Link Spamming Poses a Threat to Your Link Popularity

It's time for a long overdue newsletter. I want to tell you about the biggest risk to your PR and also mention the US court case where I have been selected as an expert witness for the defense. Ah, it's a good thing I grew up watching Canadian boy Raymond Burr as Perry Mason.

First, since we have a number of new partners in the links program I should make sure we're all on the same page jargon-wise and conceptually, and I want to address a few things I get asked about fairly often. Then we will get to the meat of this newsletter.

First of all, I think most of you know that link popularity at Google is measured by PR or Page Rank; PR is a statistic named after Mr. Page, a founder of Google. PR is an 11-point scale where PRzero indicates a web page with little link popularity and PR10 indicates a page with huge link popularity (i.e. google.com). By the way, it is a myth that PRzero is worthless (source: webmasterworld.com). The public accesses information about website PR via the Toolbar which can be downloaded to one's website or accessed via various free tools on the Internet. All the world's 9 billion or so web pages have a PR level so that means that all the world's web pages are crowded on that 11-point scale. There are millions of web pages at the PR3 level, for example.

Toolbar PR is imported (or updated) every few months. For example, there were no imports between May 1 and November 1 of 2007. But following the November 1 2007 update there was a mini-update around January 1, 2008. There is no reason to check you PR weekly or whatever. There are rarely more than three updates in one year.

Toolbar PR is not the same as Real PR. Real PR is updated continuously. The Toolbar is an unrefined tool updated only occasionally and always minimally three weeks (often 4 months) behind real PR. Toolbar PR is a simple integer. Real PR is calculated out to many decimal places. There is a huge difference between PR3.031 and PR3.949 (millions of web pages). Folks have a fit when they drop from PR3.000 to PR2.999 (seeing it as PR3 to PR2) but they don’t notice when they’ve gone from PR3.000 to PR3.945 because that information, which is beyond the simple integer, is never published via the Toolbar. Note that the loss in PR from 3.031 to 2.999 would have a negligible negative effect on rankings while the gain from PR3.000 to PR3.949 would have a substantial positive effect and yet not be noticed via the toolbar.

Link popularity gives a website buoyancy. Better PR increases the likelihood that a website will rank not only for its primary search phrase but an expanding range of smaller search phrases. A rising tide lifts all boats.

PR is not just about link popularity; it is Google's measure of how much it trusts your website. Your backlinks are a part of that but so are other factors such as how old your website is and the number of years into the future you have renewed its domain name. On the negative side, it reflects any of your practices that might be considered spammish (e.g. hidden text on the page, etc.).

Msn no longer reports back-links and the only reason, in my view, that Google keeps the toolbar alive is as a face-saving gesture to Mr. Page who unwittingly created the whole links industry by creating the toolbar in the first place. Yahoo provides the best list of back-links. The best way to check on backlinks is to type this into the search engine's search box:

link:www.mywebsite.com

While Google only provides a small sampling of credited links using this search method, you can see many more links if you sign up for a Google membership and use their Webmaster Tools function. But you have to be aware that the links you see listed there are not necessarily credited links. That tool is more for general information about who is linking to you. You may be able to mine that info for various purposes but it is not considered to be information directly related to your PR levels. In this post at Webmaster World, Google's Matt Cutts says:

Do not assume just because you see a backlink that it’s carrying weight. I’m going to say that again: Do not assume just because you see a backlink that it’s carrying weight. Sometime in the next year, someone will say “But I saw an such and such backlink show up in google’s backlink tool, so it must count. Right?” And then I’ll point them back here, where I say do not assume just because you see a backlink that it’s carrying weight. :)

So, I suggest you rely more on the link:www.mywebsite.com search method keeping in mind that at Google you will never see more than a small sampling of credited links at best.

Recent Toolbar Updates. In May and November of 2007, Google re-calibrated PR. Why did this happen? Well, one answer coming from the moderator of Webmasterworld.com (here):

...that is a very common event when Google updates toolbar PR,as they are doing right now. There are constantly more pages being added to the web, so one way to understand this is that there is less PageRank to "go around". Many pages go down by 1 in the recalculation during every update.

Most sites on the Internet lost one PR level in the re-jigging in May and November 2007. Some websites lost two levels though that is most likely an exaggeration since a drop of 1.1 could appear as two levels if a site went from PR4.000 to PR2.999. The toolbar would show it as a drop from PR4 to PR2. I assume that all partners in our program lost PR even if their toolbar display did not change. Simply, a drop from PR4.999 to PR4.000 is a significant drop even if the toolbar display remains PR4. But most SEOs saw this recalibration as harmless as few sites actually lost real rankings. That appears to have been the case in our program as well.

Just yesterday I was editing my book (The Dao of SEO) for its second edition. In the 2007 edition, I use an example mentioning that the #1 site on Google for the search phrase 'earn extra income' is a PR6 (not one of our program members). In the 2008 edition, I have edited that to convey that the #1 site for the search phrase 'earn extra income' is a PR4. That’s an illustration of just how much Goolge has recalibrated the toolbar this past year. It’s hard for many of us to get used to that but the reality is that it has largely been harmless with few sites losing rankings as directly related to that change.

Historically, I have observed that new program members go from PR0 to PR3 in about 3-4 months (about one toolbar update) and they get to PR4 within 7 months to one year. It's only rarely that I've seen sites go from PR0 to PR1 or PR2. It was though they flew right past those levels and landed at PR3. Now, with the recalibration, partners seem to get to PR2 in maybe two updates (3-7 months) and PR3 by the end of the first year.

What difference has this made to client rankings? Not much, as far as I can tell. The change really was a recalibration, not a punishment. But perception-wise it has hurt our program a bit. Just yesterday I advised a client that I would have to remove him from the program because he was non-compliant. He replied that he'd removed the partner pages because, as far as he could tell, the program had not helped; he was only PR2. He sells Feinmaster Tools. I checked Google. He is #1 for 'feinmaster tools'. The #10 site is also a PR2. Post-recalibration, PR2 is enough for him to rule his niche. Most likely his PR2 is higher than the PR2 of the #10 site. But to him he is ONLY PR2. So, dealing with assumptions and expectations in a constantly changing environment is a program challenge. (With a better understanding of this stuff, he has now restored his pages and will be program compliant.)

The #1 site on Google for 'seattle real estate' is a PR4; the #10 site is a PR3. That's the new Internet.

Folks, TBPR (Toolbar PageRank) means almost nothing. For an interesting discussion of why many SEOs would be happy to see it disappear see this article.

Spam linking and the danger it poses to your PR. In my view, the biggest danger to client PR (and rankings) is spam behaviour. But before I give you some examples of that, let me give you a prescription for success. I've always said that our program is the only 'program' that partners need. We provide roughly 200 one-way links. That's plenty. A broadly based program provides a measure of insurance or protection against the vagaries of the Internet. We might lose a few PR3 partners but pick up a PR5. Or, we might lose a PR5 and pick up half a dozen PR4s. Or, one of our partners might lose PR while another partner gains PR. Our program, due to the size of the membership, stabilizes PR in a way not possible with something like individually purchased text links. An individually purchased text link might cost $40 a month and so most small businesses can only afford a few of those. When PR is supported by only a few quality links, PR is more susceptible to changes in the PR of those supporting links. Our broadly based program is a better source of stability. And, of course, our membership club-type approach keeps costs down.

But our program is not enough. Even with the recalibration, our program can take partners to PR3 within one year. It's too early to know how long it will now take us to move partners from PR3 to PR4 although historically getting there in one year was a given (almost). But most certainly, you can get to PR4 and beyond by broadening your link base.

First, you must try to get into the dmoz.org (Open Directory Project) directory. It serves as Google's Directory. These days most folks don't even know that Google has a directory (directory.google.com/). A listing in the dmoz.org and subsequently the Google Directory helps PR. Since dmoz.org is editor-driven (rather than crawler based), Google trusts it more. A vote for your site from a dmoz.org editor is a significant one-way link. However, it's not easy to get into the dmoz.org directory. In some categories, I've heard that the waiting list is two years.

Apply to the Yahoo Directory. It is a powerful one-way link. It costs $299 annually. It is free if you are a non-commercial website. Historically, Yahoo was a directory, but a few years ago it added an engine (i.e. crawler based listings). So, today, you do not need to be in the Yahoo Directory in order to be in Yahoo. But it is a good idea. A single link from the Yahoo Directory will be worth more than 99% of the links programs out there including the links programs that promise to submit you to dozens of directories. Google views most directories as spam link programs and discounts their value. But Google takes Yahoo seriously. Again, it is about having an editor-vetted inclusion on Yahoo and the simple fact that you are sufficiently committed to your business to invest in its success.

The Yellow Pages is credible. Local geographic directories and industry specific directories are also more likely be to be considered credible. But I say that with caution. There are tons of real estate oriented directories out there that aren't worth spit. These days it seems like the world only needs 3 or 4 general search engines (Google, Yahoo, Msn, Ask), so how many general directories do you imagine the world needs? Probably, the same number. So, stay away from services that offer to submit you to hundreds of general directories. Most such directories exist only for the purpose of providing one-way links and the engines know it. You might be providing one-way links from your website but your website does not exist solely for that purpose. The engines know that, too. Your website likely has more PR than most of the directories out there. There may be a few small inexpensive general directories that will bring you some PR. Two that come to mind are JoeAnt.com and BOTW.org.

I've discussed article writing in past newsletters. Just remember that each article you write is one credited backlink no matter how many websites publish it. Actually, I just came across Work.com today. You can publish a how-to guide there for small business and get a link.

Mostly what I'm saying about getting links is to be selective and take a varied approach. Our program is a foundation that works on Google (see below). But it may not be enough. Spend your money wisely. If it's too good to be true, it's too good to be true. If it's effortless, it's useless.

Almost all backlink on the Internet are spam links. If you ask your cousin to link to you in order to help your link popularity, you are spam linking. It was the old Inktomi engine (most of you have probably not heard of it) that invented link popularity. It was Google that popularized it by inventing the tool bar. Once people became aware of link popularity as a ranking factor, they started exploiting it. That spawned the reciprocal linking industry. Google knows that almost all reciprocal links are self-serving artificial links. But Google has to assume that at least some links are sincere, but which ones? So, Google arbitrarily decided to grant a very, very small measure of credit to industry related (so-called relevant) reciprocal links. It wasn't because industry-related links were truly more relevant than other links but it was a way to rein in rampant reciprocal linking. A compromise, of sorts. However, it always has been the case that the best (most 'relevant') link for the Alaska fishing charter company is the one-way link from the banker in Iowa who recommends the service after having chartered one of company's boats while on vacation. That is much more valuable than the reciprocal link from the bait-cutter down the road. The reality is that both links might be sincere, or both links might be insincere (spam/artificial). But Google will feel some obligation to count both because the bait-cutter's reciprocal link is industry-related and the banker's link is one-way. But, the banker's link will count for much more, if my research is any indication.

This whole notion of 'relevant' links has spawned another myth and I still get emails from people asking if our links are relevant. Have a look at the PR6 (last year PR7) website at http://www.melbourneit.com.au/, a domain name registration website. Check its backlinks by pasting this into the Google search box:

link:www.melbourneit.com.au/

Notice the backlinks from the farm bureau and the stem cell research website. Relevant links, indeed! And, most of us only dream about being PR6.

We are all spam linking the engines and the engines know it. But they created this monster and there's only so much they can do to harness it. Penalties are largely a myth for all but the most egregious behavior (examples below). A lot of us enjoy a certain sense of self-importance that includes the belief that Google is after us. There are nine billion web pages on the Internet but we figure that Google has dedicated its resources just to snare us. It's just not so. Google generally ignores what it doesn't like. Just like it ignores the meta keywords tag and has ignored it since year 2000. The vast majority of links programs, over 99% as far as I can figure, use scripts. Scripts make things easy for everyone including easy for the engines to spot. Scripts deliver the links instantly. No effort required (unlike my program). Here's how the engines use this reality to catch spammers. Let's say you received three new links last week and you receive three new links next week and that's how things have averaged out for some time. But all of a sudden tonight, thanks to a script, you receive 100 new links. That's a spam alert. Google won't know if all or some of those 100 new links are spam links but since the receipt of those links greatly exceeds your average rate of receiving links Google will (most likely) ignore about 97 of those links. That's why you have to do some work when you join our program. No instant scripts. No one receiving 200 new links over-night. The fact that our program requires work on the part of partners is one of the reasons why it is fairly successful (and one of the reasons why it remains small). Our program was the first programmed approach to one-way links but most other Johnny-come-lately one-way links programs are much bigger now and richer, though of questionable value given their use of scripts.

Partners who are pretty much only in our program seem to get to PR3 within one year. A year ago, that level of achievement would have been disappointing. Today, it is quite acceptable. But there are people in our program, who after one year, remain PR0. When I see that, I pretty much know why. They have engaged in egregious spamming behaviour and they ARE being overtly penalized. Go to Yahoo and do this search:

link:www.listen-to-music.co.uk

You'll see that this site (as of February 18 2008) has 42,300 backlinks. Last month when I made a note of this example for the newsletter, the client had 40,311. Is it realistic to think that over 40,000 websites would have taken the time to sincerely link to this website simply due to its high quality?

If you do the same search at Google ( link:www.listen-to-music.co.uk ), you find approximately 80 backlinks (discounting some that are simply internal backlinks from his own site). Remember, Google only provides a sampling of credited links so this is only a sampling of an unknown total of Google-credited links. We know that altogether, this client has 42,300 links. We know that the most he could have from our program is about 200. We contribute less than one percent (.472) of his total links. Yet, 32.5 percent of the links listed (26 of 80) in Google's sampling are ours. Our program partners are doing their part; Google is accepting their contribution, but this site is only a PR1. In my view, it has been penalized for egregious behaviour. This website is hauling some heavy baggage - thousands of spam links no doubt delivered by a link farm and/or script. (Note that depending on the data-centre you access, you may see results somewhat different from the above. But the proportions should be comparable.)

Here's a site that's only in our links program, no other programs: http://www.ibuyhousesfast.ca. It's showing about 200 backlinks at Yahoo. It's been in the program just over one year. After one year, it is a PR3 and is #1 for its main targeted search phrase. Is it better to have 42,300 backlinks or 200? Is it better to be PR1 or PR3?

The above examples are NOT exceptions to the rule. I see it all the time. Every day I examine data relative to PR. Almost everyday, I see the same sort of thing.

Here's a client (http://www.christiandebtlaw.com/) who runs a modest links campaign based on our program. He is a PR3 with all of his Google credited links coming from our partners. This site has been in the program for less that one year. But the same program partner has another website that tells a different story. His other site at http://www.frequencyrising.com has been in the program going on three years . But in this past year, his PR level has dropped to PR0. He has 38,600 backlinks listed at Yahoo. When I made note of his site last month in preparation for doing this newsletter, he had only 31,000 backlinks. Is it realistic to pick up 7600 backlinks in one month (to say nothing of having 38,600 altogether)?! Less than one percent of his links could be coming from our program. And, yet a good many of the links credited for him at Google come from our partners.

When I told the client about this, he expressed shock. Later he wrote back to report that he was still being billed by another link exchange service. He didn't realize this was happening. That's true for so many. When we first start out we don't know what we are doing.

I keep mentioning PR3 partners but here's one of our many PR4 partners: link:kpropaintballnetting.com (he was PR5 before the recalibration). Google is listing about 38 external links for this client. About 30 of them are ours. The rest seem to be industry related (possibly reciprocal) links including a few from an industry source: paintballnews.com. Link popularity is like voting so 6 votes from paintballnews.com is no better than one vote from paintballnew.com. One website, one vote. Anyway, this client uses our program as a foundation (about 200 inbound links) and then more or less judiciously seeks a small number of other links partners specific to his industry (about 60 backlinks). PR4 is the new PR5; I'd like to see you all get there (or beyond).

Buying text links. Recently, I've addressed another brush fire with the announcement that Google was going after people who sell text links from their website. As important as we may like to feel, Google was not targeting our program. I do not sell text links from my website and I don't think you do either. Some people sell text links from their website either from their home page or from a higher PR lower level page. Maybe they'll sell a good PR4 link for $40 a month. This is the most difficult type of spam linking for Google to catch except for the links sellers who advertise such sales on their website. Here's what happens if you are caught. You are penalized. What happens to you if you buy a text link from a seller. Nothing. Except that the PR4 text link you bought may now be a PR0 link and you're probably not going to get a refund. I heard this directly, in person, from the mouth of Goolge's Adam Lasnick at a conference last summer. I don't think many, if any, of you sell text links from your site but if you do then don't advertise it. If you buy a text link from someone, buyer beware. But the most you'll lose is a bit of money.

Anyway, the court case. I can't tell you much about it at this stage and I can't identify the litigants. But the case contends that keywords (such as in the meta tags) are proprietary which is to say they are like private property or maybe like something copyrighted. If you steal someone's keywords, you are doing harm to their business and you can be sued. Doh! It's really dumb. The defendant tells me that the accusing litigant has now been counter-sued by two other companies and is running scared. My client's case may never get to court but he's losing money every day he's shut-down pending resolution. I have prepared a brief for the judge. I may post it as a newsletter once things are settled. Claiming keywords as proprietary is like peeing into the ocean and then trying to figure out which part of the ocean belongs to you. Well, I don't know if that is a good analogy but this has been a long tiring newsletter and the analogy just came to me.

Now I've gotta pee.

Cheers
Ken