We lost about 25% of our traffic back in the October Panda update, then lost another 10% in the above the fold debacle.
Our main focus was on GWT;
1 - Cleaned up some cats and dogs with 301s about two months ago.
2 - Placed a search box next to our site's banner instead of below it (to help with Above the Fold) Text now starts at 120 pixels. We only deployed this about two week ago.
3 - Finally deployed the canonical tag, which helped clean up a lot junk in GWT about two months ago.
4 - Starting writing meta descriptions about two months ago. This is about 80% complete.
5 - Eliminated a legacy / unmaintained links section.
We thought the above was effective. Normally, out of 1,200 pages, GWT would show about 1,000 in the index. Right now it shows 1228 out of 1232.
Comparing today to the last two Saturdays, we're showing a consistent 18% increase (not 15%, not 20%). It's very consistent, very real. The site gets tens of thousands of visitors daily, so there is no doubt about the change.
We expect the above the fold change we made to take hold in about another two weeks.
@helenp and inbound - doesn't seem like you have a panda problem. panda is sitewide, so, since the keywords that you track seem to be ok, you are not affected.
Your problem, like so many others, is longtail related. This is where google is really struggling right now. The longtail pages are often well down the hierarchy of a site, where the page rank and trust juice are thin. google just doesn't yet have enough confidence in their on-page analysis to stick their neck out and let those pages into the serps. they are still relying very heavily on off page signals to be sure that the page is kosher. since these pages are often far from the high pr and trust parts of the site, there isn't much help here, so they are getting lost. its ironic, because, these deep pages often have the most valuable content on a website. meanwhile, the deep pages from amazon that are less relevant, but still packing a trust punch, are ranking. you and the google search user lose. that is where we are now, and that is why google is broken. nobody really needs google for most 2 word keywords, we can just go to wiki, ebay, youtube, amazon or walmart with our query directly. where google really can add value is by finding the site that really can match the 4 word longtail keyword - unfortunately, google has forgotten this, or, at least lost touch with this essential part of their mission. Lets all hope that they can come back to this, because we will all win.
@BillyS I refuse to change our site design for Google as its great for users and shows off the product really well. This is getting radiculous beyond belief.
@wingslevel I think you are correct!
@petehall: I hear you about the site design. Our logo was half white space, so it was easy enough to cut the image in halve and put the search box to the left. This moved up the text about 60 pixels. Completely silly stuff, but we could do it without any real harm to the look of the site.
@wingslevel - you seem correct in much of what you say, but who's to say that panda does not produce, amongst other things, a site-wide variable (to do with how to treat exact matches on a "weak" site when a "synonym" exists for sites with more trust) which is used in the ranking process? That's the beauty of a black box, we can only speculate.
Problems that manifest themselves as to do with the long tail may still be because of panda. I doubt I'll be able to prove that, and disproving it is also unlikely.
|nobody really needs google for most 2 word keywords, we can just go to wiki, ebay, youtube, amazon or walmart with our query directly. |
I've lost most of my traffic from 2 word keywords.
We don't need Google to lift BRANDS in the SERPs.
We can just go to BRAND, BRAND, ... with our query directly.
I just did a search for my "money" keyword and at the bottom is - Searches related to "keyword": 7 out of 8 are BRAND names.
If you're NOT a big BRAND you're screwed.
I've lost at least 10 positions to BRANDS, FRAME REDIRECTS, sites involved in their own DOORWAY/BLOG networks...
Honestly, the worst results that I've EVER seen.
I just did a search for a product.
The 1st 3 results were facebook pages from Amazon associates followed by 2 results from amazon.com
It's funny, but I'm not laughing.
Is it possible this rollout could have started earlier then the 23rd? I had a significant downward shift on several previously pandalized subdomains starting on March 21st, and continuing on the 22nd.
If not, is there something else significant starting on the 21st?
I think something happened on 21st. I'm in the UK, and lost about one third of hits across site. And still like it
A 500 page site of mine is still reporting 1900 pages when looking it up via site: command. I removed the 1400 page store over two months ago now and had a removal request approved.
The pages aren't in search, when you click around the number changes to 500, but it still reports 1900 at first. The data isn't fully updated yet, or at least it's not fully rolled out.
GWT is completely wrong at this point, most of the pages/links/keywords it is reporting haven't existed in two months.
Actually GWT says "Error del servidor 0, No se encuentra 3" whatever that means, I don't speak spanish?
|Error del servidor 0, No se encuentra 3 |
Click the more link for enlightenment.
Wow... I'm looking at a POSSIBLE 30% drop starting on March 23rd.
This might be something else because I am moving my blog around and possibly messed something else up.
Oh well... I didn't like paying rent anyway.
got a 20% ave per day increase for this update. :)
Planet13 lots of people are reporting that exact drop in %. Have a look at your long tail searches as I would say these are now pretty much completely useless.
It looks like one of my company's websites took a hit, but on the 21st, so apparently not the Panda refresh Mr. Cutts tweeted about.
Another of our websites seems to have gotten a boost. The weekend traffic is about level with the weekday traffic, at least. That's not normal, but I'll have to see how this week plays out.
This update may offer some hope for white hat websites that were unfairly punished by Panda due to the innumerable bugs it has. I have seen a small recovery (around 10-20%) in some of my websites unfairly hit by Panda, despite making no changes at all during last 13 months.
At the same time I see so much spam in search results it appears that Google has abandoned the search business and is getting obsessed with Google+.
Wild ride ahead.
I am seeing the same thing traffic down about 20% but my rankings on two and three word general terms are up. It must of been a long tail kill.
Can't really tell anything on client sites because most are B2B and naturally dip on the weekends anyway.
Can't really tell anything on my sites because most are seasonal and the season's coming, so traffic is up anyway.
I know, I'll go look at competitors.
Ding ding ping..
This is what I call CLEAN SWEEP Update. Google really showed who is the boss. Verticals we monitor spam has gone down significantly.
One Adsense spammer was all over the place now he is nowhere to be found other than in 1-2 places.
So far so good. My 8 years old site got back to #2 and my remedy was I DIDNT TOUCH A SINGLE THING.
I'm seeing movement... off we go again :-)
|Can't really tell anything on client sites because most are B2B and naturally dip on the weekends anyway. |
Yep, I can't see anything significant as yet plus the weather was gorgeous in the UK and a lot of other places I understand.
I have a lot of 2/3/4 keyword phrases and many of those "seem" to have risen back towards the top of the first page, whether or not Joe Public with a browser full of cookies etc is seeing the same thing is very questionable.
|This is to do the rich richer and the poor poorer :( |
It's a fact of life :)
I don't know why however I am now definitely seeing significantly less traffic to my trade directory site and especially to the ONLY trade images gallery of its type on the Net.
We'll see how overnight goes but it's not looking good right now...Google seems to have fallen out of love with Coppermine this past year, first my May 6th slashing and now, seemingly, this.
I shalln't be rebuilding it, the amount of money that earns now it would be pointless!
|They are seeing improved rankings today, up 2 or 3 positions on their best kws. |
@Tedster - is that bringing them close to pre Panda levels ? If not, where will you be targeting the next level of improve?
btw - I'm hearing some folks relying on a bit of trust improvement in the Panda recovery mix. Can anyone add or confirm that, and what they feel that may include.
Some observations on this latest March update for a bunch of sites that are in remedial phases, having previously :
- blocked all non performing pages [ saw 30% lift in Oct ]
- added content gradually to all indexable pages [ by Dec 31 ], then gradually released more pages with new content on them
- made UI improvements to all but home page and product pages [ which are currently blocked ] and being worked on.
- Nothing happened with social buzz fresh /signals. Needs to be implemented.
1. Not all sites with same design / code base have responded the same
2. Site's with highest trust [ backlinks ] have shown the earliest signs of partial, but minor improvement. No improvement on sites with few backlinks even though they have the highest TBPR and lowest number of pages.
3.Core - competitive 2 word terms - no change
4.Semi Core - medium competitive terms - all moved up a few notches in March
5.Long Tail - non existent
Traffic has hardly moved with those "semi core" improvements in  as they contribute little compared to the long tail around the core terms.
|is that bringing them close to pre Panda levels ? If not, where will you be targeting the next level of improve? |
Yes, they are at 90% to 95% of their pre-Panda traffic levels. We've stopped focusing on Panda now and are working directly on our own user signals: improving the interactions we can measure on the site with A/B testing, surveying the industry for information gaps and writing articles that fill those gaps, etc, etc.
Something people largely overlook with their websites is synonyms. To me, that is what G is having problems with, not long-tail.
The problem is that G sees something like [popular small widgets for him] and thinks it is the same thing as [famous tiny widgets]. And that throws a duplication to G as it tries to rank one of those two. I have seen one page ranking one day and a few days later the other page will take its place. And other pages with the same issue were getting absolutely no traffic from G.
If you have a page which targets [popular small widgets for him] yet your page which targets [famous tiny widgets] is what is showing up in the SERPs, that is a synonym problem which is triggering a duplication. In turn, that is what is causing what many people are calling the "long tail" problem.
The only way to correct this issue is to change the content of the page or simply combine them. Add more content on the page about the phrase you are trying to rank for and remove content from the other page which might be somehow confused for it. If the pages truly are similar, get rid of one and 301 it to the other page.
I just truly don't think there is a long-tail problem. I have not seen it. The problem is mostly synonyms and duplication.
No noticeable change for me. Last 7 days traffic about the same as the previous 7 days.
Would all this Panda update have anything to do with my searches now leading me more and more to pseudo directories [name your widget] where they've scraped other content.
I'd rather talk to the organ grinder than the monkey. Panda was never an update. Just 19 year old ex-high school kids playing around.
I'm surprised banal "twitter posts" don't rank #1.
Google was once a supreme search engine, especially around 2002 or whenever they teamed up with Yahoo.
Then the bean counters came along...
|Panda was never an update. Just 19 year old ex-high school kids playing around. |
I've read similar sentiments from others, too. IMO, it's not a good idea to be that dismissive about Panda. At this point it's a project that's been worked on for over two years, one of them development and now live.
The goal of measuring "quality" algorithmically is extremely ambitious and certainly has had some misfires. But ignoring this new direction in ranking is likely to create frustration and anger. We may not ever reverse engineer the Panda portion of the total algorithm, but we can understand the goals. There's a lot to ranking these days that few people considered to be SEO in the recent past.
|We've stopped focusing on Panda now and are working directly on our own user signals: improving the interactions we can measure on the site with A/B testing, surveying the industry for information gaps and writing articles that fill those gaps, etc, etc. |
That's a little nug of anti-Panda hot sauce in its own right Tedster.
This is what I saw on a couple of domains.
1. with lots of backlinks and good PR, not too much change
2. newish domain, few backlinks, good SEO. 50% drop which continued for a couple of says, now slowly climbing back up about 15% down.
I suspect some of these algo changes take some time to settle down as the site is reindexed under the new algo.
Of course what I've seen may be entirely random but fits with some of the above posts. I only checked in here because I thought "whoa, what's happening".
| This 91 message thread spans 4 pages: < < 91 ( 1  3 4 ) > > |