Best SEO Blog|Launch your site.

There are some scary estimated costs-per-click associated with SEO expressions. Some of them are scary because people would be willing to pay so much for sponsored advertising on what appear to be trademarked names. Some of them are scary because of what people are searching for.

I’m going to pick on one scary SEO expression: “Guaranteed SEO”. Drop by any serious SEO forum and ask the regulars “what kind of guarantee can an SEO provide me?” and you’ll see feet shuffle, eyes cast down, and hear lots of hemming and hawing.

I hate discussions about SEO guarantees because the people who want guarantees don’t want guarantees on workmanship and product quality — they want guarantees on placement in search results.

A competent SEO technician with at least two years’ successful experience on five or more Web site should be able to help any site not receiving traffic to get at least SOME traffic. That same competent SEO technician should be able to help most sites improve their traffic, too.

You don’t have to be an analytical genius to find keywords that haven’t been optimized for. Nor do you have to reach for paid links or widget bait-and-switch schemes to figure out how to improve rankings for pages that are already getting some traffic from long-tail keywords.

But people in the industry want to act like Guaranteed SEO just doesn’t exist — can never exist. After several years of providing clients with projections and estimates and commitments, I’ve come to the conclusion that there really are certain things you can guarantee without painting yourself into a corner. You can provide some guarantees for work performance without having to meet unreasonable expectations about obtaining number 1 rankings in active queries.

Not that we cannot obtain number 1 rankings, but when you chase that top position long enough you eventually realize there is a whole lot more traffic to be brought in than any “hot” number 1 ranking will ever bring in for you. A well-optimized site with 400 pages of content should have thousands of number 1 listings in active queries.

So what kinds of SEO guarantees can guaranteed SEO offer without looking stupid and unprofessional? Here are a few suggestions:

• Basic SEO instruction — many clients don’t know enough to implement or understand what you recommend.
• Competitive SEO analyses — virtually worthless but they can help appease clients who think you aren’t paying attention.
• Intermediate SEO metrics — you can provide helpful feedback on bounce rates, time-on-page, benchmarks, etc.
• Link acquisition counts — easily verifiable if you provide a linking resource report.
• SEO Ranking Reports — everyone loves these.
• SEO plans — provided you didn’t make the mistake of including them in your proposals.
• SEO Site reviews — stock-in-trade for just about every serious SEO.
• SEO Traffic reviews — this is where you show the value of your work.

You don’t have to stop the list there but I will for brevity’s sake. Are these guarantees sufficient to justify the expense of hiring an SEO firm? Well, not, not unless you don’t mind paying for expensive lessons. You justify any expense through the return on investment, although ROI is measured in many different ways for nearly as many types of expenses.

Guaranteed search engine optimization cannot deliver everything, but very well-focused search engine optimization can guarantee quite a lot. You have to understand what you’re asking for or offering well enough to make sure you reach a common ground with the other side. In every negotiation one negotiator has an advantage of knowledge and it’s not always the expert who holds the advantage.

Guaranteed SEO works when both parties agree that the contract pertains to the work performed, not to specific changes in search results. You stay in this industry long enough, you’ll be asked more than once to deliver very specific search results. For most queries these kinds of guarantees are very feasible — that’s because most queries are not yet commercially harvested through SEO. It’s the competitive queries where people are actively vying for search conversions that one SEO’s ability to deliver the goods is compromised.

In fact, if a search engine insists on putting a specific site ahead of any others, there is nothing any SEO can do to change the search engine’s mind. That really isn’t what search engine optimization is all about. SEO is about achieving maximum possible results. In other words, the optimum performance may not be as good as the desired performance.

You cannot ever guarantee SEO success for an unreasonable request. That doesn’t mean you should turn down clients who want to rank first for “Britney Spears”. It does mean, however, that you had better reach an agreement with the client on what the point of ranking first for “Britney Spears” might be. I can guarantee you people will skip past an obviously irrelevant number 1 ranking in a search result if the site immediately below is obviously more relevant to the query.

As I am sure everyone in our industry is now aware, there is some very bad advice circulating in the SEO community about “Sculpting PageRank”. Even though it might look like Google finally slapped down the concept, it appears to me there are still some hardcore advocates who continue to preach the philosophy in secret. They may even be working on a new euphemism to disguise their intent.

Let’s face it: there will always be high profile advocates of dumb ideas in every industrial and political spectrum. Still, Google makes it clear they require some PageRank for sites that want to appear in the Main Web Index.

The Isolation of PageRank
PageRank is an entirely Googlian element. You won’t find it in other search engines like Ask, Bing, and Yahoo!. Oh, you’ll find something at each search engine that measures link value — but it’s not Google’s PageRank. PageRank is unique to Google and here is why: Google calculates its PageRank based on the data it collects (data that Google does not share with other search engines) and on the filters it employs (filters that other search engines do not use).

Since you don’t have access to that data or those filters, you have no way of knowing which links Google allows to pass value in its search index — hence, you cannot track and measure PageRank. Nor can you track and measure the other PageRank-like values used by Ask, Bing, and Yahoo!. You can collect your own data, implement your own filters, but you can only create your own PageRank-like value — you cannot track and measure PageRank that way.

How Well Can You Manage PageRank?
Imagine being given the task of altering the course of the mighty Nile and Mississippi rivers. You have immense resources at your command. You have the ability to create dikes, dams, canals, etc. Unfortunately, you’re not allowed to look at the rivers or to use scientific instruments to measure their capacities, rates of flow, etc. You have to blindly tell people what to do and those people are not permitted to tell you what happens when they follow your orders.

You do get some feedback. You can read news reports about where the rivers flood and which communities have access to water — but the reports are only published 4 to 6 times a year.

That’s a poor system for managing two huge natural resources but in this scenario you do manage them. You’re just not in a position to manage them well. That’s what happens when people go to extreme efforts to manage PageRank.

How People Can Manage PageRank
After all this time we’ve found only two ways to manage PageRank on our Web sites: add more links to increase PageRank distribution and remove internal links to restrict PageRank distribution. We have embellished both approaches with some variations. We can:

• Get more external links to point toward deep content
• Get more external links to point toward our root URLs
• Use “rel=’nofollow’” to prevent PageRank from flowing through our links
• Use “nofollow” in page meta tags to prevent PageRank from flowing through our links

Google of course recently shut down all attempts to restrict PageRank flow through Javascript, since Google now follows Javascript links. Some ideas never really worked, although people have proposed using them to manage PageRank flow. For example, using “noindex,noarchive” in your page meta tags won’t prevent those pages from accruing PageRank. If you don’t link out from the pages the PageRank will “evaporate” — more or less be distributed to the rest of the PageRank-receiving documents in Google’s index.

Some people have suggested using robots.txt files to disallow pages — but that doesn’t prevent them from accruing PageRank either. The fact is, if a value-passing link points at a document, that document will receive value. All you can do at that point is decide where to send the value next.

How You CANNOT Manage PageRank
Some people point to Google’s classic PageRank algorithm and say, “Well, they included this dampening factor that reduces PageRank as it flows through the system”. Therefore, they conclude, each page only passes on part of the PageRank it receives. Hence, one of the latest hypotheses in PageRank Manipulation Theory is that you should reduce the number of steps through which you flow your PageRank. Cut out all those intermediate directories and such.

But it doesn’t matter if you keep all your pages at the root level directory or if you spread them throughout your site. Your PageRank is still going to step gently or roughly from page to page and the dampening factor (for all we know) will be applied to every page’s accrued PageRank. Screwing up your server’s file system is not going to change the flow of PageRank. (Note: How does that screw up a server’s file system? Ask your local Unix guru how well a server handles 1,000,000 documents in one directory. The look of horror on his face should clue you in.)

PageRank flow is tied to logical structure, not physical structure. Oh, sure, a few years ago people noticed that Toolbar PR would be automatically adjusted to “estimate” what PageRank might look like if it had been calculated for a page. If your root URL had a Toolbar PR of 5 then pages in the root directory might get a 4 — as would pages in a secondary directory. Pages 2 directories deep would get a Toolbar PR of 3, and so on.

Of course, once you started getting natural links for your deep content, the Toolbar PR values would go up or down because they were being based on real PageRank. On my own sites I have had many deep content pages that achieved as much Toolbar PR as the root URL — and that was happening back in the day when people thought you automatically got “stepped down” based on folder distance from the root directory.

What You Should Not Be Doing To Manage PageRank
You should not be using non-Google resources to track and measure your PageRank. They cannot tell you anything about what Google knows or allows to happen. Anyone who claims otherwise simply doesn’t know what they are talking about. That doesn’t mean you can’t use non-Google tools for link research. If you enjoy spending your days counting backlinks, there are plenty of tools out there to feed your obsession.

You should not be attempting to block the flow of PageRank on your sites. Since you cannot track and measure it, you don’t know how much blockage is effective and how much is self-destructive. Internal PageRank is such a tiny little value, why do you think you can build it up to something useful anyway?

You should not be attempting to hide pages from search indexes. For example, if you have a privacy policy on your Web site, you should allow search engines to show it in their indexes. Why? Because people search for privacy policies. If you put “noindex,noarchive” in your privacy policy’s robots meta tag, you’re only hurting your site’s search visibility. And some people might be tempted to say, “Hey, I can put ANYTHING on this page because it’s not indexed”. Yeah, well, that doesn’t guarantee that a search engine will respect your “follow” directive.

You should not be counting the number of internal links each page contains. Link counting means you’re doing it wrong. Link counting means you don’t understand search engine optimization. Link counting means you’re thinking in terms of not passing PageRank.

What You Should Do To Manage PageRank
Start by getting more value-passing links to your Web site. Saying that is so easy it sounds trite, but that is where you should begin if you seriously want to manage PageRank for your site.

Use an open navigational system. Instead of hiding things and pretending you’re being efficient, just embed more internal links on your site. That advice, of course, leads people to ask, “But what if I put too many links on my pages?” The answer is, put as many links on your pages as will help your visitors. No more than that. How many is that? It varies by Web site. Artificial link count limits are useless pieces of fluff advice.

An open navigational system doesn’t have to be embedded in a single element. You can use multiple navigation tools. Just make sure your visitors don’t have to rely on the BACK button to get around your site.

For More Information, Read These Articles

• Large Web Site Design Theory and Crawl Management
• Optimizing Web Site Structrue Through Architecture
• Architectural SEO – Website Navigation Principles
• Site Search Optimization
• Using Multiple Domains To Build PageRank
• Duplicate Content For SEO and SEO For Duplicate Content
• Search Engine Love: Now They Crawl Me, Now They Don’t
• Prioritizing Content For Search, People, and You
• Site Search Optimization 2.0
• How To Rehabilitate SEO Advice That Has Lost Focus
• Architecting Web Sites – Design From The SEO Perspective

If I have timed this right, and if you subscribe to SEO Theory, then by now you may have read Managing Link Investment Risk. If not, go read that blog now and then come back here.

So, for those of you who understand what I’m talking about when I mention “Link Investment Risk”, here are three tips you can use to manage your risks. I briefly described one in the SEO Theory article.

• Use a spreadsheet to catalog your linking resources and strategies
• Assess your competitors’ linking investments (and link investment risks)
• Calculate risk-to-cost ratios

Before you do this, you have to make certain assumptions. You can adjust your assumptions over time as you improve your ability to assess costs. You need to assign a dollar value to every cost. For some costs the value may change over time. I recommend using average values rather than updating your assigned values frequently.

You want to assign value to the tasks involved in your link acquisition: analyzing sites, filling out forms, writing emails, whatever. If you were doing this for someone else, what would you charge them? That’s the cost you have to charge yourself. There is no profit margin to be concerned about.

Since you don’t know what your competitors earn or charge for these activities (unless they spell out the costs on their Web sites), just use your own personal cost factors for them.

Use a spreadsheet to catalog your linking resources and strategies
By “resources” I mean classes or types of resources, not individual Web sites. You decide how to group them.

You want to use a separate table for resources and strategies. You want to assign reasonable risks and costs. You can be creative in assigning a cost to a strategy. Think about how much time you would devote to a particular strategy for a campaign. Don’t think in terms of, “Well, I’ve been doing this for five years….” Most campaigns are executed in 3-6 months or less.

Be consistent with how you assign risks. If you use my proposed “foolish risk”, “prudent risk”, “unexpected risk” scheme you may find some risks are not easily categorized. It’s okay to expand beyond those three levels. I would not, however, equate “foolish risk” with “high risk” and if you don’t use “foolish risk” then you should use something else in addition to “high risk” to distinguish between risks that potentially offer a lot of return and risks that really just squander resources.

Costs should be based on time and materials, fees, etc. Use estimates where you must.

Finally, count up how many links you obtain from each resource class and strategy. If you want to really drill down, count how many of the links are indexed by major search engines.

Assess your competitors’ linking investments (and link investment risks)
Using the method described above, pick your 2 or 3 best performing competitors and do the same thing for them as much as you can. Don’t worry about how accurate your profile of each competitor is. You’re just creating benchmarks to compare your own investment costs/risks to.

Calculate risk-to-cost ratios
This should be self-evident. Let’s say you developed a 5 point risk scale, with the higher number representing safer risks. Let’s say you have identified 5 resources classes and 3 linking strategies. A risk-to-cost ratio would be the assigned risk divided by the cost of class or strategy.

If you have an average risk of 3 and an average cost of $3000 then your average ratio would be 3/3000 (or 1:1000). How many links does each resource class provide you? How many links does each strategy provide you? The lower your ratio, the better. For example, a high risk/high cost ratio might look like 1/10000 (1:10,000). A low risk/low cost ratio might look like 5/500 (1:100).

You can convert the ratios to percentages, if you like. 1:10,000 would be equivalent to .0001 and 1:100 would be equivalent to .01. You’ll never get to a value of 1 but the closer to 1 your percentage/ratio gets the better (in this model — there are other ways to measure link investment and risk).

If you obtain 100 links at a ratio of 1:10,000 and 200 links at a ratio of 1:100, which resource class or strategy do you feel offers the better return on investment? Actually, you can’t know for sure, but if you’re tracking indexed links and how fast each resource and strategy produces them — well, you can develop some pretty interesting statistics.

The bottom line here is to determine how effective, risky, and efficient your link building is (and maybe to compare it to your competitors’ performances).

I received a sad email this morning. It appears that David Castle, who has been around the SEO community a long time, has passed away after a brief struggle with an aggressive form of cancer.

He always struck me as such a classy guy, knowledgeable in SEO, and it didn’t hurt that he sometimes agreed with me on various controversial topics.

You can find some memorial threads about him around the Web, including at Warrior Forum, SEOers, and Web Pro World.

David had many online friends and admirers who respected him, appreciated his points of view, and enjoyed chatting with him from time to time. I was one of them.

My best wishes to his family and dearest friends. Your loss is greatly appreciated.

Over the past few weeks I’ve found several articles from people in the SEO industry that talk a little bit more about on-page optimization and a little bit less about off-page optimization. Even Rand Fishkin’s article from this week about on-page optimization pays some attention to matters people have long ignored (although he has yet to publish the data that show his conclusions are valid).

This gradual shift in interest back to on-page factors — which have been covered in thousands of SEO tutorials, introductory blog posts, and even a few in-depth debates — seems to signal a feeling in the SEO community that maybe not all is well in the world of links.

It feels to me like everyone is trying to make sure their on-page optimization credentials are in place just in case content optimization becomes more popular with other SEOs than link-based optimization.

I’ve always maintained that the two go hand-in-hand. After all, you only get two things to work with in this business: links and content.

Links seem to be taking it on the chin for several reasons.

First, more social media sites than ever are using “rel=’nofollow’” on their user-generated links. The SEO community is finding it more and more challenging to produce useful lists of dofollow sites that make sense.

Secondly, more bloggers are starting to use “rel=’nofollow’” in their comments, or to moderate comments, because of the abuses of the SEO community.

I’ve also noticed over the past few months a growing trend in the free article distribution business: the archives are requiring people who republish the articles to allow the embedded links to pass value (they forbid the use of “rel=’nofollow’” in their terms of service).

Who wants to link to crappy sites that no longer exist when reusing old articles?

Some link hounds must be feeling rather forlorn about now as it’s becoming more and more difficult to find good lists of resources to abuse. Of course, every now and then someone feels compelled to ruin more good resources by publishing them in a new list, but the lists keep getting smaller.

If the Web runs out of value-passing links, how will the search engines ever index all the new content being created?

There is, I am sure, a happy medium somewhere between all value-passing links and no value-passing links. We do, after all, control the links we place on our own sites. There is no need to panic just yet.

But the retreat from link gorging may produce yet another SEO myth — that links have somehow lost their power, just as SEOs have behaved for several years as if content had lost its power (ironically while sites like Ickipedia crept into more and more search results without the aid of link anchor text).

I don’t think links have lost their power — I believe that link providers have begun exercising their power. They have the power NOT to pass value and they have listened to Google explain why they should not allow indiscriminate linking to third-party sites. It may very well be that the end of superlarge sites is in sight for the SEO community.

Oh, there will be more DIGGs, other Twitters, but they will nofollow their links and they will become popular without the help of search engine optimization abuse.

Rather, I think the SEO community will have to turn back to more niche-oriented linking resources, where users are vetted in at least some minimal way. Instead of allowing anyone to blindly link out, the next wave of linking resources may only allow some people to link out (with value-passing links). We’ve already seen a primitive form of this process in blogs and forums where people earn the privilege of passing value through their links.

Think of the opportunities you have to create new resources where user trust is earned step-by-step. One possibility might be where links are submitted to community review ala DIGG or StumbleUpon and if enough people agree they are useful then the nofollow is lifted from them.

But how do you combat the inevitable link gangs that will arise? GMail accounts are easy to create and anyone can use them to create an army of sock puppets. Maybe the trick is to require people to earn the privilege of submitting links in the first place through contributions. Having to prove they are real people might be more than average sock-puppets can accomplish.

It just seems to me that the barriers being raised against link dropping are increasing faster than the barriers being raised against content creation. In the process of changing gears many people in the SEO community are starting to come out of the woodwork to affirm something that everyone has known for a long time but which they were afraid to admit because their favorite link gurus would have publicly shamed them: content is the best way to optimize for search because you can still decide whether to include value-passing links in the content.

It’s okay to remember that you still need content. You’ve been creating it all along anyway. Next year someone will come out with a great new linking resource and everyone will jump on that bandwagon no matter how brief the ride may be. But they’ll still be creating content because as everyone knows, Content Is King. It’s just not always in fashion.