Best SEO Blog|Launch your site.

Yahoo! and SEOmoz now make it possible for people to structure their link analysis reports so as to identify linking resources with diverse link profiles. That is, through either tool, you can see how many domains link to a particular domain. It is inevitable that the SEO community will begin setting minimum requirements for domain-level connections in link profiles.

That will be a tremendous mistake that competitive SEOs can take advantage of. Here is why it will be a mistake to focus on number of domains in a link profile:

1. Linking resources can fluff up their domain profiles by simply creating a lot of little sites that link to a single domain.
2. Links from large social media sites don’t necessarily pass value, but those sites have huge domain-level link profiles
3. The number of domains reported by any one tool is irrelevant to the data maintained by other services

Although there is an intuitive momentum pushing SEOs towards obtaining links from domains that receive links from many other domains, it does not follow that a well-linked domain will pass value to any sites it links to. Remember that SEO is not all about links.

Link profiles are useful for snapshot analysis. When you have to form a gut feeling on the spur of the moment in order to decide whether to obtain a link from a resource or not, a link profile may help you make the decision. Just because a site obtains a lot of links from many resources doesn’t mean those links pass value to the site in any search engine’s database. You want to know WHO is linking and HOW they link and WHY they link.

But that information is insufficient to help you understand what is happening in a search engine’s database. Knowing that site A links to site B doesn’t tell you whether Search Engine 1 knows about the relationship or whether Search Engine 2 credits the relationship. Knowledge of a linking relationship does help you assess a Web site’s probable visibility.

Given a choice between obtaining a link from a highly visible site versus an unknown site, you generally are better advised to take the link from the high visibility site. Even if the link passes no PageRank-like value or anchor in any search engine, it may still create visibility for your brand and/or pass traffic to your site. For that reason alone you should not reject links simply because they are embedded in Flash or Javascript, handled through redirection, or modified with “rel=’nofollow’”.

Link Building should be founded upon solid criteria that meet more than one expectation (passing anchor text and PageRank-like value constitutes a single expectation because they are so closely tied together). When you’re evaluating potential linking resources, you should assess their:

1. Web Visibility (how many domains link to them)
2. Search Visibility (how many query spaces do they rank in)
3. Credibility (are people likely to follow their links)
4. Ability to pass anchor text and PageRank-like value (naturally)
5. Hub Potential (are they treated as central resources by their Web community)
6. Perceived Authority (do people treat them as knowledgeable and trustworthy)

Any good linking resource should rate well on at least 2 or 3 of these criteria. You want to be careful, however, because links are not endorsements and people may not necessarily follow links assuming the destinations are being recommended.

Since links are NOT endorsements, the more value you are able to create in the link you position, the better. That is, if you control or influence not only the anchor text but also the text surrounding the link, you want to be sure you make that text useful and informative, helpful to the linking site’s visitors.

The more unique link placements you create, the better. If your linking resources are of such similar quality that you can rely upon a tool to fill out submission forms, it doesn’t matter how many domains link to your site. You’re probably not achieving sufficient link diversity to help yourself much.

If you’re going to build links, quantify everything as much as possible. Use your time wisely and don’t give in to unreasonable fantasies and expectations. Keep your link building lean, efficient, and powerful.

There are some pretty interesting theories going around concerning the Google sitelinks. I won’t pretend to have the magic formula but based on what I’ve seen other people suggest (and that I have evaluated for myself) and what I have seen, I would say there are some things you can do to improve your chances of getting some sitelinks to appear in your Google search results.

However, there are some caveats.

First, the sitelinks are more likely to appear for your site name than for any specific query, but some sites are so closely associated with queries that they do get the sitelinks for those queries.

Second, the sitelinks can appear for sub-domains but they may include references to unrelated parts of the main domains. This seems to happen with mega sites that are closely associated with multiple queries.

So, what can you do to get some sitelinks going?

1. Divide your site into categorized sections using folders and/or sub-domains
2. Link to at least three sections of your site from your home page
3. Optimize those three-plus sections of your site so that they rank first for active queries
4. Cross-promote those three-plus sections of your site from every other page on your site
5. Obtain links to those three-plus sections of your site other domains, using appropriate anchor text and surrounding text

Think of a sitelink as a brand that is specific to your site. The sitelink represents a section of your site that Google deems to be popular and important.

You can monitor your sitelinks success by checking Google. When you’re logged in to Webmaster Tools, you can see if there are any sitelinks for your site. You’ll also see when they were last updated.

Every sitelink I’ve been able to document through my own sites has met the criteria given above. Some people speculate that Google is also tracking click-throughs, and I can confirm that these pages do get click-throughs, although some of them don’t get that many (we’re talking maybe a few dozen click-throughs per month for some sections).

It may be that a site has to meet a certain “size” requirement. Google does seem to have a pretty decent algorithmic sense of what constitutes a Web site. For example, it correctly associates sub-domains on Xenite.Org with Xenite.Org, but if you look at some very popular, well-populated free hosting sites, you’ll be challenged to find any of their member sub-domains or pages appearing in the sitelinks. So it follows that the algorithm may require that a minimum number of pages be associated with a “site” before it qualifies for sitelinks.

Do all the sitelink destinations have to rank first for anything? Not necessarily. Nor do they even have to rank first for a site search. But all the pages I see in my sitelinks (and in other people’s sitelinks) do rank competitively for at least one query — which only means they are probably earning direct click-throughs on a consistent basis.

That is why I think you need to optimize sections of your site and promote them on their own merits in search engine results and on other Web sites.

If you don’t have at least three such sections on your site, you probably won’t get any sitelinks.

Search engine optimization should follow a structured process in order to achieve maximum efficiency of effort and investment of resources. Because there are many types of SEO services, you need to think about your SEO plans differently. A high-level plan, or strategy, needs to focus on which aspects of search engine optimization will be addressed. Many companies utilize their proposal literature to convey high-level plans but in a large-scale search optimization project a proposal is inadequate.

The SEO Strategy Document

A search engine optimization strategy must define the scope of the campaign: which search engines are being optimized for, which verticals are being targeted, and approximately how long will the campaign run.

The strategy also needs to set achievable objectives so that your key performance indicators can be measured against those objectives. For example, your strategy may need to build visibility for a site in several search services; it may need to improve search referral conversions; it may need to expand the number of targeted queries. The strategy can be complex but it cannot be comprehensive.

That is, no SEO plan or strategy ever lays out the end-game. Once the campaign strategy has been deployed, time moves forward and the results will eventually change; or assets will grow or change; or business objectives will change. There will always be a need for a new SEO strategy, either to cover issues that a previous strategy didn’t address or to respond to changes in the competitive search environment.

The SEO strategy document should assess the available resources and outline how they may be used to achieve the objectives. The SEO strategy can only propose solutions, not authorize them. Someone at some higher level has to approve the strategy (when you’re working for yourself, that means you need to set the strategy aside for at least a few days and then come back to it with a fresh perspective).

The SEO Research Plan

Keyword research can usually be divided into two areas: Query Space Research (where you identify the queries used to find the content you’re optimizing) and Trends Analysis (where you identify the seasonal, annual, and event-driven queries within the query space).

If you ignore trends analysis in your research plan you’ll sort of limp into competitive search engine optimization with at most half the knowledge you need to be effective.

In addition to keyword research you need to identify competitor sites and determine what strategies they have pursued for optimization. Don’t make assumptions. If you conclude they are basing their optimization on link building, back up your conclusion with hard data. If you conclude they are basing their optimization on on-page factors and internal links, back up your conclusion with illustrative screen captures. Include this data in your plan so you can go back and refer to it.

The research plan outlines what you intend to find; if you put your findings into the plan it’s actually a report and that means you’re doing this by the seat of your pants (or by rote — experienced SEOs rarely need to write out a plan for research).

The Web Site Optimization Plan

You need to set the priorities for the site. Everything should be discussed, including site architecture, navigation, internal references to other documents, how outbound links will be placed in copy (and how the anchor text for those outbound links should appear), and how the site will be leveraged to assist other sites.

Believe it or not, the majority of SEOs don’t think about how to use the site they are optimizing to build alliances with other Web sites. If there is a critical directory that your site is listed in, which your competitors have not yet flooded with submissions, but which may be useful to your visitors, find ways to link to that directory. It doesn’t have to be a directory to be a useful resource that deserves a link. The links you provide should be obviously positioned to benefit your visitors.

Your on-site optimization plan should also set guidelines for how to populate title tags, how to organize content on the pages, how to write the Hx or large-size font tags, how the pages should be named, how the images should be documented, etc.

Your plan also needs to specify what copy should be included on the site: name the articles, describe the articles (these descriptions can serve as meta description tags or as models for the tags). Indicate how long the articles should be (let the topic define length; don’t use some artificial “number of words per article” SEO formula).

The Web Partners Optimization Plan

Some SEOs include this plan in their Link Building plan. It doesn’t matter if you break it out as a separate plan or merely include it in a larger linking plan.

Web partners are other sites with which your site has a special relationship. Maybe they’re all part of the same network. Maybe they are business partners who have a close relationship with you(r client). These sites are different from every other site on the Web because they WILL promote your site and they WILL link to it per direct request.

Utilize Web partners to their fullest potential if they are willing: don’t just ask for links, ask for featured content, cross-promotional advertising, anything that makes sense for their visitors. The more they talk up the site, the more credible it becomes to their visitors.

The SEO Link Building Plan

Link Building is NOT search engine optimization. Link building complements search engine optimization and assists it. Most decent Web sites will attract links anyway. The Link Builder’s job is to accelerate the process of link acquisition and to expand the potential linking resources that will help the site build Web Visibility and Search Visibility.

Every known potential linking resource should be listed and documented: its value should be articulated in any way other than referring to Toolbar PageRank (that means nothing). Will the resource send its own referral traffic? Is it known to pass anchor text through its links? Does it get cached often? How well does it rank in complementary queries, where its search visibility lends credibility to its link destinations? How relevant are the sites it links to with respect to its own content?

Your linking plan should also explain how you intend to encourage other people to link to the site. Will you distribute press releases, widgets, special testimonials, exchange links, etc.? Write it down so that everyone involved in the project (including you) has something to refer to. No one will remember everything.

Your plan can also include link research, but frankly I feel that should be part of an SEO’s ongoing development and not part of any specific campaign or plan.

SEO Analysis and Reporting Plan

It’s standard practice today to monitor rankings on a weekly or monthly basis. Most SEOs tend to favor the monthly ranking report. Your analysis plan should go beyond that, however.

If the site is using some sort of analytics software, the KPIs that the analytics package can deliver should be specified. What units of time will serve as the reporting basis? When will trends reporting be done (this can be managed on a quarterly, semi-annual, annual, or bi-annual basis)?

What about alternative metrics? You need a sanity check. Your plan doesn’t have to commit to using any alternative metrics but it should recommend some so that everyone can do their own sanity check on the same resources.

The key performance indicators should be tied to specific parts of the SEO strategy where cause-and-effect can be reasonably expected (this goes well beyond your linking plan). How will you measure search visibility? How will you measure Web visibility? How will you measure the activity in the query space (both searches and sites positioned for those searchers)?

You need to set up a list of trigger points that determine when red flags should be raised. How many competitors can you tolerate in the top ten? How long do you want to wait before deciding an SEO plan has been effective? How much time should you allow for determining you’ve reached maximum positive benefit from implementing the strategy? How do you want to assess any possible filtering or penalization?

Conclusion

You can write an SEO plan in outline format and then fill it out with details, screen captures, graphics, and special notes. You don’t have to tackle everything possible in any one plan. You need to break up your objectives so that you can document the information in absorbable and executable chunks.

A plan that is too thin and sparse is a waste of time.

A plan that is too long, too detailed, and too complex or comprehensive is an even greater waste of time.

You don’t have to cover all the bases at once as long as you note that you need to come back and cover some other bases later.

“Why doesn’t Google see my anchor text?” Many people in the SEO community have been asking that question recently, especially since mid-August 2008, when a fair number of SEO bloggers concluded that Google is no longer placing as much emphasis on anchor text as it once did.

As someone who has long been advising Google to stop allowing ALL sites to pass anchor text, I would be surprised to learn that they had in fact devalued anchor text in their algorithm. Google’s historical approach to fixing the problems with its algorithmic assessment of Web sites has been to arbitrarily avoid looking at as many sites as possible.

In other words, Google filters Web sites out of all or parts of its index through a variety of methods. We know Google uses filters of various types because Googlers have admitted as much. So the argument that Google no longer values anchor text as much as it once did is highly speculative — it could simply be that Google no longer values as many links as it once did.

In fact, we have seen Google knock millions of sites out of its index, and filter out many millions of links, on several occasions in the past. The August 2008 Google Link Slam could have been nothing more than the effect of filtering out several million suspect links. People would have lost rankings and, not knowing which of their links no longer worked, they could have concluded that all the links lost some rather than that merely some links lost all value.

So if you’re out there pointing links at some link-poor Web site, and you’re not sure if Google is allowing your links to pass anchor text, here are a few things you can do.

1. Test your linking sources to see what they rank for
2. Test outbound links on your linking sources to see if they pass value to other sites
3. Point links at your linking sources to see if your site can pass anchor text

If your links are not working it could be that your links cannot pass value, or that your page cannot receive value, or that you’re not waiting long enough for your page to receive value.

You need to clock a linking source’s ability to pass anchor text before assuming it will pass anchor text. If you can embed more than one link on a page, put two links on there: one to help your site and one to act as a control. The control link should point to a page on a highly trusted, well-known site with anchor text that is relevant to the content of the destination but for which the destination page does not rank.

If your control destination ranks for your anchor text but your site doesn’t rank for your preferred anchor text, there could be several reasons why. For example, it could just be that you’re competing for a hyperoptimized expression and you need a LOT of value-passing links (in which case, you should not be wasting your time with that expression). It could also be that you’re trying to pass too much anchor text. It could also be that your site has been placed in the Supplemental Results Index and it just needs more value-passing links. And it could be that your site has tripped some algorithmic filter.

Algorithmic filters can be somewhat benign. For example, the presumed aging factor applied to inbound linkage may be alleviated based on the performance of the destination page. That is, your linking source may not normally be allowed to pass value for a few months unless it links to well-established, highly trusted domains.

Google more than likely sees your anchor text. But if you’re wondering why Google isn’t allowing your anchor text to pass to your chosen destination, the most likely reason is that you’re putting links on weak pages, or penalized pages, or pages that trip filters, or your own site just needs more trusted, value-passing links.

Be as choosy as possible when it comes to link acquisition. If the links you obtain don’t pass value, you’ve wasted your time. Focus on evaluating linking resources rather than on link building. You’ll develop a better feel for how to build value-passing links faster than you thought possible.