some of these people will move from the most profitable markets into many niche markets.
Economics of Search
Search engine operators know there is often more money made in manipulating their results than there are in their ad networks. They also know that it costs next to nothing to get into SEO and there are a far greater number of people doing it than they can employ.
It is not an accident that their "Google Advertising Professional" program has AdWords plastered across the top.
(Incidentally, for whatever it is worth I am qualified but thusfar see the program as nothing more but an excuse to market AdWords with a shiney logo on various marketing sites.)
It is not an accident that Google's official SEO information page mentions fraud and the FTC in the SEO tips.
In the same way they paint their broken algorithims as a fault of "spammers" they also want to frame SEO services in general as sketchy.
It works best for Google if people have to worry about how to move $200,000 of inventory and the only solution is their AdWords ad network.
With its curren P/E ratio Google needs to keep growing revenues.
The harder their results are to game the more money they make, so long as they do not sacrifice relevancy.
If relevancy goes then they lose users.
Since my site stopped listing on Google under "SEO Book" I have noticed a few things:
- My ebook sales has been consistent with their historical rates (except for a recent increase in sales due to the fine folks at Linking Matters publishing one of my articles).
- My traffic is roughly constant.
- More clicks on my AdWords ads for "seo book."
- Many more referals from seobook.com searches (since it shows my URL).
To me that means that my site would do well even if Google did not exist, which is especially awesome as an SEO.
WWGD? What Would Google Do
Google tells people to create content that is good for their users and they can't say that they are doing the same when they filter out tens of thousands of domains for their official site names.
Why This Filtering Idea Would be Appealing to Google
Yahoo! and MSN are easy to game right now. Usually you can just point keyword rich links from various sites at your site and prettymuch own their search results.
Submitting to directories, link exchange networks, link farms, or guestbook spamming are all effective, I have recently seen all of these take sites to the top of the search results in Yahoo! & MSN.
By aggressively filtering out sites which are aggressively manipulating linkage data Google can make SEO much harder for the masses.
Since its competitors engines are easy to dominate with low-grade link spam Google would hope that many would just forgo optimizing for Google and go after the other engines.
To be fair about that last sentence Yahoo! has paid inclusion and thus benefit to having search results that are easy to manipulate and MSN Search just came out of beta.
Some other people in the search space do not necessarily view SEOs in such an adversarial light. In fact, Apostolos Gerasoulis of Ask Jeeves / Teoma sounds as though he is keen on SEOs.
Things the Average Webmaster Should do:
Some of the filtered sites have already reappeared. Just because you are filtered out does not mean it will be perminant. There are a few things I would recommend doing to "naturalize" your linkage profile.
- Deep link: Work a decent number of deep links into your linkage profile.
- Don't buy too many sitewide links.
- When you buy sitewide links: try to get the people who link to you from their site to link to a presell page on their site from every page of their site and then link from that page into various points in your site.
- Mix your anchor text: If you own a keyword rich domain that will naturally pick up a ton of linkage data ensure that when you build links you use various semantically related terms whenever possible. (Using Google Sets or the Google: ~ search function would be an easy way to find a few semantically similar search terms).
Things I Could Have Done / May Do:
- Host presell pages on Ad Buys: I was renting about a half dozen sitewide links. I also get a bunch of free links from various SEO forums. I could have asked some of these people to host a presell page on their site and had them link into my site from it.
- Mix Anchor Text Better: When I first started my
other site this one was not popular. I had no idea that my other site
was going to become as popular as it did.
When I first started building links I did not mix anchor text that much because I wanted to build a strong brand. There is something contrary to building a strong brand and telling people to link to your site with fried gibronie or other random words the anchor text.
I have been mixing anchor text for a while, but it was not enough to overcome the sheer number of sitewide and keyword rich links pointing at my site.
- Host Articles Like this One on That Site: My blog does not get a ton of deep links, but is on a good number of blogrolls. The articles I write on this site build a good number of unrequested links. My information segregation hurts my deep link ratio.
- Make more long posts on my blog: Most of my blog
posts just link to a bunch of sites with little commentary in them.
In doing that and combining many topics in many of the posts it makes
the posts less likely to build linkage data.
Most of my would be longer essay type blog postes wound up as articles on this site.
- Not cross-link (or redirect the links): I was linking to SEO Book.com from a few of my sites via sitewide links. Primarily this was for advertising (as there is rarely any other ad space which has better targeted traffic than a person who just read a 3-page article you just wrote.
- Use a better affiliate program: I was using Clickbank and another tracking script, but it was using many 302 redirects. Instead I probably should have used one hosted on my site to have better control of my linkage profile.
- Create free SEO Tools: I could have created free SEO tools, added forums to my site, or create ideas designed to be viral in nature and point link popularity into other nodes of the site.
Reasons I did Not do Anything:
My site has a bunch of naturally occuring links, and seeing how some other sites do more sitewide advertising than mine I did not think my site was prone to ranking fluctuations for its official name.
My site has many natural links including authoritative or hub sites like Search Engine Watch, DMOZ, the Virtual Library (the directory which started when the web started), Yahoo! Directory, & JupiterMedia.
I also have been less risky than I was in the past. I have been focusing more on brand development than SEO, only buying a moderate number of links. I think before this update my ranking had slid to #15 for "SEO".
If I would have mixed the anchor text in every link I had some control in I still may have got flagged by this update just by being on many blogrolls and other sitewide links using my official site name.
A Couple of My Favorite Recent Quotes:
Some people argue that SEO in and of itself is somehow unethical.
If search engines can't find a way to save themselves without having to ask me, or Danny, or anyone else to adopt some sort of moral stance on their behalf and become lord protector, because of their flawed business models, then there's really not a lot to save is there?
- Mike Grehan
With the Florida update and with this recent update, this is twice in just over a year that Google has shown that they are willing to risk their own relevancy to be more innovative or to try to make SEO harder.
You never really understand this business until you get at least one site banned.
- Anonymous friend
To be fair about my own site ranking poorly, most sites are not blogs about SEO with a keyword rich domain name and over 100,000 links (with probaby 80-90% similar anchor text), but at the same time this recent change has had large affect on many businesses.
Many of them did absouletly nothing wrong other than have a strong brand not named Google.
Has talked only about SEO thusfar. SEO helped me do well, but beyond that brand development is what will help me go further.
In search engine optimization we often miss the end goal of the optimizaton.
When Google does not rank my site my sales do not significantly change, so it shows there are other parts of my "optimization" which are far more important than my Google ranking - and that is coming from a person who sells a product which is SEO related.
I am not saying that people should ignore Google, just that it is not worth it to do SEO in a vacuum.
by Aaron Wall of Search Marketing Info