Forum Moderators: Robert Charlton & goodroi
So, if I am page 1, #1, I go to page 2, #1. Page 1, #3 gets pushed to Page 2, #3. And so on. Every one I checkd was the same. +10.
Is it just coincidence? Roll out of caffeine? Or is google trying to hint something to me?
The reason I am suspicious is because every time I've "lost" rankings in the past 6 months, the rankings I track for myself have always gone down 10 spots. Every time. I didn't recognize the pattern until now - I never thought there might BE a pattern... Since this is the 6th time in recent months, it seems to me it is too much to just be a coincidence. It really feels like a systematic penalty. And in the SERPs I am tracking, ALL of the other positions around me are static.
Anyone else with this same observation?
What do you mean by "sitewide links"?
But, no, nothing black hat going on here. We do have some minor duplicate content issues, but on a percentage basis, given the number of overall pages we have, it is in the minority.
** Anyway, to bring you up to the hour - we're back now. **
We "lost" (+10) rankings sometime last night, been gone all day long, and now back to normal. So, we were out for about 10-14 hours. This is the 6th time this has happened in the past 6 months, this is the shortest amount of time we've been out.
My focus is on the +10 - I am struck by the fact that each ranking was shoved done by exactly 10 spots. And in looking over my notes, I now see it has been like this every time we've lost ranking of late.
Are the terms all of the same 'importance/competition' i.e. are they all terms that pages nearer the root file rank for or are rankings affected for a mix of terms?
Similarly do they all drive the same amount of traffic?
Just wondering if there's different ways to organise the changes so you can deduce a pattern and work out what might be causing it.
If the index around you doesn't change then perhaps you're experiencing what has been referred to as 'traffic throttling' - these threads might help.
[webmasterworld.com...]
[webmasterworld.com...]
So, if I am page 1, #1, I go to page 2, #1. Page 1, #3 gets pushed to Page 2, #3. And so on. Every one I checkd was the same. +10.Is it just coincidence? Roll out of caffeine? Or is google trying to hint something to me?
I'll probably get heat for this, but:
How would you test your results to see if site #1 belonged at #1 if you were Google and your results were dynamic?
Suppose you know the average click through rate of #1 through #10 (or #1 through #1000) and you wanted to know if the site occupying your top spot was really the quality your algo says it is, what would you do? (Personally I would (among other things) take it out for a little bit, and see if it performed at a higher click through rate in a lower spot than the average click through rate of the site(s) normally occupying the lower spot... I bet you could learn quite a bit in a day or two with the traffic Google has.)
I personally think one of the problems often addressed here at WebmasterWorld is people always think when they drop in the rankings it must be something they did or can fix, because Google would never check to see if sites are ranked in the correct order based on simple things like comparative click-through rate, right? It must be something the site did 'wrong', otherwise, the results would never change and the site (page) they are referring to would always stay at the top where it should be, right?
What exactly did you change before each drop? (If the answer is nothing, then try to figure out who changed something and why, and you'll be closer to the answer than you'll get if you study your 'inbound links' and '3 year old 301 redirects'. My Opinion Only.)
+10 'Penalty' myth or reality...
IDK Is it really a penalty? You decide.
'traffic throttling'......simple things like comparative click-through rate, right?
Actually, i've DISPROVEN this theory until i'm blue in the face. It's absolute bullocks. (See below)
Anyone else with this same observation?
This is just one form of the yo-yo, which you will find many threads on.
If I may be "self-serving" in YOUR best interest, I would suggest searching for my MANY posts on the subject.
p.s. Saying you absolutely DO NOT buy links may be part of the problem.
Your issues sound pretty classic to me.
"Been around forever as an authority"
"Don't actively get authoritative, trusted links"
"too big to micro manage site"
You slowly, but surely, have fallen into "yo-yo" land by "strictly following Goog FUD guidelines" and other bad advice.
Start with the yo-yo threads in the FAQs and follow them back through to all the referenced links in those threads.
P.S. Yes, my Adwords traffic has at least doubled over the same period that my organic traffic has been getting hit by a yo-yo affect, and even my Google Base traffic has jumped by 50%. No sure which is cause an which is effect ; ). But there is a chance that my organic competes with my Adwords/Gooble Base traffic, and it is simply a no-brainer they will improve when my rankings decline, however, they are stable whereas my organic is up and down. The cynic in me thinks they are pushing me towards Adwords by hitting my organic - "why should he get it for free? Let's make him pay for it." - this cynical explanation certainly fits the stats.
Actually, i've DISPROVEN this theory until i'm blue in the face. It's absolute bullocks.
I'm not sure if you're understanding my point correctly, because I don't necessarily agree with the 'traffic throttling' theory, but this patent application seems to indicate they base scoring to some extent on user behavior, which is likely IMO to have to do with click-thru / click-back rate, among other things:
Query Analysis [appft1.uspto.gov]
Maybe you have completely dis-proven the use of this patent application in the results and I missed something? It's entirely possible, but in reading through the patent application to me it seems they analyze queries and generate results based in part on visitor response to those queries, yes it seems the patent application applies more directly to user specific results, which could actually explain some things...
The main underlying point of my post and overly simple example is people may be seeing some type of testing or result generation based on the preceding patent application or some variant of it, but if you have definitively disproved it's use and I missed it, then you are correct in telling people not to consider it as a possibility, and thanks for posting about it again for those of us who have taken a break from our regular readings.
Besides, I've already done it ad infinitum and the mods get cranky when i repeat myself.
you are correct in telling people not to consider it as a possibility
ANYTHING can be a "possibility"... even a good SOUNDING possibility.
But i've proven it enough to post it on a public forum and unlike many forum posters, I don't post anything unless I'm sure enough to stake my reputation on it.
(as i know quite a few people here are gunning for me)
patent application seems to indicate
lol, did you just write "patent application"?!
You should know by now, That's a sure way to make me NOT want to post any more on this topic.
I'm still waiting for someone on this forum to say
"My database of 1000s of sites indicates that..."
re: this topic
Like I said, you're making the same arguments that were made the 1st 30 times this "good sounding theory" were made.
I've already TESTED each of these arguments to my satisfaction.
Thanks for taking time away from your project to post.
the amount of certainty and condescension you post
it's annoying, isn't it? ;)
however, it beats chicken bone throwing SEO every day of the week and twice on sundays =P
it's annoying, isn't it? ;)
Nope, just plain rude more than anything, but if that's your style and how you want to be viewed, more power to ya.
however, it beats chicken bone throwing SEO every day of the week and twice on sundays =P
Well, based on my join date and post count I'm relatively new to this, which would seem to indicate I haven't been around anywhere near as long as you or studied any where near as much as you have, unless of course things aren't always as they seem and there's more to the story re my experience and perceived membership time / post count here. =)
Anyway, we should end this off topic nonsense... You say anything is possible, but have tested this situation to the extent to know that it is not possible there is some sort of 'click though' (or other type of user behavior) comparison going on in the SERPs which causes sites with no significant changes to drop unexpectedly and I thank you for your insight.
I learn something new everyday here!
there is some sort of 'click though' (or other type of user behavior) comparison going on in the SERPs which causes sites with no significant changes to drop unexpectedly
If you are being serious about this, I want you to follow this line of thought all the way to its conclusion and tell me what happens...
A. You're out of a job as SEO is useless.
B. You should hate Google more than I "supposedly do" as you have absolutely NO control over your rankings.
C. Google are complete morons as clickthrus are easier to fake than links.
D. HOW IN THE WORLD can 99% of SEOs ever confirm this?!?
E. It's a "can't prove a negative" theory because of D.)
F. It leads to Lazy SEO thinking of "Oh well, i can't control it or figure it out, so let's buy adwords"
-------------------
i happen to be in 1 of maybe 10 industries that each company in top 10 has access to everyone's else's clickthru rates, conversion rates, direct traffic, etc.
BESIDES my own extensive data, this industry gets more Google eyeballs looking at the top 10 every week than nearly every other industry.
It's one of the first ones they look at for spam control, quality of SERPs, etc.
And it's the first ones they WOULD use for clickthru/user-behavior rankings.
They don't...
but if that's your style and how you want to be viewed, more power to ya.
lol if you actually a "newbie" i wouldn't be "rude" to you ;)
A. You're out of a job as SEO is useless.
I couldn't remember the login info for this username and posted as another member for a short period of time after my return to posting here, so you probably don't know: I'm committed to going Google (and all Search Engine) Free, for your reason B.
B. You should hate Google more than I "supposedly do" as you have absolutely NO control over your rankings.
I don't hate Google... I just think they can do as they please with their results, and think what pleases them (has to legally) is what makes the most money for their shareholders, which puts webmasters in a precarious position when depending on Google (or any SE for that matter) for their income, when that income is based on continued rankings.
IMO: Anyone who thinks or believes they have 'absolute control' over their rankings is sorely mistaken... Yes, you are able to have an impact in your results through SEO techniques, but anyone who thinks they are 'invulnerable' to Google's changes, tests, etc. is IMO mistaken, which is why most major SEO companies (including the first / last name SEO who is one of the longest, strongest and most experienced online) refuse to guarantee rankings. The one I'm referring to even limits the number of sites they will work on in a niche... The last I checked they limited it to ONE.
C. Google are complete morons as click-thrus are easier to fake than links.
I didn't say the entire basis for rankings was on click-thru rate, or even that was the exact basis... I said they could be using comparative data, which is based on visitor behavior in the SERPs, and indicated it was an overly simplistic example of what they may be doing.
D. HOW IN THE WORLD can 99% of SEOs ever confirm this?!?E. It's a "can't prove a negative" theory because of D.)
You're right, to definitively prove just about anything is tough, but it makes more sense than the penalty theory when stepping back and looking at the situation logically. (See Below)
F. It leads to Lazy SEO thinking of "Oh well, i can't control it or figure it out, so let's buy adwords"
Google's Mission Accomplished?
My main point was sometimes we blame our ranking drops on a 'penalty' or 'something we did wrong', which may not even be the case and by trying to 'fix' the error, when it is merely speculative for there to be an 'error' or 'penalty' received by the page for some unknown reason, we may do more harm than good. The main question addressed in this thread is speculation re 'penalty' or 'some other factor' causing the drop in the rankings.
It was my intention to highlight the often undiscussed thought that the drop in rankings may indeed be something Google is doing to test SERPs rather than a 'penalty' per se, because I think people default to 'my ranking dropped, so I must have a penalty' and don't ever explore the option there could be other factors outside of their control involved, which leads to limited testing and knowledge, because they are so busy trying to 'fix the perceived penalty' they don't ever explore other options which could be associated with and account for the drop in the SERPs...
Let's see if it makes any sense for a 10 position drop to actually be a penalty: Page drops 10 positions in the rankings for some unknown reason when no significant changes to the page. Page returns to original position after a short period of time (1 to 7 days is indicated) with no significant changes.
Which makes more sense:
1.) There was a penalty placed on the page, which caused it to drop, then without significant changes to the page it returned to it's original position on the SERPs, which implicitly leads to the conclusion the 'penalty' was originally assessed, then for some reason (with no effort from the webmaster) 'lifted', then re-assessed, then re-lifted, and so on, repeatedly?
Personally, I would think if a penalty was placed on a page the page would be penalized, until the reason for the penalty was removed from the page.
2.) Google is testing something within a result set and by moving a page in and out of the results they can gather data which in their opinion helps them rank pages more accurately within those results, and IMO the most likely thing for them to be testing in this manner has to do with visitor behavior within the results set.
lol if you actually a "newbie" i wouldn't be "rude" to you ;)
But I am, look at my join date and post count. I'm New I Tell Ya! LOL. Thanks... I wouldn't bother arguing, uh, I mean discussing, with you either... =)
[edited by: TheMadScientist at 12:58 am (utc) on Oct. 15, 2009]
Query Analysis [appft1.uspto.gov]
That patent is about Personalized Results. If the OP is forgetting to logout when they see the high results, but is logged out when they are down a page, then that patent might be in play. Otherwise, I doubt it.
For kicks and giggles, I have a new page on a non-owned site doing the exact same yo-yo that helpnow is experiencing.
Just to prove to MYSELF(again) this theory is flawed I'm going to firmly place it in the top 3 in a week.
If i can't, i'll re-evaluate my stance, but I've done it for 2+ years now, so i'm not hopeful on entertaining click-thru/user behavior theories.
Again, if i'm "rude" about this topic, it's because it leads down a dangerous road.
I'd rather be the "bad guy" than let this insidious theory spread through SEO forums as it "sounds highly plausible" BUT throws real SEO analysis out the window.
EVERY issue then becomes "I lost rankings cause Google is doing clickthru/user behavior testing"
That patent is about Personalized Results. If the OP is forgetting to logout when they see the high results, but is logged out when they are down a page, then that patent might be in play. Otherwise, I doubt it.
Yes, I understand, but it gives us some insight into the fact they use visitor behavior to some extent, and personally, thinking the yo-yo effect is a penalty or the subject of this thread is a penalty does not make sense to me, so as a 'basic premise for more thought' I looked at that patent application and it makes more sense to me they are using some sort of 'visitor behavior' cue to in a way they think helps rank pages more accurately overall, and the best 'basis' to begin the thought I can find is that patent application, which could have been expanded on, or adjusted to conduct some sort of behavioral testing in the results... I think it gives a starting point for the idea, and some idea of how they may be doing it.
Again, if i'm "rude" about this topic, it's because it leads down a dangerous road.
I'd rather be the "bad guy" than let this insidious theory spread through SEO forums as it "sounds highly plausible" BUT throws real SEO analysis out the window.
Or, leads to more testing and information being generated regarding the issue and SEOs know when to fix a page, and when to know something Google is doing is the cause of the effect and their continued, solid, fundamental SEO efforts are better use of their time than trouble shooting a perceived penalty...
I love to debate... It expands thought, which leads to knowledge!
Or, leads to more testing and information being generated...
The History of "debate" on this topic says otherwise.
as recently as this summer
It's like debating whether there's a secret hidden base on the other side of the moon.
If one believes there is, it's quite amazing to see all the convoluted "logical sounding" reasons to prove it's true.
Usually the top 3, and sometimes the top 5, are stable. The rest move.
#4 and #6 are positions I'd rather not be in- they get moved around without much rhyme or reason.
Now, the reasons for those moves are pretty unimportant. Personally, I think its usually down to user-satisfaction testing. THIS IS SITE-INDEPENDANT, AND ALL ABOUT THE MIX OF SITES being dislayed. Other causes are down to partitioning, and the effects of reranking within your partition, or the effects of straddling two or more partitions.
Conclusion:
1) Best result:
Get into the stable zone
2) Acceptable result:
a) Win top ranking in your partition
b) Cement your position within a single partition
Method:
Aquire (beg, hire, buy) good links*. "Good" does not mean volume. Competitors are a ready source of link ideas if you can't think of any yourself. Link quality is the key, and its getting more important as Google bungles its way around semantic associations (I say bungle- have you seen the "did you mean" recently).
If any of this doesn't make sense, I too have posted reasonably comprehensively about my thoughts, which can mostly be found in the threads whitenight has referenced.
*Even Google has quietened its rhetoric on link purchasing. Bought links, exchanged links, unsought links, non-recip OUTlinks. All have their place, in moderation. Mass link buying and automated link exchanging do not. Selling links is risky.
You can see for yourself. If you don't have the time to setup and monitor a wide range of bellweather sites, use the yahoo tool and scope out the backlinks of top and nearly-top rankers in competitive industries across a range of keywords (2-word phrases work well). A lot of them are bought. A lot of them are obviously so.
The History of "debate" on this topic says otherwise.
as recently as this summer
I haven't been reading too much lately, which includes this summer, and it probably shows, but if the 'history of the debate' is true, it makes me think, 'How unfortunate for the future income of most people who are dependent on rankings and claim to be SEOs...' (If I didn't plan to try and make a site or two profitable without SE involvement or dependence I would probably be testing like my livelihood depended on it... Go figure.)
It's like debating whether there's a secret hidden base on the other side of the moon.If one believes there is, it's quite amazing to see all the convoluted "logical sounding" reasons to prove it's true.
Are you saying you don't or there really isn't or something?
Can you prove there isn't?
Where do you think I was all summer?
You should see the tinfoil they have... Shiiiiiiinnnnnny!
Personally, I think its usually down to user-satisfaction testing.
Would this be some sort of 'user-behavior based' test do you think, or do you think they are using something else to test what you call 'user-satisfaction'?
Would this be some sort of 'user-behavior based' test do you think
Yes, I do. I would consider the primary metric to be "did someone click somewhere", with absolutely no merit given to their chosen destination. Google is testing whether it serves up the right Blend of results, not what results users prefer.
Running with this theory for a minute: I think it could be reasoned a webmaster really doesn't have any control over it, unless they can SEO into the top 3 to 5, and there is really not a penalty to overcome or an error to correct, so it could be people are wasting their time 'chasing ghosts', so to speak, when they look to 'find and fix the penalty' causing the drop rather than concentrating on their SEO efforts to get them in the 'steady area'... Does this make sense?
unless they can SEO into the top 3 to 5... their SEO efforts to get them in the 'steady area'... Does this make sense?
I think shaddows and I both agree on this key aspect.
Good stuff as always, well phrased.
All of my yo-yoed rankings are 1, 2 or 3.
So, you're saying that I am really at 11 or 12 or 13, but for historical reasons I am occasionally being put back to 1, 2 or 3. But that right now, as I stand, I really belong in 11, 12, or 13?
Forgive me, I must play devil's advocate, because these are important points, with potentially long-standing ramifications for many.
If it is true, that I am a 12, but sometimes given a 2, and not a 2 who is sometimes put back to 12, then it would seem to me that I would spend the most time at my "real" position. I have "always" (last 5 years, say) been in position 1, 2 or 3, and only lately have I been shoved to 12 for short periods of time. We're talking, for a particular search term, position 2 for 2 weeks, position 12 for 2 days, then back to position 2 for another 7 days, and so on. All together, I spend the majority of my time at positon 2. The longest time I spent in 12 was about 5 days. So, you would figure that my "real" position is 2, not 12.
It is an important distinction, and an important point, because it really does go to is there something I can do, or am I simply a pawn in a larger chess game google is playing with its results? If the latter, then it suggests we are at a seam between an old algo and a new algo. If the former, then I damn well better get down to work before I permanently find myself back in #12.
If the former, then I damn well better get down to work before I permanently find myself back in #12.
And you would still have to "get to work" and fight your way upstream. Your site is hinging between page 1 and 2 for keywords that presumably are important to you. Why would you even consider waiting until the damage is fully done?
