Four advanced SEO practices

by Michael Martinez on April 21, 2008

As I have pointed out myself, many people are disappointed in the SEO conferences because you cannot learn the “real” hardcore stuff from the presentations. SEO conferences nonetheless serve useful purposes because those presentations help people who are just learning about search optimization understand the basic principles and see just how much is actually going on in the world of search.

Even so, there comes a point where you ask: “Is that all there is? Is there nothing more?” (Star Trek fans should recognize the reference.)

If you search for “advanced seo tips and techniques” or something similar you’ll find humiliatingly simple suggestions like distributing articles and press releases, translating pages into other languages, etc. “Content is king” is not advanced SEO. That’s a concept for beginners. Here are some more advanced concepts.

Hoard content for future use – Hog your content like the plague is coming and you’ll never be able to produce new content again because all your hired help will be dying off. You can never have too much original content lying around waiting for publication. Advanced bloggers have figured this out. Many spammers have not. If you take all your content and publish it now you have nothing to work with tomorrow. RSS feeds may be cheap but they are easily filtered out.

Storing content for those lean, dry months when you don’t have anything useful to say to your visitors is a way to keep your site fresh and push out your indexable inventory at a predictable rate. It’s a trick that magazines and newspapers have used for over 100 years with their print editions. Even television series producers hold back scripts, script ideas, and “clips” shows to buy time in case of strikes, production schedule crunches, and budget cuts. Web site operators don’t think about scheduling inventory for publication down the road but I do.

Furthermore, if you hoard inventory you’re only one or two steps away from another advanced concept: a production schedule. Many an ecommerce site updates its online inventory often, and if these sites are not working with productioin schedules they should be. If you know the 15-20 days in the calendar when your company offers “storewide” sales or whatever, you can plan accordingly and update your burgeoning inventory of promotional content. It’s not enough that you employ boilerplate text and CMS templates. Think about how you can anticipate new idiomatic trends and capitalize on them in your promotional language.

If no one knows who you are, distributing your content is a quick, easy method for getting your name (and your brand) out in front of a lot of people quickly, but there is relatively little search optimization value in distributing content these days. The value-passing links don’t come as easily as they once did. People like me, who occasionally reuse distributed content in a thoughtful, editorially-reviewed process, are very, very picky about the free content we select. Most free distribution articles are so badly written, so self-promotional, and so uninteresting they serve no useful purpose except to demonstrate how clueless their authors are.

It’s better to save up a stockpile of content so that you always have something to work with. It doesn’t matter what industry you’re in, publishing content on a schedule makes a lot of sense for many reasons. You can gradually develop a new theme section without feeling pressure and rushing through the tasks. If you let your articles sit and age a little you’ll be able to come back to them with a different perspective and catch mistakes that could be critical.

Create Web sites instead of profiles – Everyone has jumped on the social media bandwagon these days but too many SEOs and would-be SEOs take the lazy, greedy approach. They slap out social media profiles for the sake of pointing links at their cheap, sleazy, schlocky content. Most social media links don’t pass nearly as much value as rumor and myth would lead us to believe. Why is that?

Because you have to receive link value in order to pass it. Just creating a profile page and dropping links to your badly written blog posts doesn’t provide you nearly as much value as creating a Web site that people think of as a useful resource. Just because your “web site” happens to be a social media profile doesn’t mean it can’t be well-conceived and useful to other people. Rather than trying to reach the top slots on DIGG and SPHINN you should be creating a resource that people look for when they visit DIGG and SPHINN.

Some social media sites provide you with more opportunities for creativity than others, but you don’t want to jump on MySpace and Facebook unless you have the marketing clout to create interest in pages that will otherwise fade into the framework of millions of other pages. Where you want to create Web sites is on social media sites that provide the opportunities that most of their users don’t take advantage of. Think about it: you’ll capture more attention by being the first user on super-social-media-site.com to create a really unique and interesting profile than by being the 10,000,000th user on MySpace to try it (and MySpace sucks anyway — everything is canned and all the pages look the same).

If you place personal value in your social media profiles, so that you craft them into high quality resources, you’ll be less likely to abuse them with spammy links. REAL search engine optimization is an engineering process, not a link building process. You’re engineering Web sites to perform specific tasks and functions, and you achieve more with useful resources than with mass-produced crap.

The time to create a specialty custom profile is BEFORE everyone else catches on to the value of specialty profiles.

Read the news – Leveraging interest in news stories is a secret to advanced search optimization that you seldom hear about. If you blog on a regular basis and you can draw a connection (even a far-fetched if reasonable one) between a new event and your content, do it. People will be searching for interesting content connected with the news event and if your blog appears in blog search engines or is part of large blog networks, the chances of capturing new audience attention skyrocket when you tie in to the news.

But you can’t just throw up promotional muck (which is what most people in the SEO industry tend to do when they learn about a new technique). Be thoughtful and passionate. If you are promoting content you should do so like it will be the defining content of your career. For example, if all people ever remember me for is huckleberries and huckleberry products I’m good with that. I love huckleberries and I enjoyed creating a site about huckleberry products, providers, and related information. I hope more people develop a taste for huckleberries (which reminds me, I need to add Tillamook Ice Cream to my list of huckleberry products).

When you create a bridge between hot topics and your content, you need to make sure the bridge is attractive, interesting, and useful to someone. Not everyone will cross the bridge but the better constructed that bridge is the more people that WILL cross it. Remember, search engine optimization is an engineering process.

You won’t get it right the first time or every time, but let the SEO method guide you. Experiment, evaluate, and adjust. There will always be more news that your content is relevant to. Just be patient and practice when it makes sense to.

Bridging content is definitely an advanced SEO tactic because it takes real skill and intuition — which you can develop over time — as well as a style that is professional and engaging. You don’t have to create link bait so much as traffic bait to achieve a great end result.

Shoot for the second position – If you see a query has gone hot and some spammy superstar is using his widget-bait to goof up the search results, don’t try to outrank him immediately. Put in just enough effort to get into the top five search results. If the leaders of the pack are doing anything snarky (and they often are) many of them will eventually be caught. You can capitalize on their greed and stupidity by positioning sites just below theirs, waiting for the inevitable to happen.

And if they manage to avoid tripping any penalties or filters, well, you’re only chasing the long-tail anyway. Getting into the top five results for a long-tail query is always a good thing. It’s cream. Take it whenever you can get it and be happy with that.

The more long-tail keywords you can gain top ten or top five position for, the more valuable your Web site becomes to the Web community in general. People will see it more often and begin associating it with either spam or value (the spam association only occurs if they click through and see a lot of garbage or advertising).

Optimizing for the long tail is not nearly as hard as it seems like. It’s a gradual, methodical process. You’re an engineer who is building out a large network of visibility. Not a large network of sites, but a large network of visibility.

That is what the long tail is all about. That is what separates an advanced SEO from the beginners.

{ 7 comments… read them below or add one }

wibbler 04.21.08 at 3:31 pm

Interesting this post.

One project of mine is a two worder “word1 word2″ and im on page 2 on google for it now (only 500,000 pages returned/found by g).

Im in a place where I have been before though now – thinking that :-
a) tweaking my page will cause G to dump it.
b) getting social network site profile links will cause G to dump it.
c) linking to it from my spammy sites will cause G to dump it.

Theres no way anyone would link to this page theres no reason for them to – its only interesting to a relatively small number of people, and therefore only a tiny number of those people will even own a site or blog on which to place a socalled “natural” link to me.

I cannot imagine that waiting for the 12 or so sites above me are all spammers – so waiting it out for them to get found out isnt an option.

Where can I go from here?

Michael Martinez 04.22.08 at 12:24 am

My immediate first choice when confronted with a difficult competitive challenge is to build a query space.

wibbler 04.22.08 at 5:32 am

Ok, I am going to have to re-read the query space post to see if I understand it correctly.

If I may – here is the post you made regarding the concept of the query space.
http://seo-theory.com/wordpress/2007/10/30/the-theorem-of-query-space-optimization/

I have re-read the post, however I think that looking at defining a query space begins here.

“We can group queries together in classes or sets of related search expressions.”

This I interpret as follows a Query Set is defined initially as being a list of potential queries such as this …
Britney Spears
Britney Spears Pics
Drunk Britney Spears
Britney Spears in Court
Britney Spears Concert
Britney Spears Comeback Tour
Britney Spears Kids
Britney Spears Music Download
And so on.

I will call my above set of queries (A query set) by the name of “Britney Set 1″

Following this, I look at the following from your post.
“Each Query set defines a Query Space that is populated by Web documents which are all relevant to each other because they are all relevant to the queries”

OK – :) – this is where it gets a bit tricky for me to explain my interpretation – as I am (you will notice) dragging in subtle references from other areas of the same post which I feel are important……

So – to define “Query Space” as a concept and indeed an almost “physical entity”, am I correct by thinking that a Query Space would be formed by taking each of the phrases in the query set “Britney Set 1″, and doing the following on each

1. Search one index for that phrase (say google).
2. Add all 1000 returned URLs to an array of URLs – during this, discard all URLs which are already present in the array. (There will be non in the array on the first pass for the first phrase)

Due to the high potential for a number of web pages being relevant for more than one of the phrases in Britney Set 1, we would assume that there would be less than (8 PHRASES x 1000) URLS in the array, as duplicate URLs should be huge. Infact, an indicator of the closeness of relationship between the phrases in our Britney Set 1 would be as small a figure to 1000 as possible of URLS retained in the list – this would mean that our Britney Set 1 phrases were all returning “generally” the same URLS when queried.

Now that I have this array of 1000+ URLs derived from my Britney Set 1 – I can call it (the array that is) Britney Space 1 – and I now therefore have a Query Space from my Set.

Further thoughts on this going on in my mind are as follows ….
I noted that you made the statement that ….
“A query space’s scope is determined by the number of related queries that produce a significantly similar number of documents.”

I see that this implies the inclusion of a given query within Britney Set 1 is determined by the number of documents being returned for that query – this therefore also implies that at some point before/during the “creation” of Britney Set 1, I must identify a “number” which is to be used as the “benchmark” when looking for quantities of documents. For this purpose, is it correct to say that the “primary” phrase total document count returned by g should be used – perhaps less a bit?
Im not too sure what is meant by “significantly similar number of documents.” as you can see.
As a further thought on the above perhaps from a slightly different angle – lets assume we wordtracked “britney” and got back 300 potential queries. One would assume that there would be less and less documents returned by g for each phrase as we stepped DOWN the list of phrases (sorted by frequency of search with the most frequent at the top). So, lets say that our chosen query set began at phrase 1 and ended at phrase 20 – lets say I searched for each of the 20 phrases to establish the document quantity value returned for each one – which ones would I want to retain in my list to create my set with – what “range of quantities” of should I be looking for?

In doing this phrase selection by number of documents present technique – the goal is to end up with URLs which ARE a Query Space – what the heck do I do next. :)

Now that my head is spinning – I will go and make a coffee.
Wibbler

wibbler 04.22.08 at 6:28 am

Another thought – to create my Query Space, infact I think I must do this instead

Each phrase in my list must be searched for, and all 1000 URLs in the serps stored in an array associated with that phrase.

This produces 8 arrays (from Britney Set 1)

Create a 9th array, which contains URLs which MUST be present in EVERY ONE of the other 8 arrays.

This 9th array now contains only URLs which are relevant to EVERY PHRASE in Britney Set 1 – it is thus my Query Space?

I think my initial thoughts had potential for creating a Query Space which contained URLs which were not relevant to EVERY phrase in Britney Set 1

Im off out for a drive. :(

Michael Martinez 04.22.08 at 9:46 am

Good job on the research but I think I can simplify it for you. A query space has two parts: the (class or group of) queries that people use to find content and the content that search engines show their searchers.

Using “Britney Spears” as an illustrative keyword, If you’re trying to attract visitors who are interested in Britney Spears content you have three options:

  1. Do nothing and hope the 12-20 sites in front of yours drop out eventually
  2. Pour more resources into promoting your Britney Spears content so that it competes even better against the sites above yours
  3. Build a new query space around “Britney Spears” for which your content naturally ranks highest

If the content already exists, you just need to focus on the other part of the equation: creating the interest (and thus the search traffic) in the queries that are part of your new query space. The major search engines have confirmed that they typically see 20-25% new queries each month. All you have to do is get the message out to people that they should be using different queries to find content about Britney Spears.

At its simplest level I’m talking about marketing. You need a marketing plan. Search engine optimization is a marketing tool, not a marketing plan.

If I were to get a license to create a Britney Spears Thingamajig, I would be promoting the Britney Spears Thingamajig every way I could so that people become curious about the Britney Spears Thingamajig and start searching for Britney Spears Thingamajig. I would create more content about the Britney Spears Thingamajig in different contexts so that people would broaden their queries about the Britney Spears Thingamajig.

wibbler 04.29.08 at 6:22 am

Hi,

Ok, I have had a long think here about your reply.

“All you have to do is get the message out to people that they should be using different queries to find content about Britney Spears.”

MM – That is a MAMMOTH TASK. Effectively – if I am reading you correctly, you are trying to almost create a brand awareness. For example you are wanting to turn the minds of people who search for “Search Engine” – into “Search Engine Google” – and even onwards perhaps to omit the entire “Search Engine” part and only ever search for the word “Google”.

Granted about a “marketing plan” – this is quite long term thinking though, and (not that I am only for the short term by any means) also it requires a certain amount of ranking for “Britney Spears” in the first instance. Unless I go and do offline marketing.

These sorts of brand creation marketing plans can cost millions of dollars and take years to land home – and indeed many are a complete disaster.

I have tried to understand what else could be meant by your post here though – because I do study here quite hard. The only other thoughts I can offer to your reply would be LSI related – and rather than trying to create *new phrases* and build on the traffic/population looking for those phrases, I was wondering whether you may have been edging around these **almost** non-related phrases and saying that their presence on a “Britney Spears” site, perhaps on a subpage of that site, would possibly boost ones ranking for the target initial phrase of “Britney Spears”?

My own research shows that the answer would appear to be NO – however there were other parameters at play which I cannot discount for the failure of my testing.

Any thoughts would be appreciated.
Cheers
Wibbler

Michael Martinez 04.29.08 at 7:41 am

You don’t have to convert the masses all at once. Reaching a new audience with a message that some of them will respond to is pretty easy. Keeping the message fresh and interesting long enough for the converts to grow in number is a bit more challenging but advertisers have been doing that for years.

Since latent semantic indexing is not being used by any of the search engines there is really no need to dwell on it. Nor would it have any impact on the growth of new query spaces (since in an LSI search environment query spaces would become larger and starting new ones would require more unique concepts).

Every snowball starts out as an ice crystal. “One drop raises the sea”. Patience is the SEO’s ally.