Link building versus search engine optimization

Posted by Michael Martinez on May 7, 2008 in Link Theory

In case you missed the online reputation management asssociation post, there’s a link for you.

Debra Mastaler wrote about the May 2008 Virginia SEO meetup yesterday and I read with interest the couple of paragraphs beginning with “It was really surprising to hear the diversity in the answers to ‘what’s working in link building’”.

I was not surprised to see that people were claiming various things were still working. In fact, everything works in search engine optimization. Things don’t work when you cross the line between tolerability and intolerability. That is, search engiens are not so much accepting of content as they are tolerant of content.

Tolerance is the measure of how many straws you can pile on the camel’s back before the camel collapses. In search engine optimization, too much of anything is a bad thing. So if you’re building reciprocal links with relevant sites whose content actually has some connection to your own content, that’s a good thing. It always has been and it always will be. You’ll get better converting traffic from links followed by people who are already looking for your content than from randomly placed spammy links.

That is, search engine optimization complements good Web marketing practices. Although it would be difficult to argue so, I think that has always been the case. After all, before search engines allowed links to pass anchor text you were still more likely to be crawled through link-rich sites than through link-poor sites. Crawling is an incredible gift to bestow upon any Web page and it is very much under-rated in the SEO industry.

Debra also pointed out that press releases work very well for many people. Online press releases have been a tool for SEOs to work with for years and I’ve used them many times, but not for link building. You can get links from press release sites but the time and effort it takes to create a press release make the links expensive. Regardless of whether you buy links, create links in free articles and press releases, submit listings to directories, etc. the time you invest in capturing one link to your Web site is a real cost of acquisition for you.

If you’re averaging 2 hours or more per link, you’re creating very expensive links. How much is an hour of your time worth? What do you get for that hour? When I create a press release I may invest as much as 1-2 days in the process. I want a lot more than a mere link for that kind of investment. It’s not worth it to me to create press releases for the sake of links. I want more from a press release than just links.

Some people are a little concerned that Google Analytics has excluded link building from SEO. I’m not. Their definition of search engine optimization is technically accurate and I have to agree with Danny Sullivan that people are reading too much into the post.

Just because you can easily find a lot of SEOs who wrongly equate link building with search engine optimization doesn’t mean that link building is search engine optimization. Link building doesn’t optimize anything. It doesn’t in any way improve performance. You’ll be quick to disagree with me by pointing that “one really good link can make a difference!”

Absolutely. One really good link can pass you traffic, PageRank, anchor text, visibility, credibility, and much more. And none of that has to do with search engine optimization. SEO Theory defines search engine optimization as “the art of designing or modifying Web pages to rank well in search engines.” The sexiest man in search likes to add “and drive converting traffic” or some variation thereof to the end of the formal definition, but I often exclude that qualification because in fact sometimes your search engine optimization is not about driving traffic.

For example, let’s say you’ve read all the SEO blogs and have gone out to social media sites. You’ve created profiles for yourself and have dropped links to your content. Although some people report success from that tactic, many people fail miserably to improve their search results. Why? One reason is that they are creating worthless links on worthless pages.

It’s one thing for you to create social media content that other people link to (either from within the social media resource or from other sites). It’s quite another for you to create instant profiles that no one links to and which accrue no perceived value. Unless you can naturally attract value or are very good at choosing social media sites that just happen to share their value with your randomly created content, you have to build value in your linking pages.

That is, you have to optimize your linking pages for linking. Linking optimization is all about the PageRank, which is why Google opposes the creation and use of “links to manipulate PageRank”. Linking optimization builds and passes PageRank — it manipulates PageRank. When you run around to SEO blogs and forums, dropping your signature links — or when you build your social media content — you’re optimizing your links because you are building links.

That is, you place your links where you think you’ll get value (PageRank) to pass value. Technically, any link that uses “rel=’nofollow’” would be in immediate violation of the “don’t build links to manipulate PageRank” rule were it not for Google’s explicit exclusion. There are implicit exclusions, of course, such as “links that you don’t control and which manipulate PageRank” are fine. If CNN points a link to your site from its home page just to create a little visibility for you, Google probably won’t blink (NOTE: I’m not talking about the buying and selling of links).

A visibility-building link would imply a relationship exists. I’ve linked to friends’ sites from my own sites on occasion. There is nothing inherently wrong with linking to someone’s site because you want to help build visibility for them. That’s what the World Wide Web is all about and search engines have to respect that fundamental aspect of linking. But if you’re linking to manipulate PageRank, PageRank really isn’t visibility. The manipulation of PageRank is a deliberate attempt to alter search results in spite of the algorithmic preference to allow PageRank to flow naturally.

That is, manipulating PageRank is contrary to the best interests of a search engine that relies upon PageRank. Hence, manipulating PageRank is not optimizing the search results. But does that mean that moving content from Google’s Supplemental Results Index to the Main Web Index is not search engine optimization? Not at all. After all, if Google is allowed to create exclusions and exceptions for its rules, search engine optimization specialists are, too.

But the exclusions and exceptions have to be reasonable. Link building for the purpose of accruing PageRank looks very much like link building for manipulating PageRank. It’s too fine a hair to split, in my opinion, but nonetheless you have to accrue sufficient PageRank to move your Supplemental pages into the Main Web Index and that is search engine optimization.

Link building stops being search engine optimization when you build links because SEOs tell you that is what you must do. The difference is that SEO link building tends to be a mindless exercise. There is no goal in mind except to “improve rankings” — meaning you want the anchor text in those links to boost your destination page higher in the search results for a specific keyword.

To put it in a much broader context, which is the more optimal process: to refine a specific page’s ability to rank well for a small number of keywords or to empower a page to rank well for many keywords?

Search engine optimization fails when you obsess on specific keywords rather than set the pace by emphasizing specific keywords. That is, if you create a Web page, you should be asking how to get it to rank for as many active relevant query expressions as possible. I include relevance because irrelevant traffic is not optimal. That is, if your page only ranks for non-converting traffic then you’re not optimizing for search.

You can point a lot of links to a page with different anchor text. In fact, it is easy to get a completely blank page to rank for numerous expressions by pointing links at it. Many people have done just that (and that is one reason why expired domains have become so popular among domain speculators). But it takes more time to get the links than it does to put optimized content on the page and include it in a larger Web site that is indexed and uses internal links to help search engines find and crawl pages.

Let me put it this way: I can create several thousand long-lasting value-passing links in a day. How many days does it take you to obtain 1,000 value-passing links that will last a year or longer? My links pass value because they are designed to do just that without crossing the implied boundary that Google has set. That is, if I create a 1,000 page Web site (that contains unqiue content and serves some useful function) the internal links I create for that site will make or break its search visibility.

If I can only acquire 100 value-passing external links per week for any Web site, but if I can create unlimited content for the same site, where is the most link value going to come from? Link value is not measured only in PageRank (which, like peanut butter, can be spread rather thin). Link value is also measured in traffic that is passed, visibility that is created, credibility that is conferred, etc. An ecommerce site that won’t link to its own content has a credibility issue as much as it has a crawling issue. Why should the consumer trust the ecommerce pages if the site won’t?

It is of course obvious that internal link building is vital to search engine optimization as it controls the crawling process, but external links are not vital to search engine optimization. You may want to point to all those competitive queries where you need links to rank well but those queries are the exceptions, quite rare compared to the millions of query spaces where rankings are not determined by link weight.

The number of truly competitive queries that require link building remains less than 1,000,000. Don’t believe me? Let’s play with some numbers. The United States Census lists 1265 incorporated cities with populations of 25,000 or more as of the 2000 Census. How many competitive verticals can you think of that are broken down by city? Real estate, travel, hotels, vacations, cars, etc. Each of us — without any help from our SEO friends — could probably easily think of a dozen, with some effort we could probably list 20-30. Get together with some friends and you should be able to compile a list of 40-50 such verticals.

50 verticals that can be optimized for U.S. cities would produce about 65,000 competitive queries. Now, I’m not talking about how many pages you can generate based on city and town names. I’m talking about using how markets are worth competing for as a metric for estimating how many competitive queries there may be.

In addition to activity and industry verticals there are products that are highly competitive. For example, everyone loves V*iagr*a and its variations, but there are also shoes, shirts, cheap leather purses, discount coupons, etc. But can you think of 50 off the top of your head? Most people cannot. The relatively few spammers who specialize in product-based spam undoubtedly have lists of thousands of items but most people are not competing for those queries.

Hence, the number of competitive product-based queries is relatively small. I’ll be generous and estimate that maybe 100 products are highly competitive (requiring links to rank) in our 1300 city markets. So there is another 130,000 competitive queries for you.

Now, where things get interesting (in my opinion) are in news and entertainment. Celebrities and news makers number in the thousands — perhaps tens of thousands. Unfortunately, no one optimizes for “Britney Spears Miami Beach” so we cannot multiply our celebrity and newsmaker estimates by 1300. But I’ll be generous and conclude there may be 25,000 people who create headlines often enough that their queries are competitive. Personally, I cannot think of 25,000 celebrities and newsmakers, and I doubt I could produce such a list if I gathered 50 people together.

65,000 + 130,000 + 25,000 doesn’t even bring us close to 1,000,000 competitive queries. You can throw in the Fortune 10,000 if you want, or the Forbes 5,000 (I don’t even know if such links exist). You can throw in all the sports teams, all the schools and high schools with sports teams, all the athletic and intellectual competitions — you’re not going to reach 1,000,000 competitive queries.

That’s a very rough argument and it would be helpful to the SEO community if people spent some time trying to estimate how many competitive queries there should be (it won’t break my heart or embarrass me if someone finds a way to top 1,000,000 — search engines report 7,000,000,000 queries are performed every month).

To be competitive, a query has to have at least two competitors and in this context I’m requiring that the competitors rely on links for their rankings (that is, they MUST be link building in order to rank). But just because someone chooses to build links doesn’t mean they have to. It only becomes necessary when you cannot outrank anyone on the basis of natural relevance (which can be VERY strong if you repeat the words hundreds of times on the same page, which DOES happen naturally in many different types of Web documents).

Search engine optimization is not concerned with how many queries are competitive. Search engine optimization is concerned with ensuring that appropriate content is visible for relevant queries. Although you can do that through links it’s better to do it through the combination of content and links (with emphasis on content rather than links). You can rank for more queries through content than you can through links; hence, link building fails to provide optimal results. Therefore, link building is not search engine optimization.

Q.E.D.

3 Comments on Link building versus search engine optimization

By Carlos on May 7, 2008 at 9:11 am

I like to add the “drive traffic” addendum because e-commerce and lead generation really have concrete end goals. I realize that in a brand protection or reputation management campaign you probably don’t want traffic for your phrases.

By Wiep on May 7, 2008 at 11:35 pm

“If you’re averaging 2 hours or more per link, you’re creating very expensive links.”

If you just look at the link itself, this may be so, but this also depends on the quality of the link. But a link is never just a link. It’s a brand exposure mechanism, it’s a way to attract traffic, it’s a way to attract more links, etc. A link that cost you over 8 hours to get may be a good catch, while pursuing that 5 minute link may have been a waste of time.

By Michael Martinez on May 8, 2008 at 5:54 am

Carlos, in brand or reputation management (or brand promotion) I most certainly WOULD want traffic for the protected queries. I think most if not all traditional marketers would expect some ROI in their brand-oriented campaigns.

Wiep, there has to be a metric behind the 8-hour link to show that it’s worthwhile. Most people chasing links don’t measure the value they are exchanging for their links. That’s really what I hope people will think about. It’s a business-oriented approach to link building: get more bang for your buck (your buck being the time you invest).

Comment

Log in or Register to post a comment.

More

Read more posts by Michael Martinez

About the Author

Michael Martinez is the Director of Search Strategies for 1st Query, an Internet Marketing firm offering organic SEO and PPC services.

Online Reputation Management Association Brief hiatus