I received This email from one of my ad companies today, it is filled with great information and statistics about penguin. I was just looking to pass this info on.... Depending on your approach to SEO April 24th was either a terrific day or a very, very bad day... and so began Google's "Penguin" roll out. Since then our team has been working around the clock to figure out what happened, what it means for our affiliates and what you can do to recover lost rankings and/or protect your rankings going forward. So in this issue we've come up with a list of actionable strategies you should be employing to prevent further penalization and/or regain lost ground... Strategy #1: Present Natural Signals (Onpage) Penguin is the result of a number of algorithmic changes, not the least of which is a recent change to their "Keyword stuffing classifier" algorithm (http://insidesearch.blogspot.ca/2012/05/search-quality-highlights-53-changes.html). This means that affiliates need to watch out for "onpage spam" which is a euphemism for both overly specific (exact match queries) and overly repeated onpage semantics in relation to other "spam" or "quality" issues. In short watch for these onpage issues: Keyword stuffing - don't have the same keywords in your title as in your URL. Once in one place is fine. Optimize for secondary keywords in your title. Google knows what your site is about. Also: Don't exceed 3% keyword onpage density, especially for a "money" keyword. Just write it naturally, Google is very good at figuring out what a page is about;Don't put the same money keywords in your META keywords or Description tag. If you don't need to rank for Bing, don't even USE a META keywords tag. Write your META description to sound like a natural call to action for people, not search engines; Page stuffing - don't have numerous pages on topics so related that the titles and URLs would share 70% of the same keywords in the same order. If you do, make sure the pages are +80% unique in content and (as a rule of thumb) over 800+ words in length so they are providing real, individual, value; Avoid Duplicate Content (onpage) - if you want it to rank it has to be +65% unique. [*=1]Use a rel=canonical or rel=author tag to protect the copy; [*=1]Index or ping the content as soon as it is posted; [*=1]Use copy protection to make it harder to steal; [*=1]Update regularly so your site gets spidered more often; [*=1]Add date semantics into your content (e.g.: "updated 1/12/12); Avoid Boiler Plate content - don't have blocks of text that are duplicated or not adding to the value of the page - change that text to GIF or PNG text to remove it from being counted against you; Avoid Spam Linkouts - Part of a site's trust signals are what sites it recommends (links out to). Link out to a few reputable and highly trusted / professional on topic sites only; Have a Big Authority Site - Part of a site's trust / authority signals include how many indexed quality pages it has. Also this diversifies the anchor proportions that Panda / Penguin uses. See strategy #2: offpage issues; Avoid hiding keywords, cloaking, doorway pages, etc. - These are short term tricks that will not work for long, especially against Negative SEO "outing" of sites (Google's webspam and paid link forms, etc). These tactics are not sustainable long term; Increase Quality - Although not necessarily part of Penguin, the following is always a good idea: [*=1]Usage Metrics: Add movies, images, charts, lists, how to's, and dynamic content to keep eyeballs on pages longer; [*=1]Add Trust Semantics - like reviews, accreditations, awards, recommendations from NPOs, charities, officiality signals like TM, R, etc.; Strategy #2: Present Natural Signals (Offpage) Of course, Google Penguin also checks for unnatural and manipulative signals offpage. This typically includes links and the quality of offpage backlinking pages. In short, our research indicates you should watch for the following: Too high percentage of exact match queries (EMQs) in backlinks - Our research indicates that, to be safe, any given anchor text cannot be replicated over 10% of one's total backlink portfolio. For e.g. if these were your index page backlinks: Red apples... 19% of total backlink anchors Buy red apples ... 18% Red apples review ... 12%.... The site in question could have an "over optimization" issue especially if any of these keywords are ALSO matched in the URL and or Title of the page. Add any of the onpage issues (from above) to the page in question and a penalty / ranking issue is highly likely; Total EMQs exceed 30% of total portfolio overall - Some research indicates that "normal" sites have a much lower EMQ / "money keyword" threshold: under 40% of total link profile. The following total anchor percentages are a good rule of thumb: [*=1]30% EMQs, e.g.: "buy red apples here" [*=1]10% Partial Match, e.g.: "Want apples? Go red here!" [*=1]30% URL/Brand, e.g.: "www.bigred.com" *** Note this only applies if your URL/Brand is NOT an EMQ. If so, add 30% partial matches; [*=1]30% Generic (non EMQ or PMQ), e.g.: "food", or "get fruit info here", "click here", etc. Remember: Exact match queries (EMQs) include phrases, e.g.: "red apples" and "nice red apples review" both have the same EMQ in it; In theory, this applies to any anchor, not just money keywords and we have seen accidental non-money keyword repetition in anchor text still cause problems for sites (that also had other onpage issues). However, because few are optimizing for non-money keywords, it rarely happens that there are enough onpage and offpage signals on those keywords to cause issues. Lack of Natural Signals in Anchor Text - Another obvious problem is that manipulative linking targets EMQs and does not have random natural signals. Our research indicates this is strongly correlated with recently penalized sites. Make sure that: All Anchor text has natural signals - a backlink is a label. The question is how would another person linking to my site (because it is so awesome) label that link? Over 70% of your backlinks need to have "click here" like natural signals, e.g.: "this is a great site for red apples" not "red apples"; Anchor Text matching linker not searcher - natural links label what a site is about, not what searchers look for. Google is smart enough to tell the two apart. Links are not naturally made like this "want a bigger apple" but "good blog about how to get a huge apple." The former is what a searcher is typing in, and is clearly manipulative (especially if it lacks natural signals, and is over 10% of the anchor text portfolio, etc.), the latter is more how a linker would naturally label a site he was linking to because he truly thought it was a great resource. Less than 30% Duplication on Backlinking Pages - Our research into Penguin showed a strong (90%) correlation between affected sites, and backlinking pages which were often over 60% duplicated (from article syndication). This is a very high correlation. Even if Penguin 1.0 is not triggering on this, Penguin 2.0 is just around the corner. Avoid having unnatural signals in your semantic index; Avoid "linking" sites and "directories" - Our research also found an equally strong correlation with sites that were affected by Panda/Penguin and sites that had backlinks from: banned / deindexed blogs and article sites, low quality spam flagged or penalized sites (sites that were not ranking for their title tag, or has a "null" pagerank), sites with "free links", "submit your link here" and paid link semantics. Again, we do not know that Panda 3.6+, or Penguin 1.0 is using, or even causing these patterns, but it is a very, very strong correlation, and Penguin 2.0 may very well use such signals. Avoid putting backlinks on these "spam" sites. Instead... Maximize Natural Offpage Trust Signals - Try to maximize any trust or authority signals (this is used in Panda and may affect Penguin): Get natural +1s, Likes, Shares, Followers, Tweets, Re-Pins, Diggs, Tumblr, etc. Get natural looking links from high authority sites like news, product sites, education, and government publications (not open blogs, not where everyone else is spamming) - note: low quality relevancy links (like comment links) still "work" as long as the rest of these rules are followed; Reviews on trusted high quality review sites (like Google reviews, etc.); Strategy #3: Present More Brand Authority Part of the Penguin effect has also been to prioritize brand and "under-optimized" sites (see under "More Authoritative Results" -- http://insidesearch.blogspot.ca/2012/05/search-quality-highlights-53-changes.html). Our research into this algorithmic change indicates that Google has altered their pre-existing Query Deserves Diversity (QDD) algorithm to serve approximately 20% more "brand" (e.g., not www.redapple.com but www.bigred.com or www.johnsmithapples.com) and or (ironically) 20% more "new" or otherwise young, under-optimized exact match domain (EMD) sites. This is part of the reason why many search engine ranking pages (SERPs) have had blank, or low quality pages served higher than other more robust, but perhaps overly optimized sites. We do not know how long this specific change will remain in place, but to take advantage of this algorithmic change to QDD we suggest either: Make a Brand Domain - In the past using an EMD has been a favorite tactic given the boost such a domain used to give, and still does if it is not over optimized. However, now that Google is essentially reserving spots for qualifying brand sites to counter this tactic, it may be strategically viable to try promoting brand sites to take one of those "reserved" spots, and as a way to differentiate oneself from the herd of EMD review sites that come up; Stay with EMD sites - BUT stay under-optimized - New domains often get a boost, especially if they are an EMD. But inevitably they fall due to low semantic index (links, mentions and social signals). Now it seems google is willing to reserve a spot for a new / younger / less optimized site if, and only if, it seems very relevant (exact match on that query) AND has no over optimization profile. So the trick with EMDs now is to make sure one does not over optimize them, but instead targets at a more general link on and off page profile. In Summary: Present Natural Signals In short, the final strategy, the only viable strategy long term, is to simply present natural signals in everything you do. Ask yourself these questions: How would someone who liked my site naturally link to me? What would a normal, non-seo author write for my content? What would my link profile look like if I was referred to by other people, or in a white hat market? On average, Google is very good at determining what a site is about. There is no need to try to game relevancy signals, only authority signals. The only manipulative tactic that is always going to work is making your site look like Google's preferred site - i.e., the "natural" one (if such a thing exists). Presenting natural signals to Google is, and always has been, the end strategy.