@wneennsdjj33 - yes, that could cause a problem. Links and anchor text are very sensitive areas for Google, and repeated changes can look like attempts to reverse engineer the algorithm or otherwise manipulate your rankings. Titles can also be sensitive to frequent changes - members here have reported ranking losses after making several changes in a row.

@tedster - thanks for your reply, i have one more question:
If i have a link twice on a page of my site , and one of the link has a title and the same link is on the other part of this same page but it is with another title, does this affect SERP? Is this good or bad?
Thanks in advance!

Some tests that were discussed last November at Pubcon showed that only the first link on the page gets counted. Other tests from early this year suggest that a page fragment identifier in a second URL (# - hash mark) will allow that second anchor text to be picked up.

This is the kind of factor that Google might shift regularly, so up-to-date results might be different than what was just recently the case.

thanks for reply again, then i think i'll leave it like this and hope that fixing the random title stuff might lift the penalization ...

oh , now i realize, i was stupid, cause all that about randomly changing title tags was wrong... sorry for that, cause i meant TITLE ATTRIBUTES (you know it appears when you hover on a link) so now please answer again, does randomly changing title attributes effect SERP in any way? Cause i read many articles saying that link title attributes does not have any effect on SERP...

1.) Don't believe everything you read, there's a bunch of FUD out there, so if you haven't tested something personally, then take everything with a grain of salt.

2.) The algo changes all the time and what's counted today might not be counted tomorrow or what did not count may count at some time in the future.

3.) How does it provide a good user experience to randomly rotate something like what a link goes to? As a user I would like to know what the link is about if I hover over it and if it's two things, then IMO the title should be two things, not one thing on one visit and a different thing on another visit, or even on a refresh.

4.) Something that does not necessarily have a 'positive effect' does not mean the same something could not have a 'negative effect' if it appears 'spammy' or as an attempt to influence rankings artificially.

does randomly changing title attributes effect SERP in any way?

Take it off and you'll know the answer if you don't make any other changes and your rankings improve.

Personally, I would think a SE would look for 'stability' in the results to some extent so they can determine what a site is about, and by rotating something that defines what 'action' a link will have randomly I think you send a poor stability signal, but that may be just my thought.

Like I said, remove it and you'll have the answer... If your rankings change for the better and you haven't made another change you could probably reasonably determine it had a negative effect, and if you change it and nothing happens you could probably reasonably determine it's totally discounted and if your rankings immediately decrease you could probably reasonably determine it had a positive effect.

I can't figure out why you would do it, except as an effort to influence rankings, personally, because IMO it certainly doesn't provide a better user experience to have the text defining what action clicking a link has or what the link points to rotate randomly for people. It's the same link to the same page, so it should 'do the same thing' and 'have the same end result' all the time for all visitors, IMO.

@TheMadScientist - thanks for your answer!

Well those two randomly canging title attributes, both have something to do with the link, so it is useful in a way, and it both has connection to the link... But of course i did the random change in the hope that it might help my pages rank better in google... So it might be considered "spam" . Anyway, now i changed it back. I'll post the results here, however if i am penalized, reconsideration period might even be 2 months... Hope they're fast this time :)

I personally wouldn't worry about the reconsideration request if you haven't filed it... I think something like that is more of a 'fix-it ticket' ... Something didn't work the way it should, you fixed it. If there's nothing else wrong it should recover naturally, IMO.

But of course i did the random change in the hope that it might help my pages rank better in google... So it might be considered "spam" .

I think that's a fairly reasonable conclusion... If you think about it and people were trying to 'game' your system all the time, don't you think it's reasonable to look for any signs of 'gaming' or 'artificially influencing', even within areas that might not 'boost' a ranking?

I think I certainly would look for any signal that might identify a site as trying to 'artificially influence' things, even on aspects I didn't 'give value to' for 'determination of topicality' at a given time.

yes, you're totally right, but damn google, they can't think of every little tricks...

And i didn't tell you but the behaviour of this "penalty"-like thing is the following:
I had a lot of pages scraped by google, and all ranked on the first 5 page of related google searches, and then all dissapeared, and the strange is that now my main domain is showing instead of these pages... This might be a tipical problem for a specific penalty i suppose...

The past two weeks or so my site has been all over the place. I've been ranked #5, #4, #3, #2 and now #1. I've also seen a mixture of universal results appearing on various places throughout the SERPs. Google has been erratic at best!

I've been ranked #5, #4, #3, #2 and now #1.

Aye, it's been a fun couple of weeks for us too! ;)

Got a feeling that this level of flux is going to be permanent. Probably not for super competitive terms, but those SERPS where the top 10 used to change every few weeks will probably be changing every few days now. The SERPS that used to change every few days will now change every few hours.

Maybe "Caffeine" isn't just a description of how so many of you seem to be up nights checking datacenters ... ;)

I have found that my rankings have gone up, but my traffic is gone down. Long-tailed traffic that I'm used to getting has gone way down; yet the rankings for these long-tailed terms has gone up. Have no clue what Google is up to these days.

I have found that my rankings have gone up, but my traffic is gone down.

That's probably because there is no way you could have known all the rankings of all the thousands of long tail keywords you were getting traffic from, even if only 1 click per day. Once the trickle-down of long tail keywords dries up, your overall traffic would be down because a few good ranking keywords cannot carry the day just by themselves.

I'm in the same conundrum. On multiple sites overall Google traffic is down as much as 90% (and even more on some sites). However, if you look at G Analytics, you can hardly notice a difference in rankings on most popular keywords. Granted, a few of them are gone but most within the first 20 are still there, roughly around the same positions, give or take. However, once you get past the first page of KW stats, you can just see the picture of devastation that long tail keywords had suffered. Most of 1-2 clicks/day keywords are simply gone or can be found on positions behind 200 or in some cases 300. As it turns out, they were bringing 90% of all Google traffic. I really miss my long tail keyword traffic now!

This happens because with a super low crawl rate most of long tail URLs will not be reindexed, have no cache or an old cache and drop from the index.

Welcome to the stone age ;)

Actually 1script, I have kept track of my long-tail traffic for quite some time. For one site in particular, I have 10 long-tailed (4 KW's) phrases that account for most of the traffic. The Google rankings for those particular few have risen, but the traffic for them, has all but ceased. I have checked across servers, and the ranking are consistent across the board. This is what puzzles me most.

SeoptI, what you describe, is not what I am seeing.

This happens because with a super low crawl rate most of long tail URLs will not be reindexed, have no cache or an old cache and drop from the index.

Well, that's the thing: I cannot say that the crawl rate is anything out of the ordinary. Certainly not "super low crawl rate". I do see a bit of an issue with indexing - I've been always checking the freshness of the home page's index. Although it varies on different sites, most had their homepage visited and cache updated daily. On my largest site I have seen the homepage cache date updated 2-3 times a day. Right after the traffic dropped on March 15th, the homepage cache dates have been rolled back 7 to 10 days and are no longer updating daily. They've been sort of catching up and now hover at March 18th to 25th (it's March 29th as I'm writing this).

So, I can't say anything about "quality" of Googlebot visits (as in whether they retain and process the data it gathered) but the quantity appears to be normal during and leading up to the rankings crash.

Has someone a constant appearance of Caffeine now? Perhaps in Europe? I haven't seen Caffein-Results for over a week now. Which means in summary: Caffein has been rolled-back here (again...), sites are spidered but not indexed and new links have not been influencing ranking since 3 month now.

cangoou - I concur - no caffeine sighted lately, and new links still seem to be largely ignored this year.

1script - my home page is _unfailingly_ indexed daily. But my deeper pages are not getting indexed anywhere near where they used to. When you say your crawl rate is not out of the ordinary, what metrics are you monitoring? Just your homepage? Or do you have a way to monitor the rate of crawling for your whole site? Because... it may be misleading to gauge the overall rate of crawling your whole site is getting by what is happening with just your home page...

@doughayman

I have 10 long-tailed (4 KW's) phrases that account for most of the traffic

I'm no purist when it comes to SEO related terms but I think your definition of long tail keywords differs from mine. I've no idea if mine is more correct but for the purpose of this discussion I think we should settle on one.

I call long-tail those more obscure phrases and even sometimes individual keywords that are rarely searched for. IMHO it has nothing to do with the character-length of such words or phrases. I had traffic coming in on 5 or 6 character words that were actually engineering part numbers. maybe one person on the whole Web per day searches for it but if you rank for 1000 of those obscure part numbers, you get 1000 visits (provided you're high enough in SERPs and the title makes it sound a visit would be worthwhile)

If you are saying that most of your traffic came on those 5-word phrases, it makes me think those aren't obscure phrases even though they are on the longer side. So, in this case they would be "short tail" (is there such term?). Someone correct me if I'm wrong please.

BTW, after the crash on March 15th my by far most trafficked individual phrase is a whopping 7 words long. I wish I had more of those but I no longer do. Sigh...

When you say your crawl rate is not out of the ordinary, what metrics are you monitoring? Just your homepage?

Oh no, most certainly not just the homepage. I process my logs locally on my workstation using VISITORS Web Log Analyzer [hping.org] and I'm looking at its "Googled Pages" metric. I am also paying (much less) attention to the "Adsensed Pages" metric for the other bot. Visitors has limitations in that it does not break Gbot visits by day but you get a total for the time period being processed and you get a nice list of the latest X visits. I usually look at the last 1,500 Gbot visits. Most of the time it comes for really deep pages.

When I say "deep" I mean forum-deep or sliding scale deep meaning that a page can be only 1 click away from the home page if it's the most recent post. But then, as the post ages, it moves further down the linking structure until you may need 4-5 clicks for sure to get to it.

So, in my observations crawling of those (sliding scale) deep pages has not been affected.

Is anyone else noticing the difference in the results between the caffeine results and the regular Google results is slowly getting less and less as time goes on?

I have been watching closely to see how this new caffeine was going to go and waiting for a big change when Google put these new serps on all the ips but I am seeing that the caffeine results are staying fairly steady and the regular Google is slowly and steadily changing to more closely match.

For instance when I first started watching closely on the regular serps for one keyword set I would be on #15 and on caffeine I would be on #6. After a month or so the caffeine serps are very close to the same but the regular Google that was showing the results at #15 is up to about #8. I am only watching about 10 keyword sets but it appears from my limited observations that there may not be such a big bang of sorts on April 1st or whenever but there is a possibility that the two may just slowly merge into one. Is anyone else seeing this sort of thing?

difference in the results between the caffeine results and the regular Google results is slowly getting less and less as time goes on

Yes, that is exactly what I have been seeing! Kept thinking about posting that, but didn't. It seems to me the google.com results are morphing into caffeine.

I can hardly tell them apart now. Can others see a distinct difference still?

I dont see the morph yet, I still have a client who is #6 on caff and #56 on normal google result, thats just 1 example of many keywords...

ya, I can still see the difference but it is becoming much less. Maybe when it gets close enough google will put it live?

I can see some movement 2.00 am GMT (on a night shift..), surely is not leading to April Fool's predictions, that's all what we need, and it's not even funny!

<b>1script</b>, I bet you're going to see some positive changes and revert-backs within 48 hours, probably started a couple of hours ago. It seems G* has started releasing another chunk of data and that may reveal the mystery where your sites were buried and now finally will start surfacing again, hope I am not wrong!

Oh, will someone edit the bold tags to [] instead, forgot it was a forum.

dusky: I am desperate to see something, but alas, in my sector, on an east coast proxy, and also in the north-east, I see 0 changes so far. But I pray you are right.

It's probably like a tsunami, sweeping from east to west, starting from Japan, Russia, W Europe...US in that order and deliberately climaxing on the US based sites bang on 1st April, what a Joke if that's the case, a good one for some and a bad one for others that is!

Whats the Google IP address you are on dusky?

I use .com which as noted countless times on this thread, it jump from one IP to another, so it's usually difficult to say which one has returned the search when you do one, but if I ping the domain, 74.125.39.104 is the address at the moment. The latter address however returns almost the same search results as .com and does not seem to be varied much!