|Google Traffic Recoveries? What (if anything) did you do?|
If your traffic has recovered from any of the recent Google updates it would help if you could post about your experiences in this thread. Tell us what you did if anything, when you did it. How big was your fall and recovery? It would also help if we know a little about your website - Eg - forum, blog, eCommerce, affiliate, custom or whatever.
Although these aren't my site, I read the report on WebProNews with some interest. The recovery being discussed came with no changes to the site, even though Panda first demoted it six months ago. Either that's a significant tweak to Panda, or there are strong off-page and off-site components that shifted during that period.
|Google continues to tweak the update and make additional algorithm changes, and we’re hearing more stories about recoveries from Panda. SearchMetrics recently reported that various sites it had previously reported as losing search visibility due to Panda had bounced back. |
Remember FonerBooks that we talked about earlier this year? Morris Rosenthal, who runs that site, had written a guest post for us about Panda. Now, he is talking about a different site of his making a recovery. "My IFITJAMS site, which got crushed by Panda in April, recovered with no changes in the September update, and is now getting a 3X to 4X boost from the latest Panda," he tells us
Daniweb was also mentioned although I don't know if its up or down this week I did read there were a lot of changes made.
I apologize for taking this thread in a slightly skewed direction by linking to second hand recovery reports. Kidder clearly asked for FIRST hand recovery reports - along with some detail about any changes they made.
We do have some recoveries mentioned in various threads around the forum, but it would be good to collect such report in one place for reference. So let's hear it (and let's avoid any more digressions, please.)
From my own personal experience, removing technical issues proved to be the biggest change (went through WMT's HTML Suggestions), and probably contributed to the partial recovery the most. Other than that, page speed improvements are also important, and as my website is content based, I've also added more fresh, relevant, high quality content (not necessarily more content in total, just more of the type of content that, regardless of Panda and whatnot, would attract visitors to stay longer and to come back).
It's also good to get more non Google traffic, which is a necessity these days anyway to counter whatever concoction Google cooks up next. I have a feeling Panda seems to have a less full on effect on websites that does not rely on Google traffic as much, which kind of makes sense when you think about it, as content farms and scraper sites would all have most of their traffic coming from Google, whereas a high quality website would have traffic from everywhere.
|and probably contributed to the partial recovery the most |
How much is partial?
One of my sites is pretty well back to pre-panda 2.1 in terms of traffic and serps as from 10th Oct.
Traffic was down by about 30%. Serps was down by on average 3-9 places.
What did I do? Very little as I didn't know what was wrong in the first place (no scraped or copied or duplicate content, no weak pages) so just decided to wait with this one. It's also still my best performing site (even postpanda and i didn't want to do anything that might make things worse!)I subscribe to the "guilt by association" theory with this site - that in fact it was more likely others of my sites that Goog didn't like, sites that got more heavily pandalized.
Another site I moved the hosting (part of testing my guilt by association theory) and it recovered two months later and is still ok after a further two months. I subsequently moved another site to different hosting, but after two months that hasn't made any difference.
Sorry - not suggesting any cures for Panda here - just adding my experiences to the overall picture.
We were hit in an early panda iteration and had a lot of trouble coming back. It wasn't until last week that we recovered, and the traffic graph from Google organic search looks like a capital L flipped on its back followed by a hill that continues to rise.
We tried EVERYTHING. Go through any list of stuff to try to come back from Panda and we probably did it. But when it comes to "removing low quality content" we at first chose to use follow, noindex tags because our "low quality" content was still useful to some users if they were searching for it on our site. It turns out, at least with our experience, that Google actually wants you to remove the content from the site, not just noindex it. Once we deleted the content (thousands of pages) and either let it 404 or, in some cases in which they had external links, 301 it to the closest category page, we made a comeback in the very next update.
We redesigned the site, went through days and days of filing DMCA removal requests for old, syndicated content (9 times out of 10 they'll remove it instead of digging out the contract or email that says they can use it); deleting content when the syndicating site refused to remove their version; cutting down on Adsense units and moving the remaining units below the best content; removing any content that was too similar to another piece of content (e.g. ten ways to skin a cat Vs more than one way to skin a cat); searching with a fine-toothed comb for any duplicate content problems and fixing those; continuing to build very high-quality links the entire time; increasing time spent on site, pages viewed per visit and conversion rates while reducing bounce rates... You name it, we probably did it.
We were at the end of our rope. Our last effort was going to be a TOTAL site relaunch with new content, new taxonomy, new design - but on the same domain and with the same brand since we've built an impressive, defensible, high-quality link profile over the last couple of years, not to mention our branding efforts. But if that didn't work we were going to scrap the site.
Two weeks before the launch of our new site we came back. Go figure. But the only big difference between now and the last Panda update was that we deleted the content instead of noindexing it.
The thing that I wonder is, at the intervals that they've been running these Panda updates, how can you know really what "fixed" it? And even if you recover with a new iteration, maybe that's just got the tweaks from two iterations previous?
I'm just real wary about the "deleting content" thing. Have been from the first.
Netmeg we are wary about that too. That's why it was sort of a "last resort" type of thing. I agree about not knowing exactly what brought us back, and I think Google has designed things that way for exactly that reason - among others.
For what it's worth, that was my Panda experience with one site. It may not be a scientific study or a controlled experiment, but perhaps if fifty other SEOs chime in with their success stories (please, please stop citing DannyWeb and Hubpages. There has to be other comebacks out there.) then we can develop a more substantive picture of what has really worked.
Step #1 - I removed all Google beacons including search box, analytics and +1.
Step #2 - I cut the number of adsense units in half (was only showing 2 per page, none on most).
Step #3 - I carefully reviewed webmaster tools and removed an entire directory from Google, the one Panda beat up the most.
Step #4 - I moved Pandalized articles out of the other directories and into their own, a new one just for them.
Step #5 - I updated my internal link structure so that no page interlinks with any page in that category.
I saw a moderate improvement but a subsequent "minor" iteration (as per Matt Cutts Twitter message) took that away. I refined my site to deal with a few more pages that seemed like they might be pandalized and am seeing more traffic daily, though not as much as before the first smack.
HOW do you fight a bear? You don't, you avoid it, and avoiding Google products isn't hurting, at least not yet. I'm working on a system of throttling Googlebot because prior to my Panda smack Googlebot went crazy for a day. I see no reason to allow that again. The biggest benefit to the changes is not wasting time worrying about Google issues anymore and simply building a better site with the time savings, it's refreshing.
Yesterday's stats showed Google referral traffic was down around 15% compared to the same time last year. This figure was 45% during the worst of Panda for this website, back in early June.
The "recovery" is not so much getting our old keywords back, since a lot of the major traffic getters are still in the doldrums, but we've managed to get new keywords to rank (mostly keywords for fresher content).
|There has to be other comebacks out there. Then we can develop a more substantive picture of what has really worked. |
Via Quantcast stats - answers and encyclopedia (add .com) - seem to have recovered - suite101 appeared to have recovered and then was smacked down very recently. Not sure about the accuracy, though, as I've noticed a number of sites have removed Quantcast altogether from sites over the last few months or they forbid full access.
I've noticed an increase in G traffic over the last couple of days, nothing earth shattering but we are certainly up with no obvious change in the rankings. It just looks like we are getting a bigger slice of the available searches. I've made no changes to our website, its a US based blog and the 14th was the first time its been hit.
|Google actually wants you to remove the content from the site, not just noindex it. |
I can confirm this fact too in one of my sites.
Even to remove with no 301. Just remove and copy somewhere else (if you want).