175 replies to this topic

[SmellieNellie]

Hi Qwerty -

Well now the honest truth is I have only just learnt that as my website is hosted in Canada people searching UK pages only, will never find me!

As far as the logs for searched conducted world wide, no, there is no indication that anyone has searched using the town in their search. I guess what I ought to do is go back to wordtracker and see whether that comes back with any logged searches would you say?

Thanks

Smellie

[qwerty]

Yup, and you should also check to see if people are searching using something slightly less geographically-specific, like your county or region (assuming you were referring to a city before -- I don't know much about the UK).

The location of your server is also an issue... and do you have a .com or a .co.uk?

[OldWelshGuy]

Bob , Nellie is in the worse possible position, having a .org domain hosted in Canada , but she is attacking that as we speak lol

Nellie, this might help, it is the UK search check for Overture:-

UK Overture search Stats

[peter_d]

Write a quality CMS and, Voila, pretty good, detailed content that should rank well accross multiple terms.

That's right.

If you take the spirit of Ammon Johns Three Page optimisation technique, a database, and a competent developer, you can quickly generate a large site that pleases both search engines and users.

This technique is useful on sites where you can't alter the sales copy, so you optimise the structure instead. The more pages you have in the SE database, the wider your keyword net becomes.

[peter_d]

within intensely competitive markets such as IT, they all come up with a low KEI

Disregard it.

The most important piece of information I know about search marketing is this:

most search queries are unique

The English language is so rich that there are billions of possible combinations.
While there will always be common terms at the top of the hierarchy, (i.e."health"), the further down you go, obviously, the more diverse it gets (i.e. "oral cancer treatments"). Don't stop there. There might be a hundred different ways a person could search for "treatment for oral cancer". There's just two . You can still go for the common terms, however it's a matter of effort vs reward.

One common term may cost a thousand times what one hundred lesser terms cost to achieve, yet both may produce the same amount of traffic. Nine times out of ten, the later produces better qualified traffic.

So, you want many pages in order to catch the myriad of variations. One of the best techniques on how to exploit this fact was written by Brett Tabke, and it still holds true today. Search on "Successful Site in 12 Months with Google Alone".

Edited by peter_d, 04 February 2004 - 06:45 PM.

[BrianR]

Yup, and you should also check to see if people are searching using something slightly less geographically-specific, like your county or region

Smellie

As I'm just over the border from you in Chester, I'd say that it's very important to target geographic keywords: Wrexham, Wrecsam, Clwyd, North Wales, N Wales, Llangollen, Oswestry - for web design.

BrianR

[SmellieNellie]

Thanks all.

OK, so you've talked about the importance/non-importance (is that a word!!??) of KEI but, when you are looking at the number of searches vs the number of competing pages, at what point would you think, nah, that's not worth targetting.

For example: Count = 47 Competing = 0 or Count = 646 Competing = 3180

I am trying to get an idea of when to disregard an overpopulated keyword/phrase.

Brian - thanks for that, but you know what worries me, I actually put wrexham into wordtracker, and apparently nobody searches for any of my IT services in wrexham??? Hmmmm

Nellie

[DanThies]

OK, so you've talked about the importance/non-importance (is that a word!!??) of KEI but, when you are looking at the number of searches vs the number of competing pages, at what point would you think, nah, that's not worth targetting.

For example: Count = 47 Competing = 0 or Count = 646 Competing = 3180

I am trying to get an idea of when to disregard an overpopulated keyword/phrase.

I like to follow a process, the process sort of tells me what to target.

First, you have to do a thorough assessment of the site, and carry out in depth keyword research. If you do this well, you'll likely have at least a couple hundred search terms to work with, often many more.

Then we take that list, sort it by popularity, and start looking at relevance. Relevance, we measure as a percentage. What percentage of searchers, who have just used that search term, would find your site relevant?

This gives me a "weighted popularity" for all the search terms. Then I can sort my list again, and the best search terms will be at the top. I will definitely target anything that is highly relevant, so that we get the highest quality traffic.

In terms of measuring competition, you can see my initial post in this thread. You have to be realistic about your ability to build links and create content.

I have a quick tips guide online if you'd like to see how this is done. It's an extract from the guide we send with our keyword reports:
http://www.seoresear...a/quicktips.pdf (PDF, 196KB)

[SmellieNellie]

Thanks Dan - I have saved your report and will read it thoroughly later.

Nellie

[steveeyes]

Hi Dan

Since the latest google update, I have read several posts on allinanchor. I know how to use it, but I do have a question.

If you do an allinanchor on your keyword and your site is ranked high compared to a regular search which ranks low, what exactly does this mean.

One post said it means your site is being filtered on a regular search if it has a low ranking compared to an allinanchor ranking that is high.

I followed up asking if being filtered meant I done something wrong and being penalized or if filtered meant google filters certain keywords even if I did not nothing wrong. Never got an answer.

Any feedback would be appreciated

Thanks Steve

[Randy]

Hey Steve,

I'm not so sure that "filter" is the correct term, though it has been bandied about a good deal. It appears that several factors are involved, so the New Google is more of a many-faceted shift in G's algorithm. You'll see discussions about many of these factors all over the forums, especially down in the Search Engine News category.

The "filter" idea came to the forefront early on with the Florida update in mid-November because at that time it appeared that only certain highly competitive (some theorized Commercial-only) search terms had been affected. Since then, we've seen more and more terms hit by this "filter" with the Gladys update.

However if it were simply a "filter" of some sort (Keyword Density, Anchor Text, etc) being put into place, re-establishing a site's previous position would be a piece of cake. That's note the case though. For some sites/pages one set of changes will help, for other sites/pages a totally different set of changes seem to help.

So IMO it's more of a shift in how Google ranks pages, not necessarily some type of all-encompassing filter. Then again, it's still my opinion that Google is simply broken right now. I can't see any reason why so many sites/pages rank Top 5 on ever other engine, yet not be in the Top 800 on Google. That's just silly by my way of thinking.

[MUTTLEY]

Well it is comforting that I already came up number one on most of mine, the rest were number for or more. Yet I'm far from number one in my category. At least it appears as though I have a decent start !

- Ray

[memmert]

Hi there-

I'm using WordTracker for the first time and I've found that many of the words I selected in my initial keyword list have very high competition numbers. I read Dan's information on how the competition numbers in WordTracker aren't a good indicator, so I'm going to perform the steps he listed but I've got a question.l

After I've done a search in wordtracker and get back my list with count and 24 hours numbers, the words that seem to me to be worth optimizing are a count of 100 or more (my self imposed indicator of value?- when I see that perhaps a version of that word has appeared 24 times in 2 months, it doesn't seem like a very good word to optimize your page for. Am I correct in my thinking? I realize that if those words that have appeared 24 times are more targeted and therefore are more likely to drive traffic to the site that is really interested in your offering then great, but these are most likely just different versions of the same word.

The web site I'm optimizing is for a local business and the most targeted and relevent traffic will come from people within a certain geographic range (our state and the bordering states). I'm tempted to include the city names "such in such in Dallas) in my keywords to the extent that I can use copy that supports this.

Does anyone have any suggestions? Am I thinking about this incorrectly?

Thanks so much for your help!

[BrianR]

Hello memmert - not seen you posting before, so a hearty welcome to you.

Two points from my experience that might help:

1. I could be misreading you on this, but it seems that you're wanting to optimise for single keywords - please correct me if I'm wrong. But if you are, then that is difficult in highly competitive areas - you may have more success optimising for relevant two or three-word keyphrases.

2. For local businesses, geographic indicators are critical, IMO, simply because that's the way that people will search, especially in competitive categories. So I would try to optimise for geographic centres on pages where it would fit most naturally - contact us, about us, delivery service, and so on.

Hope this helps.

BrianR

[DanThies]

Take the top 100 or so search terms you get from Wordtracker to start with, and consider how relevant or targeted they are. I like to use a percentage value, just an estimate - what percentage of people using that search term are likely to find your site relevant?

If someone types "marketing" in a search engine, they could mean almost anything. If they type "search engine marketing," that's a lot more specific. If my site is just about search engine marketing, it's really only relevant for that more specific query.

Once you have those percentages, multiply the counts by the percentage. So if you have a search term with a count of 400, but it's only 10% relevant, the "weighted count" becomes 40. Compared to a 100% relevant search term with a count of 60 searches, it's no contest, right?

It's not about "how much" traffic, it's about how well you target the audience you need to reach.

If you are targeting the most relevant search terms, should you really worry about competition? Only if it's really extremely stacked against you. If you're serving a local market, it's hardly worth worrying about the global competition, you just have to make sure you have the "local" keywords (city name, neighborhood, whatever) on your site.