Ranking Check Depth Options

[whitemark]

Many of you monitor sites to detect rank changes for particular keywords. What is the number of search results you start to search with initially - the first 200, 500 or 1000 search results? Why?

[bragadocchio]

Hi whitemark,

About 200, but I consider that pretty far out.

I think that when you set out to get pages to rank for certain words, you can get a pretty good sense of how difficult it might be based upon the site being optimized for, and the competition that it faces.

I'm not sure that there's a lot of value in going from 999th to 853rd. Chances are that people aren't going to go that deeply in search results. So, the return on your time, your investment, isn't going to be worth the time in tracking those changes. It's probably better spent figuring out how to make the pages rank better.

[Michael_Martinez]

My SEO universe has changed since I first got into the field. Now I divide Web sites into two groups: those that make it into the search results listings (that you can actually click on) and those that do not make it. You're usually talking about 1,000 Web sites or less in the first group. You're looking at more than 8,000,000,000 in the other group.

Considering that most (80%) queries are non-commercial in nature, and that most queries (including most commercial queries) are random (meaning, the SEO search results queries are only a very small minority), I think it's amazing that search engines knock out 8,000,000,000 or more sites by one or more methodologies and come up with anywhere from 1 to 1,000 sites that may actually be relevant to a query.

So, just getting into that top 1,000 is a mark of success IF you manage to do it for any specific term.

And while it's true that the very first listing in the search results gets more click-throughs than all the others combined (generally), my referral logs have indicated an increasing trend among searches where people click deep into the search results, sometimes digging down 7,8, or even 10 pages.

Why?

Because they are looking for something specific, and whatever they are looking for, they decide one of MY pages was worth a closer inspection.

That is why good SEOs spend time sifting through their logs. If you find some obscure phrase people are using to find your pages, you can help them (and others like them) by optimizing for that expression.

So, today's results "dreck" (sites below the top 10 listings) can turn into tomorrow's gold. Neither we nor the search engines can really influence what people search for.

Hence, while I celebrate the fact that people click on some of my top 10 results, I take no joy in the fact that I have top 10 listings which produce few or no click-throughs, and I devote some time to finding new possibilities buried deep in my logs.

So, how deep do I go? I go as deep as the referer link data tells me to go (they usually indicate which page in the search results for the query on which your content was listed at the time of the click-through).

[bwelford]

Being the cut-to-the-quick kind of person I am, I set all the major search engines at the maximum allowed by their Preferences. So I look at 100 results for Google or Yahoo! and 50 results for MSN Search. If I can't get a website into such a group I still have a great deal of work to do.

I am also surprised from my website traffic log referral data how far some searchers will go through the SERPs. However I do not believe such people are at all typical. In practical terms I believe you've got to be in the top 20 and ideally in the top 5 in a search where all results are relevant.

[whitemark]

I go as deep as the referer link data tells me to go (they usually indicate which page in the search results for the query on which your content was listed at the time of the click-through).

Hi Michael. This is what had me confused in the first place. The most important data the refer link data provides is the keyword. If you decide that its an important keyword and it doesn't rank within the first 100 or 200 search results, that means you need to optimize your site pages for it. As Bill and Barry suggest, it is more resource intensive to monitor its ranking from its current position and hence not the best approach. But if you have a different reason or approach, could you please clarify?

[Robert_Charlton]

Depends on why I'm checking them....

If I'm reporting position to a client (which I do only rarely), I'll look at the top 20 results... sometimes 30 if we have a phrase we're working on. I use the default of 10 results per page, as changing the default number changes Google's clustering.

On the other hand, if I'm checking a site to see where it has potential and how it can be tuned, I use this "ie" Google display, designed for Internet Explorer, set to 100 results...

http://www.google.com/ie?q=&num=100&hl=en

It's quick, but has the disadvantage that if you have two pages ranking for the same term, the second will be "clustered" with the first. So if your first page were #4 and the second were #99, and you displayed 100 results, you'd show in the #4 and #5 positions.

To determine the position of the second page, you can change the number (num=xxx) in your address bar and click "Go" to quickly narrow down where the second site is. In the case above, if you reduced the number to num=90, your second page would drop out, telling you that it was somewhere between 90 and 100.

I also use this "ie" display when I'm diagnosing new client sites, to see if they rank for anything helpful.

[AbleReach]

I usually search with 100 results to a page at a time, then let firefox's find in page function tell me if the domain of the site is present. It's pretty quick, except Google doesn't number results. Yahoo does. Manual counting is what eats up the time.

With http://www.google.com/ie?q=&num=100&hl=en open in a second window, I can run the same search, get numbered results, then use firefox to look for how the page is titled in the search results.

Pretty slick combo, and doesn't take much time.

If I find myself doing this more than once a month, I chalk it up to obsessive compulsive procrastination. Anything over result 10 or 20 won't make much difference. However, curiosity is a powerful motivator. If a previously invisible (sandboxed?) site appears in the 200-300 range, it is encouraging, if not yet productive.


Elizabeth

[bwelford]

One thing you may find useful, ablereach, are bookmarklets. These are small amounts of code that are loaded up in to your favourites list. So in my favourites list I have Google #s and MSN #s. When I have a Google SERP on my screen and I want to add the numbering, I just click on the Google #s favourite and the numbers are added to what I'm seeing. Here's one prepared by ILoveJackDaniels.

Numbered Google results
http://www.ilovejackdaniels.com/resources/...-google-results

I wanted to have the same feature for MSN Search and that fine fellow, ILJD, produced the following at the drop of a hat.

Numbered MSN Search results
http://www.ilovejackdaniels.com/resources/...red-msn-results

Try them both, ablereach, you'll find they're incredible productivity boosters, particularly using <CONTROL>F to find the particular web page you're looking for.

I've always been delighted that Yahoo! does number the entries in its SERPs. Why Google and MSN Search can't adopt a 'customer-centric' view and give us what most of us want, I cannot fathom. I think they're trying to say, "Don't sweat the exact ranking, they're all relevant." Well if they're all relevant, perhaps none of them are relevant. That may be Google's current problem.

[Robert_Charlton]

Barry - Thank you for that one....

[whitemark]

That's a great bookmarklet. Thanks Barry (and ILJD).

I use the default of 10 results per page, as changing the default number changes Google's clustering.

That's something I always keep forgetting.

[AbleReach]

Yet another gem!

Thanks ILJD and Barry.


Elizabeth

[Michael_Martinez]

Michael Martinez wrote:  
I go as deep as the referer link data tells me to go (they usually indicate which page in the search results for the query on which your content was listed at the time of the click-through).

Hi Michael. This is what had me confused in the first place. The most important data the refer link data provides is the keyword...

If all you are concerned about are the keywords, you are missing out on a lot of useful information. Referer data will tell you a great deal about where your traffic is coming from.

...If you decide that its an important keyword and it doesn't rank within the first 100 or 200 search results, that means you need to optimize your site pages for it...

Not necessarily. People often find my sites through variations on the phrases I am targeting. It would make no sense to target a variation that generates 10 searches a day if I am already optimized for a search that generates 1,000 searches a day.

Just because I find their keywords in my logs doesn't mean I'm going to drop everything and re-optimize.

...As Bill and Barry suggest, it is more resource intensive to monitor its ranking from its current position and hence not the best approach. But if you have a different reason or approach, could you please clarify?

I search the referer logs to find out what new, unique phrases people are using to search for. I then do some research on those phrases to see if they are worth optimizing for. If I cannot find evidence of substantial searching for those phrases, I usually don't do anything.

The referral data also gives me an idea of how well my rankings stay consistent. If a lot of referrals come from the first page of search results, then I feel satisfied there is little to no fluctuation in the search results.

Referral data will also tell me which search services are more likely to be used by people seeking the kind of content I provide. I have good listings with many search services, but relatively few of them send me traffic. Occasionally, I'll make some adjustments in my content to try to entice more traffic from a smaller service, but that is normally experimental in nature.

[AbleReach]

I search the referer logs to find out what new, unique phrases people are using to search for. I then do some research on those phrases to see if they are worth optimizing for. If I cannot find evidence of substantial searching for those phrases, I usually don't do anything.

And don't forget to keep a record of what search terms you have used, when, or you'll be looking at your own keywords! LOL...

Warning - I am about to preach to the choir.

The most fun I've had optimizing has come from FAQ and testimonials. A business owner's emails can be a goldmine. Sometimes the owner is so close to their topic that they don't think to mention industry specific keywords. They often use more general terms like "quality." Their fans, though, tend to use or infer specific terms out of hand. It's amazing to me when business owners are slow to see the benefit of these words - maybe that's why I want to do what I am doing? :-)

Example from Google
wax carving - 187,000
jewelry waxes - 104,000
vacuum casting - 807,000
vacuum casting, jewelry waxes - 29,300
custom jewelry waxes, vacuum cast - 20,900
vacuum casting, custom cad cam jewelry waxes - 548

548 is not a big pool, but it is a specific search that those who know what they want may run. These words can be interwoven naturally into text that would also come up on a more general search.

The business is one person with occasional help. She does custom wax carving. It didn't occur to her that writing a page about the process using industry specific words would be helpful. People who are in the industry would already know the info, so why tell them what they already know?

I got more terms to work with by asking specifically what happy customers are saying. The result will be a little bit of education for the curious wanderer, and an easier to find site for those who know the industry specific words and want services like hers.

My theory is that this will bring targeted traffic and good ROI. I think that knowing her site is built around keywords from testimonials will give her a personal boost, too.


Elizabeth

[bragadocchio]

I got more terms to work with by asking specifically what happy customers are saying. The result will be a little bit of education for the curious wanderer, and an easier to find site for those who know the industry specific words and want services like hers.

It's a very good idea. Sort of like having a lawyer write multiple pages about what to look when hiring an attorney, or a plumber explaining some of the basics of what plumbing entails on their pages.

Emails are just a great place to find words that people will use to find your site. They are also a good place to find out the words that your customers will use when they are referring to something that you might use technical jargon or industry slang to refer to.

Looking through log files, emails, and customer correspondence are some good places to find words that your targeted audience will use. It's a slightly different topic than the post which started this thread, but it is interesting. The words you find in your log files may not always be ones that the site does well for in the search engines, and someone may have drilled down through quite a few listings befor clicking on that link. But they are worth considering, and it might be worth seeing how far down in the rankings they are.

[AbleReach]

Yeah, I did pull it OT.
To me, the distinction between keyword rank and keyword value is artificial. Rank is nothing without the value. Fuzzy logic, maybe?

For the specific keywords I mentioned, I did not know that an acid test of a highly detailed custom cad cam wax carving (her keywords) is how clean of a result it produces with vacuum casting (keywords from testimonial.) I had researched jewelry waxes and come up with "lost wax," a more general term that she confirmed was appropriate. Someone who has had a casting fail under vacuum casting wouldn't look for "lost wax," they'd look for something to do with vacuum casting of jewelry waxes. These motivated and targeted searchers would not drill down from "lost wax." Even if they don't know how to search effectively, they'd still want to try for some combo of lost wax and custom wax carving with vacuum casting.

Hitting the top of a term with over a million results is a coup, perhaps bringing in researchers who are at an early stage of a decision and will hopefully bookmark, remember, and pass on word of mouth info. Hitting the top of a specific term is another story - like the difference between - laptop - and - IBM ThinkPad with DVD/ROM.

Here I am OT again. I think I'm done now. :-)


Elizabeth

[whitemark]

Email - now why didn't I think of that? That's a great tip. An idea to expand on that - join mailing lists of related subjects to get more terms.

[AbleReach]

I've done quick googles for industry specific forums to see what industry specific vocabulary and issues they use, especially when naming threads. Will those keywords give targeted results? Which general kewords will target the search? Will the targeted terms appear at a reachable level in a general search? Will asking the client to explain the term get me more to go on?

More mailing lists and I would surely drown! :-)


Elizabeth

[bwelford]

Another way of keeping your mind open to other keywords is to look at the keywords meta tags on competitive web sites. That may indicate that someone else looking at the same market place thinks some other concepts are worth thinking about.