Welcome to the Q&A Forum

Kent - Just guessing here...  IF you had to change your settings, did you clear your cache? Might be worth a try to see if that solves it.  Good luck.

Joshua - I moderate all of the comments to my blog before and remove any links before they go live.   I find this effort worthwhile because the comments seem to increase user engagement (time on page) on the articles that have extensive comments.  These pages seem to become reliable traffic generators and seem to withstand the algo changes.

Nor would it help local search.  This could be a major factor in deciding where to purchase heavy printed materials.

I found Marie's comments very interesting.  Unfortunately, that doesn't seem to be my problem.

I agree with Marie's advice not to do anything to respond to a perceived penalty.  Doesn't look like your site is "penalized" but the swings are extreme.

My EMD site dropped from 12 (on last Saturday's MOZ report) to 41 (last night) to 32 a few minute ago for the EMD keyword.  The site has also been affected by Panda (April & again recently) but, as Elia's response suggests, G is probably adjusting the EMD values.

My Moz points seem to increase immediately by a couple of points when I post a question.   I know that Moz awards a point for the question so I assume the "extra" 2 or 3 points come from some sort of caching issue which updates when Moz awards "automatic" points for something.  Just an observation if that helps your techs who are looking into this issue.

Darin & CMC-SD.   Thanks for your comments.  I am highly experienced in my field.  Very few guest blogging opportunities are available in the properties that occupy the SERPs I'm competing in.  Smaller indy blogs (whose owners I've known for years) probably offer realistic opportunities for guest blogging. But I don't want to convert their customers to mine. (The product should be inherently local but there are national providers.)

My concern is how to counter the "authority" of the copywriters.  I'm concerned that G will look at the authority, page rank and engagement (bounce rate & time on page) of larger sites and reward the copywriter with authority status based more upon the platform they publish on than any real authority on the matters they write.

Although G could begin to compare the number of fields an author publishes in and add an educational history and licensing status to G+ pages, I suspect that G wouldn't bother to go through any extra steps (costs).  Perhaps I'm overly cynical, but selling Adwords is G's business and search results just need to be "good enough" without being too darn good.

Again, thanks to both of you.

To "Guest Blog" or "Ghost Blog"?

I've been wondering which would be better given G's "authorship" tracking program.

"Onreact.Com" indirectly raised this issue in a recent blog post "Google Authorship Markup Disadvantages Everybody Ignores" as :

"Google might dismiss your guest articles. Your great guest blogging campaign on dozens of other blogs might fail because Google will count the links all as one as the same author has written all the posts and linked to himself. So maybe the links won't count at all."

Assuming all other things are equal, would you use "Guest Author" with G Authorship attribution (if allowed) or just ghost the article and include an in-text link without attribution to you as the author?

Eliathah - I agree.  Blog commenting is extremely inefficient use of time. I wonder how many of the people who promote blog commenting as a method to obtain traffic would pay someone else to put "meaningful comments" on relevant (no-follow) blogs.  I wouldn't and I doubt many people would.

Dexm10 - The substantive text on the sample page consists of 47 words.    In the U.S. 10 on similar legal topics, I doubt 470 words would work on a legal site nonetheless a page discussing such a common subject.

I noticed 10+ sub-topics on the sidebar.  My guess is that you'll need to combine the child articles into the parent and offer much deeper content to be competitive.  It will probably require a substantial investment of time by people who know the subject areas.

Ryan - Thanks for the quick reply.   Soooo, I am overly concerned about strangers who own "spammy" sites taking the widget and putting it site wide on their spammy mortgage sites ?   After what some of the developers went through, I'm concerned about Penguin or its future mutations.

Marie & All - Excellent Discussion.  I've been very concerned about site wide use of widgets and inbound links from penalized sites.  I've been considering developing widgets and licensing them out to particular sites with the restrictions that :  the widget appear only on one page (such as a blog post).  Since the underlying data would require periodic  updates, I could build in an "out of date" statement in case someone hijacks it to a spammy site or an authorized user doesn't listen and installs it site wide.   I view this implementation of widgets as more analogous to guest blogging than developer's site wide footer links.  Providing people I've had contact with a plug in for their specific locales should result in links without much asking.  So long as the anchor text is selected by the site owners (who are even encouraged to use the URL if they ask), I view this as less risky than the web developer's site wide footer links.   Am I still missing something important  / risky?   Thoughts ?

Panda 3.5 and Penguin as it turned out.  You and Miriam were spot on.

THX to both of you.

Wayne - I share your frustration about competing with transparent networks that have few legitimate links and appear to be sustained by continuing growth in the network.  I ran Omega through OSE though and found "tons" of legitimate links from authorative sites (edu sites from major universities) which were NOT hacked.  OSE shows IBL's from 1,241 different websites.   It will require a lot of time and resources to compete head on with the omega site.  Without comparing the competing site or knowing the key phrases, my guess is that the Omega site has enough juice coming in from authorative sites and has not reached  tipping point for manipulated (money) keywords.

Alan : Thanks for sharing your experience in such detail.

I'm asking people who have recovered from Panda to share what criteria they used - especially on sites that are not large scale ecommerce sites.

Blog site hit by Panda 3.5.   Blog has approximately 250 posts.  Some of the posts are the most thorough on the subject and regained traffic despite a Penguin mauling a few days after the Panda attack.  (The site has probably regained 80% of the traffic it lost since Penguin hit without any link removal or link building, and minimal new content.)

Bounce rate is 80% and average time on page is 2:00 min.  (Even my most productive pages tend to have very high bounce rates BUT those pages maintain time on page in the 4 to 12 minute range.)

The Panda discussions I've read on these boards seem to focus on e-commerce sites with extremely thin content.  I assume that Google views much of my content as "thin" too.  But, my site seems to need a pruning instead of just combiining the blue model, white model, red model, and white model all on one page like most of the ecommerce sites we've discussed.

So, I'm asking people who have recovered from Panda to share what criteria they used to decide whether to combine a page, prune a page, etc.

After I combine any series articles to one long post (driving the time on page to nice levels), I plan to prune the remaining pages that have poor time on page and/or bounce rates.  Regardless of the analytics, I plan to keep the "thin" pages that are essential for readers to understand the subject matter of the blog. (I'll work on flushing out the content or producing videos for those pages.)

How deep should I prune on the first cut?  5% ?   10% ?  Even more ?   Should I focus on the pages with the worst bounce rates, the worst time on page, or try some of both?

If I post unique and informative video content (hosted on site using Wistia), what I should I expect for a range of the decrease in bounce rate ?

Thanks for reading this long post.

The links from low authority, unrelated sites are not likely to help much.  The anchor text (all keyword rich and targeting a couple of phrases) is scary.  Warn the client about the Penguin mauling you expect when the client reaches the tipping point.

For whatever it's worth, have a site which was also mauled by Panda & Penquin in April, 2012.  Probably lost around 30% to Panda and Penguin knocked it down to around 25% of the pre-mauling traffic.   The Penguin mauling is algo based and is probably caused by anchor text in manually written comments on relevant blogs or pages.  Without doing anything (as in taking an extended vacation from blogging, linking, etc.) traffic is now at 50% of pre-mauling (probably half of the Penguin effect is gone).  Seems to benefit from each Penguin update by 10% - 20% per month.  I am curious how this compares to "link removal" efforts.

Local - My site was severely hit by Panda 3.5 and Penguin 1.0.   Bing results held steady.  Eventually a slight decline in Bing but I attribute it to the loss of the FB Likes that resulted from my taking a break in the "community" (think "Free Beer") and taking a break from developing IBL's, new postings, etc.

As a Penguin victim, I believe what you're recommending is not enough to be safe.   I was even penalized for using my name as the anchor text.  Some of the sites where I used my name might have been unrelated and a few might have been moderated too loosely but the high degree of danger greatly outweighs any benefit.  Whatever you do, do NOT point the links at your homepage!  Other people seem to believe there is a benefit but it's probably too marginal to be worth the time required and the risks.

Penguin attacks rankings of the pages which ranked for searches where the links that caused the penalty were pointed.  So, you could have a home page that is penalized (b/c the offending links pointed to the home page and the offending links contain high traffic terms) and some, all or none of the individual pages in the blog.  The wild (temporary) fluctuations in the SERP results might be Panda related.

Carl - A couple of peple have questioned whether their listings in local caused declines in non-local results.

You mentioned that the problem you described coincided with a Penquin update.    Penguin appears to be penalize the keywords contained in the offending links as well as set a sitewide ceiling on organic referrals per day.

Is there a reasonable chance the site is penalized by Penguinn?  If so, are the offending keywords the non-local or local variety?   You've probably done so, but I'd inspect the IBL's carefully before I'd delete any reputable links.

You probably omitted the "www" prefix.  (Without the prefix, OSE says DA = 1).

I searched first with the "www" and came up with the same numbers as Dana.

Francisco raises some excellent points.  In evaluating the risk of the link being bad -- I'd avoid sites that are filled primarily with  "guest posts" or have much (perhaps no more than 10 %) "off topic" posts and also consider how close (or far) the topics are "off topic" on the blog you're considering contributing too.   Many websites feature almost entirely guest blogged stuff but remain consistently close in the topic.   Most that accept lots of guest contributions seem to be no better than the old BMR sites (i.e., lots of off topic stuff).   I expect the day is not very far away when only the strongest sites which consist primarily of guest posts pass along much juice.   My $. 02

@ Bizzer -  I don't disagree with what you're saying.  The issue is more complex and isolating one factor (even a major factor such as duplicate content issues) is often very difficult to do if you are comparing small sites with very large ones such as Mashable of HuffPo.  Mr. Cutts has avoided answering whether non-analytics data about time on site is a ranking factor.  I believe it is.  Many other factors favor larger "high authority" sites.  Even if you select better material and make more useful editorial comments about it (as evidenced by better time on site), G is going to favor the larger sites.

@ Marie : I agree with you about much of what I'm observing about guest blogging.  I get neg'd whenever I say so.  Many people must disagree.  But G appears to be culling the heard of sites.  (I expect this trend to continue and the price of entry into the game to continue to steadily increase). The sites that post a couple of articles a day utilizing entirely guest authored blogs who are likely to be devalued substantially in the next 1 - 2 years regardless of their DA/PA or psuedo-PR.

@ SEO Bunny - It's increasingly coming down to who you know.   I'd start by approaching my friends within my industry who also have blogs.   Although the stats of many of these sites are quite low (I'm not a prof. SEO), some of them may improve over time and the contextual links will be HIGHLY relevant to both sites because both of the sites concern the same fairly narrow topic. In a sense, these links are "harder" to get.  There will be few OBL's to brand sites because of the level of trust / personal relationship required.

Ryan - Thanks for taking the time to discuss Penguin recovery.  Your initial thread and follow-up Moz article were very informative.   On a lighter note, I moderate all of the comments to my blog.   I almost fell out of my chair laughing when an automated spamer targeted the keywords "click here".  LOLZ.  I suppose they were going to use a bot to attempt to spam a more "natural" link profile.  Only one comment like it I've seen and I was tempted to award a small amount of creativity points although their clients are probably being sold a terrible bill of goods about how to acquire a "natural" link profile.

Thanks.  I've watched the video before but it's worth reviewing.  Still seems a bit strange that someone can violate terms of service which G never bothered to enforce for years and get slammed with "Double Secret Probatiion" while a malicious site can clean up and eventually get the penalty lifted.  No doubt a malicious site manual penalty should result in a long time in the penalty box but at least it's obvious what to fix.  There doesn't seem to be a reliable consensus or even many case studies on garden variety Penguin recoveries yet.  Not knowing what Dean Wormer wants me to change is irritating.

InHouseSEO - It's not an e-commerce site. (It's a blog with a couple of hundred posts many of which need pruning but many of which are high informative and written by someone with substantial experience in the subject.)

Sounds like you're telling me the best gamble is put in the work on this blog to try to grow the legit links so that the bad ones dip below the "tipping point" which prompts the Penguin attack.  Have you had success with this tactic?

The home page appears to be penalized b/c of keyword rich text from relevant blog comments on mostly relevant blogs/pages. (It's also quite possible it's just a rather severe devaluation 30 or so spots in the SERPs for the EMD keyword).  Other pages are hit or miss but the stronger pages (high bounce but very high times on pages) are beginning to return to some of their former strength (probably 50% of peak traffic).

Site traffic declined just before the 25th (the date that is associated with  Panda 3.5) and resulted in a 20% hit.  After Panda 3.5, the G traffic dove steadily (which I assume is Penguin added to the mix).  Traffic is now off by around 2/3 without excluding the Bing traffic.  (Have probably seen 15 -20% improvement recently with no new posts and only added one authorative directory link (Nat'l Trade Assoc. picked up the blog).

I just reread all of the comments in the thread you linked to.  (Never received a warning in WMT so I assume the penalty is algo.)

Reading your comments, it sounds like you recomment attempting to remove any blog comments that I created. (I don't expect much success based on what people are sharing.

If  my pet Penquin is algorhythmic and isn't scheduled to lift anytime in the next several months, should I try to guest blog my way out of the penalty?  (Assume I have access to decent releveant indy blogs that are low authority but extremely legit.)

Thanks for the reminder to re-read the thread with you and Egol.

InHouseSEO - this is a GREAT question.   I wish there were more discussion of realistic case studies like this one rather than so much "focus" on negative SEO and a handful of high authority sites that were probably hit by mistake.

The consensus seems to be that you can file for lifting a penalty IF you can show you removed bad links AND document the efforts you made to remove the bad links that remain despite your efforts.

Matt Cutts appears to say you're more screwed if the penalty is algorhythmic.  Huh?  Buy BMR links, remove them and escape the penalty G imposed on your site for 50  -100 presumably manual and relevant blog comments?  Gimmee a break!

The 50 -  100 blog comments are probably going to be the worst of the lot to attempt to remove.  Have you had any sucess removing the trash directories?   You might be able to out grow the penalty by developing new links so that the number of suspicious (or bad) links falls below the tipping point.  On a recent WBF, Danny Sullivan opined that Penguin is just a devaluation of the bad links. (Not my opinion but it's an interesting opinion.)    No one has shared results but some people have suggested combining removing links with developing new strong ones.

Penguin is bizarre.  Some of my pages are (very) slowly returning to their former top positions even when some of the bad links point to them.   New pages with extensive content (think 2,000 words of unique/expert content) were among the first 2 - 3 to cover the event but now rank around 120. (Ouch).

I share your suspicion that for many of our sites, it's aggressive use of anchor text.   Developing non-aggressive links may dig us out.  Would love to hear from anyone who had tried this and what results they acheived.

Ben - I agree.   Using a shared hosting service and a CDN provider, 1.5 is probably better than my average load speed especially on pages that have earned extensive comments.  Everyone using shared hosting would be penalized.  Sounds like someone is selling patent medicines to me.

Alan - I haven't seen any postings by anyone who has tried this but several people have suggested this tactic in various forums.  If you try this, please share whether it works.  Given the lack of success in response to deletion requests, a more direct route would be much more effective and preferred if G wants to favor it.

Riplash & PVB : Thanks for sharing your observation about local.    2 - 3 months ago, I joined a thread and asked whether anyone thought optimizing for local would decrease national search SERP rankings and no one seemed to think it would.  My main site got mauled by B&W critters in April.  So, if I set up a new site to replace it, I'll set up a micro-site optimized for local and move the good local links there rather than to any replacement non-localized site.  Again, thanks for sharing this useful info.

Cesar :

I looked at your Scrabble Dictionary site. (Nice).

I probably wouldn't blog but would spend time trying to involve the user with the site and try to build loyalty or continue to build authentic socials.   (On other info search sites, forums seem to do pretty well and moderating the UGC to exclude spam is far easier than sustaing a blog.)

The scrabble game is buried in a secondary page and takes a LONG time to load.  (The ads for "Buy A Link" which appear to delay the game loading up in Java do not help.)

I'd consider moving the game to the home page or at least makiing it more prominent as a choice on the home page.  Then, I'd consider awarding some sort of prize (a Scrabble game?) to whoever scores the highest on site that month.  If you can restrict the prize to those in your FB community, you'd limit the CPU drain and gain community in FB. (Sort of a take off on EGOL's "Free Beer").

Just my .02 drachma/euro....   We'll see what far more experienced people suggest.

The "Talk to a Local Bankruptcy Attorney" looks like a banner ad positioned between the title and the text.

Depening upon how deep you are in the SERPs, the visitors will likely have already found a much better answer to the question presented in your title "How Often".   The page does not mention the length of the periods where they are prohibitted from filing a subsequent chapter 7, 11, 12, or 13.  None of the prohibitionary periods are addressed in the page.   Depending upon the keywords your visitors are streaming in on, the title appears deceptive or the page is not helpful.

Way too many links.  And, the links are so short, I doubt a reader would find them descriptive or useful.

My .02 Drachma (or Euro)

Marie - I share your concerns.

I looked at MBG  1 - 2 months ago and carefully looked at the age of the sites which were seeking guest postings.   The vast majority of these sites were launched right after BMR etc. got delisted.  This did not appear to be a coincidence.

There are a few quality sites with invitations to guest write on MBG.  For that, we should be greatful.  Most of the select sites have strict content requirements which appear to be highly selective. (Think the kind of product that could take someone familiar with the subject 2 - 3 days to do even with the running head start).  For a 50 - 60 site, a small author attribution and no key words in the text seems rather "pricey" in terms of author time.  But, I suspect the stricter blogs are more likely to be around and pass along some authority beyond the next 24 months or so.

I agree with many of the reasons EGOL often states for not guest blogging (why would you ever create a competitor for your keywords by giving someone your valuable content?)   An exception I see is where the content is a "one off".  If you market only residential real estate and there is a quality, high authority website that will accept a great article you wrote about commercial real estate, the trade off between the link and giving up the possible links to your own site becomes more attractive for those of us who are not over the hump and need opportunities for quality IBL's.

The remainder is a rant but I feel like venting my frustration today.

Whether we want to recognize it or not G is culling the heard of websites and will continue to do so for the next couple of years.  The cost of entry and wait required until a site is profitable has increased over the past 2 - 3 years and will continue to climb.  G's algo is weighed toward "high authority" sites written by content writers (not experts and seldom local) rather than weighed towards experts (why not favor blogs written by licensed experts in the searcher's local rather than "about.com" and the like?)

I fear that after G devalues the MBG sites (for whatever reason G makes up), any professional sites which allow others of the same profession to guest blog could easily be deemed an impermissible "network" and would not carry much link juice anyway.

Whatever Penguin is, I will bet many more people on these forums will be affected by the time it reachs Ver 3.5.   G sat on so many violations of its sacred "Terms of Service" for so many years, it would be deemed to have waived them in any private commercial transaction.

Perhaps another search engine which weighs authorship authority on more than "links" from the "abouts" of the www and even weighs in locality more will pick up market share.  I'm not going to hold my breath waiting though.   Selling ads in organic and social is far to profitable to give the traffic away to content providers who are unlikely to advertise.

Zachary : I believe Penguinn attacks for a lot LESS than the volume you're describing especially if there are few powerful links to begin with.

Donnie - Great point about avoiding the Fiverr.  LOL.

Right now, I'm watching competitors stack on modest amounts of FB Likes & Tweets (say 10 - 15) but very few G+'s.  Since Penguin, I believe I'm oberserving more of this sort of optimization.

I've been wondering what the "over / under" is on when the search engines decide "enough social spam", let's penalize them.   Based upon how horribly long it took G to unleash the nasty little Penguin, I'd give it four years.   Probably worth it on a disposable site but I would not set up a flagship site for penalization / fail in the future.

Wianno168 - Did you change anything that you believe may have lead to the partial recovery (from 95 hits to 550-700)?   If so, please describe what you did.

Medlin04.  I agree with Robert, it's a cost-reward trade-off.

Since it's the paving industry we're talking about, how many nice relevant sites can there be ?

I doubt there are very many that carry substantially higher PA/DA.  Your customer's competition will face the same problem of available relevant links.

Unfortunately, G's continuing attack "freshness", Panda, and now Penguin will continue to favor the sites Robert dubbed "PA-60" sites which have high "authority" but are usually written by content writers who, no matter how good they are, often do not fully understand the subject however experienced and skilled the writer may be.

No matter how anyone characterizes it, G's culling websites.  Saves on search, helps sell adwords.   If there is a huge (intentional) weakness in G's formula, it's underweighting relevant small business links out to other sites in the same business.  I believe this part of the algo is intentional because the big sites often sell the advertising which directs the traffic to these small business.

The penalty varies widely by page.   The penalty appears to hit certain keywords associated with individual pages.

Been wondering about this myself.   Did anyone who had a site that was otherwise clean get slammed by Penguin because of IBL's derived from widgets?

@ John - Sounds like a useful experiment for someone who can code and wants YouMoz exposure.

@ G.L.A. -  G's "business model" is to use search to sell ads and not to provide the "best" results.   I would not be surprised at all if G is now penalizing large contact forms, contact forms above the fold, etc. at least as harshly as some of the earlier Panda updates penalized over use of AdWords.  If Adwords ads are not immune under certain conditions, why would contact forms for the site itself be treated more generously assuming all other things are equal ?

@ Ryan :  Have you actually observed 301 redirects associated with transferring content to a new domain result in a penalty against the new site ?

I ask because one of the potential Pennguin recovery strategies is to transfer the best content on the old (penalized) site to a new site and start afresh.  Part of the plan is to use 301 redirects so that G does not mistake the pages transferred to the new site as duplicating what G believes is still on the old (penalized) site.

I understand that G may do whatever it arbitrarily wants to its algo.  For those of us with large (long tail) sites, creating a new site from scratch to protect our sites from further algo changes may not be the correct choice unless there is already a penalty that you described.

By the way, if penalties do pass through a 301, some of us have suddenly valuable domains available for sale to the ethically challenged. (LOLZ)

Nathan -  Someone recently posted a thread on BHW which suggests Penguin penalized sites similar to what you're describing.  (The examples are worth looking at and comparing to your site.)  Other people are insisting that Penguin does not apply to any on page variables.  Without setting up some throw away sites and testing, I do not know the answer and wish I did.

Has anyone observed the effect on G organic traffic when a site which has little or no location information suddenly registers with the reputable "local" directories?

I am especially curious about results observations based upon G's behavior during the past several months.

It might be a hosting problem (the host is performing some non-routine mantenance) or possibly even a HUGE change in G's algo but I've observed a huge drop in my traffic after claiming a couple of the local listings earlier this week.  Until then, I doubt G had associated my site with my city.

A couple of other explanations are possible but the timing leaves me to doubt it's a coincidence.

T.I.A.