I recently installed Google desktop search on my home computer, and have found it to be quite useful for finding all of my information quickly. In fact, I've found it so useful that I'm beginning to wonder--does the desktop search use the same type of algorithm as the traditional web search?

If it does, then I've already seen some information that could definitely be used to help my future SEO--it places pages that have been modified very recently (within the past two days) far ahead of pages that haven't been updated in a while.

I knew that update frequency is important for Google's algorithm, but I didn't think it was this important. I haven't yet seen any document on my computer overtake a document that was modified in the past 48 hours for any given keyword.

Does anyone know if this is indeed Google's traditional ranking algorithm?

I believe there may be some form of "freshness" filter - but not to the extreme you mention.

Do you have the software installed on your computer? I'd like to see if anyone who's using it has noticed a similar trend (or, to that same extent, if any Google users have noticed a likewise trend in search results there as well).

It's two different pieces of software, why do you suspect their algos are related? In the desktop search there are no such thing as backlinks. Anchor text in backlinks is the largest factor in Googles rankings.

I think Google are smarter than this.

You can test it just by looking in the results. The newest doesn't always come out first.

True, but the fact that Google weighted it so highly in their desktop search leads me to believe that they may hold it more important in their algorithms than I had previously thought. Obviously, it's not going to have the same effect on Google's main engine (as you said, due to the lack of backlinks on your PC). Thanks for pointing that out; I hadn't even thought of that beforehand.

But looking at Google's results, you will often see more frequently-updated pages near the top of the results. This also explains how "recent news" can jump so high even when it is only a day or two old. I believe that at least some of Google's algorithm has gone into their desktop search. The main draw of their search engine is an algorithm that most users find to be efficient and helpful; I don't think they would want to deviate too far from that. Perhaps they've replaced backlinks with something such as number of accesses/modifications? Just a thought...

While it is possible that they used similar variables, they would have given them different weights. Just think of it logically. What makes a document on your computer 'important?' It's the age, filename and (if it can do this) containing the text within the document. What makes a website important and recognized? A lot of backlinks to show the site is established and not just a quick fix spam site. The anchor text for the link shows what it should be ranked for. The title also tells Google it has something to do with that topic. Having the text within the webpage does the same. Freshness? It's a possibility but if they gave a lot of weight to freshness then a quick spam site could get in the top 10 without trying. I highly doubt they would use this as a 'major' factor in the Google web search, and I would be inclined to believe they even try and prevent a site instantly ranking in some way.

Originally Posted by snowblind

Freshness? It's a possibility but if they gave a lot of weight to freshness then a quick spam site could get in the top 10 without trying. I highly doubt they would use this as a 'major' factor in the Google web search, and I would be inclined to believe they even try and prevent a site instantly ranking in some way.

Quick spam sites get into the top ten by having other parterned/networked spam sites link to them, and by repeating their keywords over and over. Indeed these sites do actually get up and running quite quickly, despite Google's efforts to engineer their algorithm to prevent such spamming.

Also thinking logically, what would be more relevant to the user: an article from yesterday or an article from last year, with the two articles both discussing "new electronic devices?" Obviously, the newer article (and this is the case with most news) will provide the latest information.

Here's what I'm thinking: if Google does respond favorably to pages that are updated more frequently, than perhaps having a "timed release" database feed would be beneficial to a site's ranking? Just a thought...

The sandbox was created to slow down and stop most forms of spam. However spam does happen and this is why Google encourages you to report it.

Also thinking logically, what would be more relevant to the user: an article from yesterday or an article from last year, with the two articles both discussing "new electronic devices?" Obviously, the newer article (and this is the case with most news) will provide the latest information.

Of corse, but what would be more relevant - and article from yesterday or an article from a week ago on that topic. Neither would be any better. Now add in an amateur creating a website "new electronic devices" and really talking about what he's brought lately in the line of appliances. No one interested in new-on-the-market electronic devices would be interested in that guys fridge. So therefore an established site that has a lot of backlinks would be more important than a "fresh" site.

Also, in most topics, freshness does not create relevancy.

There may be some kind of freshness filter but it would give little weight when compared to other, more 'important' things.

Right... but, even if it isn't much of a positive weight, I believe that by simply creating a script that would cause a part of a page to be "updated" on a weekly basis (ie--having a database of articles, with the article on the homepage automatically changing to point to a new one every month) would help boost rankings at least a little bit.

If all other factors are, theoretically, the same for two given websites (backlinks, keyword density/frequency, anchor text of links, etc.), I believe the one that is more recently updated will rank a good deal higher than the other.

In fact, taking a look at MSN's new search engine will show you that one of their three sliders is used to weight the age of the page being searched...

In fact, taking a look at MSN's new search engine will show you that one of their three sliders is used to weight the age of the page being searched...

Isn't this an optional filter? In MSN's mission statement it promises to be able to get the newest news etc. I saw one of their first press releases and that's what they wanted to do. If they have an optional filter for freshness, there would be little point building it in their standard filter.

If you had a database of articles - why wouldn't you just put them on individual pages? This would give your site more pages, thus creating more 'themed backlinks' which is clearly a large part of the MSN Beta algo at the moment. A larger part than in Google, actually.

It may be considered a good idea to have some type of random generator, but I would only suggest it on small-scale. Things like a text counter, or a random quote. However you have to be aware that this would make you an easy target for someone who wanted to get you for duplicate content penalties. Google looks towards the newly updated page to ban, whether you stole it or they stole it.

I think that backlinks, with appropriate anchor text, is always going to be a major factor in Google - and what looks to be the new MSN search. While freshness could be applied on large scale, it doesn't appear to be at the moment. This is not to say it hasn't already been applied on a small scale.

"But looking at Google's results, you will often see more frequently-updated pages near the top of the results. This also explains how "recent news" can jump so high even when it is only a day or two old."

I dont believe this is true. more like, these news sites are well established and hence able to push these newly updated pages to top of results.
but again, you may be searching within a small list of search results since the news are fresh. meaning it prolly has like a few hundred returned results in google search.

2 cents worth

A good point madmonk.

Very true. Do you think it has any sort of relevancy, madmonk, or is it not a factor in the ranking algorithm? As far as I have always heard, it is a small factor... this post was just my crazy hypothesis that maybe it's not as small a factor as I had once thought.

There has been no proof that there is a freshness filter at all. However most will accept that there is, on a small basis, simply because it makes sense. Google has over 100 variables in their ranking, and crediting this to one of them isn't too hard. I'd still like to see some testing, though.

As would I. Where did you get that "100 variables" figure from (or is it just an estimate on your part)? I would like to find a partial list of variables that influence Google's ranking algorithm, if such a list exists.

Obvious ones are known to most webmasters (keyword placement, keyword frequency, backlinks, etc.), but there are sure to be others that should be taken into consideration when optimizing a site.

As would I. Where did you get that "100 variables" figure from (or is it just an estimate on your part)? I would like to find a partial list of variables that influence Google's ranking algorithm, if such a list exists.

Obvious ones are known to most webmasters (keyword placement, keyword frequency, backlinks, etc.), but there are sure to be others that should be taken into consideration when optimizing a site.

Google states it on their Webmasters Section:

A. How does Google rank pages?
1. The basics

Google's order of results is automatically determined by more than 100 factors, including our PageRank algorithm. Please check out our "Technology Overview" page for more details. Due to the nature of our business and our interest in protecting the integrity of our search results, this is the only information we make available to the public about our ranking system.

We only have small list compared to their total. What we do know however is that for on-page optimisation the title seems to be the largest factor - and the largest overall variable is anchor text.

They keep this list to themselves for obvious reasons.

Originally Posted by Snowblind

They keep this list to themselves for obvious reasons.

Of course, but I'm sure at least some die-hard search engine optimizer out there has tried compiling a partial list.

(This is when the other people who've been reading this post without replying should respond!)

Indeed, that's how we know about headings, anchor text, title, bold, etc etc.
If people had discovered other things previously unknown I think we would know about them It is pretty easy to conduct experiments though - just make sure there is not more than one variable in any testing.

Another card to throw on the table is the target attribute on links. At WebWorkshop (another forum I talk frequently at) the testing concluded that using target="_blank" gives more weight to a link. The only problem is that it's not very user friendly to open every page in a new window.

That's what these forums are for, isn't it? Testing and publishing results

I dont think there is such a thing as freshness filter , - just guessing.
but there is a fresh crawl and deep crawl from google every now and then.
so i guess thats juz about the freshness trigger you are talking about :-)

about relevency, not too sure cuz google dont release such stuffs.
but coming from a established/authority site point of view, anything that is release by them, can be easily found cuz their network of spreading the article is huge

Google algo is based on hundreds of factors. So, unless proven that this particular bit is highly important, i doubt you should actually listen to someone claims that this bit is huge.

Sure.... this is a factor in google algo.
huge or small doesnt matter cuz these articles are released by authority sites. and to get that level, you will need massive support in terms of finance, publishing and network affliates.

in other words, contents is Not King! But instead, Money.!! Lol.

I dont think there is such a thing as freshness filter , - just guessing.
but there is a fresh crawl and deep crawl from google every now and then.
so i guess thats juz about the freshness trigger you are talking about :-)

How is a fresh crawl and a deep crawl anything to do with a freshness trigger? This is completely irrelevant.