6 Months Later: Google Penguin Reactions, Predictions, Tools and Tips

In the 6 months since Google gave birth to the Penguin algorithm update it has had a dramatic effect on the SEO industry. For years Google’s rhetoric had been about quality of content and links, but they’d been unable to back up what they said you should (and perhaps more importantly shouldn’t) do with what the industry saw worked.

It was a cause of frustration for large numbers of white hat SEOs to be doing the sort of things that Google recommends (give or take), and still be outranked by the people spamming anchor text with low quality paid links. Wil Reynolds’s complained about exactly this when he spoke at Searchfest just 8 weeks before Penguin, stating that a client of his is “getting killed by a website who is just targeting tons of anchor text only links on GARBAGE sites and is KILLING my client in the rankings” (he later wrote it up for SEOmoz - get it here). So many were happy to see Penguin, whereas many were less so.

So to celebrate Penguin’s 6 month birthday, I did a survey, asking 78 SEOs about their experiences with Penguin, their reactions to it, predictions for the future and any Penguin tips.

I presented some of these survey results at SMX Stockholm, where I spoke with Per Eriksson from MediaAnalys about Penguin (my slides, Per’s slides). Per and I come from agencies that are quite different in many ways, but also similar in others, yet we found that the reaction to Penguin seems to be pretty consistent!

This post isn’t going to be a lot of commentary but is going to focus on a few interesting results from the survey, and what the SEOs had to say.

Prediction: The next big algorithm update?

In the last 2-3 years, we’ve seen May Day, then Panda, and most recently Penguin targeting authority, quality content and over-optimised links, with a bunch of other updates in between. I asked the SEOs what they thought Google would be targeting next. Here are the top 10 responses and how many votes they each got (the responses were freeform so I’ve categorised them as neatly as I could managed):

Chart showing what algorithm update will come next

Low quality Infographics and Guests Posts both make the lists, which makes sense as both have been mentioned by Matt Cutt’s as candidates for future updates.

Four people predicted Exact Match Domains would be hit, which they were 3 days afterwards, so prizes to John Trimble, Felix, Kyle Mitchell, and Distilled’s own Hannah Smith.

The top response though was that Google could be targeting fake social signals. Something that struck me, is that with that response 3 of the top 4 responses are all about social signals, trusting them, authoring them (which is a specific means of attaining trust, and one I think is important), and  penalising abuse. I found this interesting because it still isn’t clear exactly how such signals are used, and whether we are seeing causation or simply correlation.

Observation: Where would you typically expect to find a Penguin link hiding?

Another interesting result from the survey was regarding where you’d expect to find Penguin links living. This isn’t particularly actionable, I know - but I thought it worth a brief mention:

Graph showing where Penguin links are found

The interesting part of this for me is the emphasis on sitewide links being a strong Penguin signal. All the widgetbait we spent all that time crafting is now coming back to bite us in the ass.

Recovery Tips

There was lots of data to dig through, but amongst it were some interesting tips that are worthy of sharing:

“Group sites using IP address and quality metrics (e.g. PR) and target to largest groups for efficient removal.” - John Trimble

Gianluca Fiorelli suggests using Buzzstream for link cleanup - somewhat the reverse of what it is normally used for, but a cool idea.

On a similar theme, Mike King, has a great suggestion for using Boomerang for following up on link cleanup requests.

Not a tip per se - but a fantastic slide deck from Ian Howells that everyone trying to recover from Penguin should read: http://www.slideshare.net/ianhowells/life-after-penguin

“Sadly, smaller sites are often better starting over than trying to clean up, depending on the severity of the bad link profile.” - Cyrus Shepard

“Link Removal is NOT cost effective, if you can move domain you should. Even if you’re a large brand, if you can get a domain which is close enough then move. I’ve worked with clients who have cleaned up 1000s of links and huge portions of their profile to essentially be fobbed off with the standard Google response.” - Mike Litson

Recovery Tools

The top of most lists were the usual suspects MajesticOpenSiteExplorerSearchmetrics, Google Analytics, Ahrefs.

Two breakout tools which came up often and are specialised for Penguin recovery were SEO Gadgets Cleanup & Contact utility Tool which does what it says, and the totally awesome PanguinTool which overlays your analytics data with Panda and Penguin dates so you can identify easily if an update was responsible for your woes: 

Panguin tool showing Penguin strike

(thanks to Per at MediaAnalys for the image)

Other tools that got a few mentions were Link DetectiveLinkDetox report from LinkResearchTools.com, and Niels Bosma’s awesome SEOTools for Excel.

Reactions and Quotes

There were  too many to pull them all out, so I picked out some favourites. If you want to read more, see below to access the spreadsheet of the full answers.

“[Penguin targets] unearned links - those that aren’t naturally created as a result of doing #RCS” - Rand Fishkin

“If my company had embraced #RCS (Real Company... Stuff) a long time ago, we’d be in a much better spot today. We’re now focused on tactics that create real relationships, real content, and real influence (you know, with people). IF we succeed, no matter how Google decides to measure authority and trust, we’ll have it in spades.” - Nathan Grimm

“If [Google] are going to make huge changes like Penguin they should be better at running it more regularly and also provide more transparency to webmasters about what links they consider bad.” - Chris Clayton

“I have to say that overall I feel favourable about Penguin. It hasn’t been 100% successful, but on the whole the SERPs are more agreeable now than before.” - Iain Bartholomew

“They need to run a data refresh. I think the reason they haven’t done a data refresh is because something went wrong. Of course, they’ll never ever admit it.” - Sean

“I worry about the folks who purchased existing sites, or hired a VP of marketing they trusted, or otherwise got themselves in trouble. What’s their recourse in a system like this?” - Ian Lurie

“The use, or not, of anchor text shouldn’t steer you to the types of links you get. It’s rather the situation and the type of link that should dictate the anchor text. There are no hard and fast rules or percentages that you need to strive for... it’s all about keeping it natural.” - Paul Martin

“There’s a lot of blurring of Penguin penalty vs. manual penalty, despite what Matt Cutts has said recently at conferences.  It’s interesting however how many times a Penguin penalty hits and then the site ALSO gets the manual penalty email.” - Michael Cottam

Wrap Up

There were more questions, so if you want to dive into the data you can grab a copy here.

From my point of view - it is frustrating that Google, directly or indirectly, encouraged certain activities (optimising anchor text) by creating the eco-system that they did, and now slapped some people for exactly that when there never was a clear line of how much is too much. However, I feel like Penguin was a good update on the whole, and once the dust has settled it will have helped good sites more than it harmed.

My concern is, the main signals for Penguin links will be easily enough circumvented (tone down anchor text, cut out site-wide links, and try to avoid being on pages with people who are leaving a footprint) and we’ll end up with a not dissimilar situation again.

I’d love to here from people in the comments about their experiences, and particularly interested in what other people think is going to follow on from Penguin.

Tom Anthony

Tom Anthony

Joining Distilled as an SEO, Tom comes from a background in freelance web development. With a degree in Computer Science, a PhD in Artificial Intelligence (almost – he is still writing his thesis!) and having taught himself to program on a BBC...   read more

Get blog posts via email

23 Comments

  1. This has all happened before and will all happen again.

    Your last statement pretty much nails it. Savvy spammers have penguin safe anchor text now and whilst results improved many areas we work in are still rife with pay-to-play SEO.

    Will be interesting to see what 'jolts' are to come and how they turn the dial up with penguin.

    reply >
  2. Hey Tom,

    Thanks for giving my the credit for my prediction.

    reply >
  3. Personally, I don't worry about this too much as long as I know I'm writing honestly and to the best of my abilities. Sure, SEO is something I definitely utilize but at the end of the day, I don't let it overwhelm the main meat of a good article/blog post: excellent and valuable content.

    Thanks for sharing this!

    reply >
  4. Sandeep Gulati

    Hi Tom
    This is frickin the best post, great insight. the above tools are great in detecting it and since Google allowed Disavow Tool webmasters can remove links they are not able too.

    reply >
  5. Great article, and thanks for the new seo tools to check out - thought I'd seen them all until this post!

    reply >
  6. Great post Tom;

    In light of this,

    “Link Removal is NOT cost effective, if you can move domain you should. Even if you’re a large brand, if you can get a domain which is close enough then move. I’ve worked with clients who have cleaned up 1000s of links and huge portions of their profile to essentially be fobbed off with the standard Google response.”

    (I know it's not your quote but presumably your use of it means you agree with it)

    What do you think you should do if you have a client that's paid a few hundred thousand pounds for a domain name, that's saddled with thousands of bottom of the barrel links? Keep trying to clean up the profile or count your losses and move on?

    reply >
    • Hi Amy,

      If there links are deep links, I'd abandon those URLs (404 them) and hope that helps (it should - provided you don't have loads of Penguin links to the homepage). Check out Ian Howell's presentation that I link to where he tested this.

      If the links are to your homepage, then you have more of a problem, but maybe 404ing the others makes cleanup more accessible. However, if beyond that your choice is forced down to abandoning the domain, then I'd be tempted to try the link disavow tool (that didn't exist when I got the survey responses). However - I'd be cautious and only use it as a last resort currently.

    • Bill

      If google penguin is algorithmic then why would like removal make a difference?

  7. Tom!

    Fantastic post and thanks so much for taking the time to put together all that data!

    I'm working with one Penguin-hit site now and have had the chance to review dozens more from the Moz Q&A. Thing is, they all have verrrry suspect onsite issues - namely, keyword stuffing in the titles, URLs, descriptions, on-page and definitely on-site anchor text. As well as lots of spammed comments (with spammy followed links) that were not cleaned up.

    I have a high suspicion that Penguin is about more than just back links. I'm not sure where the assumption came from initially that Penguin was only about back links - I recall the words from Google being more like "webspam" and possibly "over-optimization".

    Couldn't it be possible Penguin could also be about "webspam" on your own site? And over-optimization on your own site?

    I'm just so skeptical of it ONLY being back links because I think we've yet to really see a 100% confirmed Penguin recovery - "recovery" meaning a screencap of analytics showing traffic tank on a Penguin update day, and then return on another update. Of course it can be debated there are more updates than what we are told, but still haven't seen a traffic pattern resembling that on any date. WPMU can be debated whether or not this was because the case went so public.

    Anyhow, until there's a confirmed recovery, I think we should be looking broader than just off-site back links. And thus looking beyond just link cleanup as a means to recover. I feel as if we need to show Google BOTH improved on and off-site signals.

    I could be totally 100% completely wrong in my hypothesis, so I'd love to hear thoughts.

    -Dan

    reply >
    • Hey Dan,

      I'm tempted to agree with you - I like the theory. Some of the respondents I quoted (and others too) also weren't convinced Penguin is a simple as just links.

      It would certainly make sense that they can have far more trust that they have 'caught you' if there are also on-site factors too. Especially when you consider the potential negative SEO connotations.

      In one niche, there is one player I have seen who I can clearly identify have paid for, and are renting, links but they have good on page, and haven't been hit. They have begun slipping slightly in the last couple of weeks though, which is interesting, but certainly no dramatic crash.

      Hard to get any concrete evidence either way though!

  8. From my experiences, all of my sites were hit in the original Penguin update. The problem was that while I had good sites with high quality links, I also ended up buying a lot of low quality links from places like DigitalForums.

    Panda and Penguin were the first time Google actually used negative rankings to your site, which is why it caught so many people out. Before that we all knew it was just buy tons of links and get as many positive signals as possible.

    It takes time to recover and make various changes to sites to remove the negative signals. I'm on the right track now, building white hat sites, but then this EMD thing seems to have hit a few of my sites including StudentMoney.co.uk which is a nasty pain.

    I don't think SEO and Google's algorithm has really changed at all - it's just some extra filters like Panda and Penguin that you need to get through.

    reply >
  9. @Dan & @Tom -

    I am also inclined to agree with your proposition Dan. I have found a number of instances where keyword stuffing on site, specifically in page titles and URL's seem to be keeping otherwise good sites/content down in the SERP's.

    To clarify I'm categorizing 'stuffing' based on keywords being representative of a large percentage of total kw's both in titles/anchors/on-page. For a bit more context feel free to take a look at this convo on Google+ and share your thoughts.

    Thanks for all the research in this post.

    reply >
  10. Im complete agree whit your pint of view.

    But I also thing that If GOO are going to make huge changes like Penguin they should prevent us , to be more transparent... and give us more tools to react.

    reply >
  11. Great post, Tom. I just wanted to expand on the quote you got from me. At the time it had been 5 months since the initial Penguin launch. Overall I think Penguin was a success though if you read the "black hat" forums they'll tell you it was a complete failure and that their "totally relevant" site deserves to be ranked despite them taking the easy road.

    I still believe that something went wrong for Google to have to wait 5+ months for a Penguin refresh. EIther that, or they wanted to wait and do a refresh after testing out the Disavow Links tool. According to one of the companies that tested the tool before it was launched, their "penguinized" sites recovered after the most recent Penguin update. Could that have been because of the links they disavowed? Who knows.

    reply >
  12. It's really a sad thing to be at the bad end of the Penguin update. I personally think that the disavowing tool still has a long way to go to get re-ranked if you got a penalty and it's really true that cleaning up bad links isn't cost-effective at all. More reason for SEOs to invest on increasing their unearned links. Cheers to a Promising and exciting future for us.

    Thanks for this interesting entry Tom!

    reply >
  13. I really agree with Chris Clayton that it would be preferable for google to increase transparency. I am simply a website owner so am far from an SEO professional but to me it is akin to bad parenting -> rousing on your kid for not doing the right thing but not teaching them the proper way to get it spot on in the future.

    reply >
  14. I wish I'd read this before I came in on Friday, could have said thanks for the mention in person.

    I definitely agree with your conclusion and believe that Penguin is an acceptable initial step in addressing the problems that led to frustrations like 'Google is making Liars of us', etc.

    We're not all the way there yet, but it's clearly a step in the right direction.

    reply >
  15. Zach

    @ Dan:

    I also think that part of the reason you are not seeing the recovery is that the traffic and rankings were manipulated in the first place. Without the penguin links, many sites wouldn't be so high in the rankings in the first place. I look at it as a "correction".

    reply >
  16. Not sure if i just missed it but a couple of things that I also feel were major players in the Penguin update were the network links as well. Much like sitewide links if a site had many links on different sites with the same IP it posed an issue. The other factor that made the update so major was the fact that it considered previously built links. In my experience it was the article syndication and heavy anchor text usage that got me.

    reply >
  17. Thanks for a great post, and specially the PanguinTool, It helped me to identify few of the updates which affected my website.

    reply >
  18. Penguine hurt the spammers but i observe that only directory submission spams are hurts badly while black-hat tactis are still effective. comments spammers are little bit hurts.

    If any body can precisely elaborate the penguin algo update. then i will be grateful to you.

    reply >

Leave a Reply

Your email address will not be published. Required fields are marked *

*
*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>