Cassandra: Google update algo analysis thread.

NO whining or cheering about how your site is doing in this one.

[rfgdxm1]

For those who have been here a while, this is a thread topic I have started before for earlier updates. I'm starting this thread because another member suggested such would be a good idea because the main Google update thread is cluttered with posts like "OMG, I've been dropped in the new index!" and "Yippee, I'm now #1 on a key SERP". This thread is ONLY for serious, generic discussion of changes that you are observing with the new algo in this update. As in things like "Looks to me like PR is less important this month, and anchor text of inbound links counts more.", etc. How your site is doing has no relevance here unless you can explain why you think so in terms of a general algo update.

My observation for this update: More of the same old, same old. No dramatic changes from last month. Inbound anchor text in particular seems to be very important. However, I rarely look at spammy SERPs, so those who are familiar with those would be better able to evaluate whether Google is doing better or worse on squishing spam.

[mrguy]

My observation is the algo is now favoring pages that are much higher in keyword weight. So much higher, that you can now stuff it full of keywords and actually rank higher.

It appears to be the algo used years ago.

[Chris_R]

Everything looks the same to me. One of my old sites that was apparently banned "don't know why" is now back in "don't know why". I gave up on it long ago and only ran across it by accident.

[Oaf357]

It almost seems as if older sites are being favored a little more this update.

[stuntdubl]

I saw a lot internal pages showing up as backlinks across 4 different sites. May have just been my badluck this time, but it seemed that I had nearly the same number of pages in the index.

[kevinpate]

I guess I woulda known this if I'd read more first, but google apparently really does not like:

301 redirects from index.html to index.htm
Even though there had not previously been a
problem with having both index.htm and index.html
online, I decided in March it was time to get past the lazies I had caught myself in and take the site down to only one index page instead of two options existing.
I guess of the two, I shoulda elected to keep the .html instead of .htm

a link to the index page at the bottom of other
site pages
Most site pages end with the tag of:
This [name of NFP] [2-3 word summary of page topic]
page was last revised on: [date]
The [name of NFP] was linked back to the index page on the long pages (most site pages are long as opposed to being broke into 2-4 pages to icnrease page count) mainly so there could be some form of navigation available without scrolling back up and without using a TOP link (Those are ufly to me).

Other than mild content revision and some new pages, the above were the only big change related to the index page. but with the index page totally removed from the index, it seems apparent (now) that these were not sound ideas.

Oh well, another fun life lesson.

If this is in the wrong thread, don't hesitate to delete or move.

[Oaf357]

Another interesting aspect of this update was that a lot of people (myself included) were seeing Freshbot crawl while the update was in progress.

[taxpod]

I don't see much algo tweaking this time or the last couple. Things look pretty much the same to me.

[pixel_juice]

Nothing major here. Same old same old. Fresh tags on update serps was slightly unusual.

[MetropolisRobot]

27th last update. Added more links. Changed content around.

So did everyone else. 27th this update.

No real change to algorithm except that PRs seemed to be dinged on all sites in the category. If there were big changes in the algorithm then I missed it.

[rfgdxm1]

>301 redirects from index.html to index.htm

This shouldn't be a problem if they had identical content. Just yesterday I cleaned up the awkwardness of having some of my internal pages pointing to index.htm instead of root, and corrected that. Hopefully Google realizes this is really no change at all.

[BigDave]

Not really algo related, but it looks like Google google crawled about 10% deeper in some of the listserv archives that I decided to start watching last month.

[gopi]

Only one big change that is the drop in PR across industries ...

If it happened in a single industry i will say this is due to some big directory/Hub/vortal losing PR , but as it noticed in many industries i suspect some change in the PR calculating algo.

[Chico_Loco]

My opinion is that they still rely too much on off-page criteria. Whilst anchor text is at the core of their algorithm - i think it's a little too big of a core. Whilst I still like Google (even though this update I received my first PR0 ever), I do think that Fast & Inktomi are coming up behind them faster than they'd care to acknowledge.

[canuck]

One thing I noticed is that Google may really give a lot of weight to a DMOZ listing - not in terms of PR but relevance. One of my sites finally got the Category showing in SERPs and its doing amazing things this update, especially with my homepage.

I wonder if Google's algo gives heavy weight to DMOZ's unbiases vote... perhaps works into the 'site-theme' aspects of Google's algo too.

Any others see similar things once a DMOZ listing appears in Google's SERPs?

[willybfriendly]

I can tell that there have been changes, but I have yet to determine exactly what. On a single keyword I went up 10 positions. On a two word combo I went from #8 to #3. No significnat changes in the number of backlinks, and this is the first update since a DMOZ editor moved my listing from a PR5 cat to a PR1 cat. I would have expected to drop in the listings, rather than rise.

Another of my sites with only a PR of 3 managed to get two #2 positions (not highly competitive terms, but one of the #2's is out of 145,000 and is 4 positions above a PR8 site!)

So, it looks like on page factors are having more impact than in recent updates.

<edit>Wow, I see this post elevated me to a Junior Member :-)<edit>

[NotePad]

It might be a bit early for this thread, doesnt seem like it's nearly done at all.

Don't chew me out for posting this please.

[dvduval]

GoogleGuy responded to one person in the update thread about a new site ranking really well. He said they must have had a good site stucture.

I have a new site that I didn't even think would rank. The site is basically, a navigational shell with zero content. Basically, there are text links on every page to each category (or section) and that's it. I'm already ranking for keywords and the site isn't even ready for prime time!

So I say these are important factors:
1. Link Text
2. Navigational Structure (utilizing link text)
3. Page Title

Heck this new site is nothing but...the above.

[rogue]

I agree with dvduval. I have basically the same situation, only with a very small amount of content. I think it has to do with keyword density. This site is at almost 35% density.

[Powdork]

My site moved from 24 to 4 and 39 to 5 for its most important 2 and three word phrases. I had added a few links but not enough to justify that kind of improvement IMO. It does have a higher keyword density than most of the competitors (43% for the two keyword phrase). It's not just a shell, it has about 400 words of content on the page. I would assume (totally not sure though) that of two pages with the same kw density, the one with the most words wins.

[NotePad]

powdork, can you stickymail your URL to me? I'm reaally curious to see a KW density that high. :)

[shaadi]

powdork, can you stickymail your URL to me too?

[indiabackoffice]

Powdock, Please sticky mail your url to me also.
Thanks

[Powdork]

To those I just sent my url via sticky I forgot to mention that the settings would be to uncheck meta tag keywords and meta tag description but to include the title. Also, the two word keyword phrase that I'm targeting actually came in at 32%, the one at 43% is not the target phrase. For the 3 word keyword phrase it is 44%.
I really don't place too much emphasis on these numbers as the amount of multiple word phrases is highly dependent on what is considered a phrase, especially when three letter words are omitted.

[Liane]

Once again, they kicked out more links by decreasing their PR value. In the last two updates I lost a full 1/3 of the links I used to have. I also dropped from PR5 to PR3 last month and that has remained the same.

However, this has not (substantially) changed the SERPS as it was done across the board for all sites in the category. I can only assume that Google are fine tuning PR and making it much harder to reach PR 7, 8 or 9.

It seems (to me) that they are trying to make authoritative hubs even more authoritative by way of putting a greater distance between the "everysite" with hundreds or even thousands of links and "authoritative sites" with hundreds of thousands of links. Considering the daily growth of the web, which offers sites many more opportunities for links, this makes sense to me.

PS. Thanks for this thread rfgdxm1! :)

[Edouard_H]

Looks like slightly greater weight for key word density; I notice many main and indented results have switched places, the new main having kwd for target phrases between 9 - 15%, the former between 6.6 - 8%. Just a casual observation though.

[jrobbio]

See if you can work this one out. My friend has a site and he seems to be particularly good at optimising the code with rewrite rules and navigation etc. A page has gone from nowhere (actually I don't know where it started, but he never mentioned it) to number 1 beating out a page from a PR10 root and is actually PR6. Now here's the interesting bit, this page has no outside links into the page (and only 2 internal) and is a PR4, the particular keyword weighting is light with the headings and the title being the highest percentages, but well below limits. It seems ever more important to put your keywords correctly and in important places rather than stuffing them everywhere.

[sullen]

well, to me, it looks like the algo has not changed by very much at all. In my keyword sector, the same sites are doing well as before (some with high kw density, some without very many keywords on the page and 1 in particular with no title, no metatags and only 1 instance of one of the words in the keyword phrase. Wish I knew how they did it).

Just thought I'd add that to counter all the comments about keyword-stuffing being so much more important.

Also my sites have not dropped in PR or backlinks, despite not having very many (or in fact any) links from "authoritative" sites.

[sullen]

jrobbio - have you tried checking out the link-text of the links pointing to the PR10 site? seems to me link-text has been very important in the last few updates.

[glengara]

On the link text question, using an allinanchor search for a major keyphrase a site is at 12 up 1-2 places, but has fallen 5-6 places in the serps.
Only change from our side was dropping the <H4> tag in favour of size 3 <strong>.
Seems to go against my assumptions on the importance of link text for ranking purposes.