| 8:09 am on May 28, 2010 (gmt 0)|
@tedster - on the "who got the long-tail traffic" topic (sorry - have to work out how to make a quote from a previous post)
One explanation could be that the long-tail traffic now is getting distributed on a much larger number of (perhaps smaller) sites...
| 8:45 am on May 28, 2010 (gmt 0)|
There could very well be something to that. There may have been a certain set of SEO "methods" that Google felt concentrated the long-tail traffic on too few sites (they do have a lot of statisticians on staff!) and they found a way to lessen that effect and spread the traffic around.
It's as good a theory as any I've heard - only trouble is, I don't know how to test it ;(
One key bit of knowledge could be what kind of pages lost their long-tail traffic. Has anyone who got hit seen any patterns they can report on? This thread should be a kind of think-tank, I'd say. we sure don't need the absolute definitive word - that's too much to hope for anyway.
| 9:07 am on May 28, 2010 (gmt 0)|
I mentioned a bit further back in the thread, our traffic has stabilised near its previous level, but with a lower ranking across more keywords. That may indirectly support this theory. In theory with caffeine its possible we are getting more pages returned across all the long tail keywords we rank for, but they are perhaps being ranked slightly lower than before. Therefore some space has been created for other (perhaps a greater variety of) sites up in the top 5 or so spots (we previously tended to rank 4-6 for our main keywords, whereas it now appears to be 6+). Admittedly that's a generalisation across a lot of long tail keywords, and it always struck me as odd how few of these keywords appeared any higher than position 4.
| 9:15 am on May 28, 2010 (gmt 0)|
Having read vanessas article it would be very easy to take it out of context, i dont doubt the algo update would focus on long tail terms, this is one of the main aspects of the caffiene update, they have more resource to handle such. However, the interview was a pr man and his comments were then 'interpreted' they could be interpreted in 101 ways so have to be taken with a very large pinch of salt. It is very probable that the algo change has meant a reindexing from scratch of the long tail onto the caffine system, therefor large sites are taking longer to crawl hence taking longer to reposition.
Traffic in the mean time has gone to the smaller sites which can be quickly reindexed and evaluated. Larger sites are going to take longer.
Until things settle and some reliable information can be obtained, wmt and search correlate or something definitive from the horses mouth then it is business as usual here; quality content, new links etc etc
I am going to make a prediction, sometime between 3rd and 5th june there will be another massive shift.
| 10:11 am on May 28, 2010 (gmt 0)|
|One explanation could be that the long-tail traffic now is getting distributed on a much larger number of (perhaps smaller) sites... |
This would make sense from both Google's point of view in delivering traffic to the most relevant sites (niche as opposed to generic) and why we may not have heard about anything about traffic increases as if the traffic love was spread around it may not have impacted enough for any one site to have noticed the increase. And mom and pop sites tend (broad sweep) not to hang around here to comment.
| 10:19 am on May 28, 2010 (gmt 0)|
Andylew, it would be terrific if what you said were to be true. Just hoping for the same.
|therefor large sites are taking longer to crawl hence taking longer to reposition. |
I don't have a very large website..It is a very medium sized website and still I have lost a lot of traffic (of whatever existed before). So I really don't think it's only because of the website being "large".
Anyway, when I now key-in "example" on Google for my example.com domain name, the first two results are for two interior pages - these were two of my hotly linked pages where several PR 7 and 8 (PR just for reference sake) websites had linked to. I have to click on 'Show more results from example.com' to actually see the home page.
I think this is an indication of my earlier speculation that Google now ranks webpages independently. A 10 year old site with several thousand IBLs does not matter to Google now. What matters is how many IBLs you have for that one specific page. And since long-tail pages rarely get that kind of in-links, they have suffered. I have a feeling news portals should be the beneficiaries out of this. Their articles are syndicated across the web which means they generate a lot of IBL to every single article they create.
I had not monitored my keywords before MayDay happened and so I am not able to verify this. It will be interesting to see if members here see a pattern between IBL to individual pages and corresponding ranking of these pages.
|Martin Ice Web|
| 10:47 am on May 28, 2010 (gmt 0)|
In my niche we have a distributor who feeds his customers with his database.
Almost all of my competitors who uses this database for their webshop are gone out of serps. Instead of the customers now the distributor shows up for widget searchs.
We do not use the database for our webshop but writing own descriptions. We got a cut of 50% but were not vanished.
We find heavy loads of price comparing sites in top ten and plenty site from Swiss although we are located in germany.
Strange is, that sometimes for a widget search comes up plenty of shops and search for an other widget comes up information pages mixed up with brands.
Maybe google set up a database where they store information for widget searches, like:
-> widget a - most poeple looking for stores -> show more stores
-> widget b -> most poeple love to find more information -> show information site and brands
| 10:59 am on May 28, 2010 (gmt 0)|
A quick update: I have seen something today which MAY be a breaking of the downgoing trend for my own website(s).
I suddenly seem to reappear on 1-10 positions again while these queries where gone or bashed down to the cellar by younger and crappy websites. I must say that it still isn't showing up all the time but this may be a "syncing of datacenter" thing. I also noted that my pages in WMT are increasing.
It still seems to fluctuate but in my opinion it is a good sign most of the time.
| 11:09 am on May 28, 2010 (gmt 0)|
As far as I can tell, all that this update would appear to do is encourage paid or spammy links to deep pages especially for e-commerce sites. After all, who is going to naturally link to or write a review on a product page for something like say a paperclip or a lightbulb?
Does this mean the end of e-commerce websites for anybody other than the big boys ?
It seems to me like it is time for Google to take some responsibility for it's monpoly like status with regards to search and realise what kind of impact a change like this will have on the economies of independant retailers (particluarly those with limited markup or margins ie with small budgets for adwords that can't compete with big organisations) who have relied on the relative uniqueness of products via long tail searches to eke out a living.
I believe that it's almost at a stage now where Google could bring down a whole country's economy if it implemented the right (or wrong depending how you look at it) kind of algo change.
| 11:21 am on May 28, 2010 (gmt 0)|
Couldn't agree more Steerpikegg. Reading the whole thread is depressing, this is putting some out of business. We aren't suffering that bad but sales are down 50% this week.(last week was near normal). It chops and changes so much at the moment that i can't get too depressed. Once we know where we are & i have bad week after bad week...well, it's game over, thanks Google!
| 11:41 am on May 28, 2010 (gmt 0)|
Sweden has the answer!
Sweden is the only country in wmt that is showing a consistent recovery to april levels, massive drop 1st may then recovery 13th may.
A site: search on google.se is also the only version of google that i can get what i class as a reliable answer, it is half what every other google location shows but critically directly corresponds with the clickthrough to the site. It corresponds 100% with clickthrough, wmt and all previous statistical data gathered over the years.
So the conclusion is, dont trust site: (except in sweden for us). The pages reported as listed in sitemaps must just be on the list to go live but are not yet accessible from searches.
Same conclusion as all month - google is broke! OK OK not quite as broke but doing a massive update which seems to take us back to where we started.
To get to this ive tried over 300 different google ips - same site: result. Every country version of google, same site result. Its is only sweden that seems to give me numbers that make sense.
| 11:43 am on May 28, 2010 (gmt 0)|
Andy what do we make of this? Was Sweden affected in the same way?
| 11:57 am on May 28, 2010 (gmt 0)|
|Andy what do we make of this? Was Sweden affected in the same way? |
All countries dived on the 1st, some recovered mid month then dropped back (see earlier post) however sweden has stayed at this new may 13th level ever since. It is also the only version where a site: search gives a different number to all other google varieties which also happens to correspond with all my own statistical data. We dont really sell to sweden so it has no effect on sales but it is nice to find something that adds up - if there are the number of pages listed that sweden google is reporting then it would correspond directly with the mayday drop and our new numbers.
april site: 100 pages = 100 sales. (statistically normal)
may site: 100 pages = 10 sales. MAYDAY!
sweden may site: 10 pages = 10 sales. (statistically normal)
The sweden numbers exactly fit on our historical graph for pages listed to average sales. So based on wmt reporting more pages it is just a matter of these being added to the live search and we should be back to 'normal'
Also going back to server hardware, after yesterday saying since moving to more powerful hardware we saw a massive increase in googlebot activety (30 odd distinct concurrent ips) we are now looking at over 100 distinct googlebot ips concurrently, crawing 24 hours. One thing we have noticed is that a single ip will never crawl more than 30k pages per month so more ips = more pages.
| 12:18 pm on May 28, 2010 (gmt 0)|
Google.co.uk is broken again for me. Results are all .ie, .nz and .au -- why does it keep doing that? Is this Google's understanding of geography?
| 2:17 pm on May 28, 2010 (gmt 0)|
I want to stress that in my article, the two official statements from Google are 1) it's an algorithm change impacting long tail queries; 2) it's a ranking change.
I tried to be very clear in my article that everything else I wrote was my speculation. Matt Cutts specifically said that the algorithm change was intended to surface "higher quality" sites/pages, so in the article, I speculated on what that might mean and how someone might go about figuring out how to improve their sites to meet that higher quality standard.
None of that speculation is based on inside knowledge of any kind, I just hoped it could be starting point for anyone hit by this to start some creative thinking about analyzing the situation.
It's unlikely, btw, that this is related to caffeine, since caffeine is an infrastructure change, not a ranking algorithm change.
| 2:23 pm on May 28, 2010 (gmt 0)|
Getting the same. UK filter is broken on Google UK... one piece of software I use to track rankings is showing many international sites I've never heard of for my main search term I follow.
| 2:28 pm on May 28, 2010 (gmt 0)|
|Matt Cutts specifically said that the algorithm change was intended to surface "higher quality" sites/pages |
If that's the case, then there's MANY postings at this forum that would give that effort a failing grade.
| 2:37 pm on May 28, 2010 (gmt 0)|
Hmm, well we used to sell hundreds of products each week, now we don't. Those customers obviously liked our site & products now they can't find us! FAIL! I think the silence from Google is terrible, the least they could do is stick a statement out so we know what's what. Did anyone on here experience the Florida update? What was Google's standing on that situation? I understand it cost many $$$ for US retailers?
| 2:52 pm on May 28, 2010 (gmt 0)|
Unrelated to my earlier posts I have been doing my weekly roundup for clients on positions and noticed that one client who has a monopoly in the uk and the one and only expert source of knowledge in a particular area, has completely original content, technical data and 'online shop' selling their ~700 products which cant be bought anywhere else, every product has extensive technical data and the site is updated hourly. They are the only manufacturers and duplicate content cant be found anywhere on the web. They rightly so have always been at position 1 for their main term. They have been dropped to position 13! and replaced with spam (blank wikipedia page amoungst others) on the first page.
It is completely beyond me now, google has become a joke this month, I can see many reasons why big sites, duplicate sites etc etc would be dropped but it is just nonsense for a site like this to be dropped and replaced with spam sites. Im completely baffled!
| 2:55 pm on May 28, 2010 (gmt 0)|
|Hmm, well we used to sell hundreds of products each week, now we don't. Those customers obviously liked our site & products now they can't find us! FAIL! |
Look, it's not up to Google to make your customers find you. Go buy some adwords, or send product to bloggers to rate, or undertake any other kind of marketing that doesn't depend on free traffic.
Quite probably, your potential customers are now finding a similar product, and are quite happy with the situation.
Money is not being "lost". It's being redistributed. If you're not happy, then change what you're doing. Adapt to the new reality. Launch a bunch of highly focussed, highly targetted micro-sites. See how that goes, or re-write your content so it is totally unique. Go build some links. Anything other than demand Google gives your traffic back
| 3:02 pm on May 28, 2010 (gmt 0)|
Shaddows, please tell me how our customers are buying elsewhere when we are the only people who make these products!(much like Andy's customer) Adwords? What a joke, been there and got ZERO return. Oh, and remember we used to send £££ through Google Checkout thus earning Google money, that isn't happening any more.
| 3:19 pm on May 28, 2010 (gmt 0)|
Just to add, we have been selling online since 2002, since then sales have increased steadily. You say go & get traffic, easier said than done. One week we are "normal", the next well down, even this week one day is DEAD the next is manic. It has NEVER been like this before, until we see things settle down there is nothing we can do. I can assure you money IS being lost, it can't redistribute when no one else makes something!
| 3:39 pm on May 28, 2010 (gmt 0)|
Many people (including us) make AdWords work very well. ROI is up since MAYDAY, possibly down to a combination of improved targetting and shoddy O-SERPs.
Rarely in world history does one product uniquely fill a niche. And when it does, sales are substantially above a few hundred a week.
Let's say I'm making a left-handed, rotarised, pneumatically powered elecrto-widget. I may be the only manufacturer in the world, but someone else might be making a symetric, fly-wheel, spring-loaded electro-widget.
Or, say I make nice unique classic car minitures made from seashells. Someone elsew makes ANY OTHER NOVELTY ITEM IN THE WORLD.
The point is, your customers did not know they wanted your product until they landed on your site. Now they "know" they "wanted" whatever product they got from the site they landed on.
If they did know, and went looking, then frankly AdWords is the way to go. That, or some actual, proper marketing.
I apologise for coming across harshly, you just happened to be the most recent poster on the common riff of "Google owes me the traffic they used to give me for free" which always follows major updates- including Florida, which you referenced above.
| 3:45 pm on May 28, 2010 (gmt 0)|
No problem, & sorry if i sounded off in reply! Believe me when i say what we do is unique-back in 2002 we would have customers say i looked in all the shops then i found your site! Adwords made no difference then because the search terms were so well convered and all over the main page. It may be worth a look again now. I mentioned the Florida update as i thought even Google themselves had referenced that it didn't go quite to plan? Google doesn't owe me anything but if they want us to spend on advertising with them(& send sales via GC which costs MORE as sales go DOWN!) then they need to be more transparent on issues like we have now. Why would i advertise when tomorrow i may may £10K from FREE traffic?
[edited by: ohno at 3:54 pm (utc) on May 28, 2010]
| 3:54 pm on May 28, 2010 (gmt 0)|
I think the point here though, is not that anyone expects anything for free particularly, it just that for many online only based companies without massive budgets, long tail was a way for us small fries to grab a small piece of the action from the mega-corps. The mega-corps in my sector ensure that anyone without a 60% or more markup on product is completely pushed out due to minimum bid requirements.
I think it would be a reasonable expectation with a search engine that someone searching for a left-handed, rotarised, pneumatically powered elecrto-widget would expect results for those, and not Google's interpretation of what it thinks you want and how many people are linking to it.
Also, I don't think backlink weighting works for many 'products' - most products are not interesting or exciting and therefore you would not expect people to link to them (at least without paying). Let's face it, someone in the UK searching for left-handed, rotarised, pneumatically powered elecrto-widget is almost certainly looking to buy it, not read a 2 year old blog post from somebody in Japan no matter how many people have linked to it (which is the type of things I've been seeing). Just because something has lots of links to it doesn't mean it's good or relevant.
| 4:12 pm on May 28, 2010 (gmt 0)|
One word, Monopoly. ohno I'm with you.
| 4:47 pm on May 28, 2010 (gmt 0)|
Another word: Antitrust.
From today's news:
|Federal antitrust regulators may be able to build a case against Apple over its iTunes business because the company has a dominant share of the U.S. music download market, an antitrust lawyer said today. |
Read Full Story [computerworld.com]
I understand we are comparing apples & oranges when we bring up AT&T or Microsoft or Apple, but a similar scenario does exist in each example: One company too thoroughly dominates its field. Google is in that position and yet, to the best of my knowledge, we don't even hear the gov implying that maybe the situation needs to be examined. In fact, I have never even seen it discussed outside of the webmaster community. Be that as it may, I can only hope (but do not expect) that will change as this behemoth continues to move any way it wants, without the slightest regards for the damage it renders.
| 5:14 pm on May 28, 2010 (gmt 0)|
I have ONE question: where is Caffeine?
if it's on Google.com across the world, why isn't Google saying it's live? After all they have to, given the prep that went into this and pre-2009-holiday statements.
| 5:24 pm on May 28, 2010 (gmt 0)|
Caffeine isn't live yet.
| 5:31 pm on May 28, 2010 (gmt 0)|
Hi Vanessa. Thanks for the update, many of us have read the post in regards to the update that has happened, and I think you have explained it fairly well. Perhaps you can educate the forum in terms of the update from your point of view.
| 5:31 pm on May 28, 2010 (gmt 0)|
Ah, so Matt meant after the MEMORIAL day holiday :)
Thanks for jumping in here Vanessa. Sounds like we have more churn to look forward to/dread.
[edited by: trakkerguy at 5:34 pm (utc) on May 28, 2010]