SEO – Things you need to know

The current state of SEO

Search engine optimization (SEO) is an ever evolving, fast paced, constantly changing field. What has worked in the past, may or may not work today and is even less likely to work in the future. This results in a common mistake caring webmasters keep running into. Most SEO’s want to increase their organic traffic with highly targeted keyword rankings through various (classic) methods and when their efforts don’t reflect the intended result, they start to care too much about SEO. In fact, they spend most of their time doing SEO that could better be spent in producing more quality content with a far greater result.

That said, as a webmaster, you should definitely not ignore SEO completely. There are many more ways of doing something wrong, which results in a serious disadvantage, than doing it right. This page tries to give you a better understanding about the current situation of SEO in 2014. What should I do? What to avoid? What is important?

Since Google is by far the most used search engine and market leader, it makes sense to focus our SEO efforts primarily on Google.

Panda, Penguin and Hummingbird – Information and observations

Panda (launched 02/2011)

Goal: Increase search ranking of pages with high quality content while decreasing rankings of low quality sites like content farms significantly.

How Panda works

If the Panda algorithm finds lots of low quality or duplicated content on a given domain, it is likely that Panda not only reduces the ranking of those pages, but the entire domain. Even if you have a good portion of valuable, unique information, you may still be affected. I tested this theory by setting up a low quality site consisting of 150.000 different products (which made sense, because duplicate low quality content is prevalent across e-commerce websites and there was also almost no competition anymore except Amazon, Ebay and very few others. Google eradicated most of these sites successfully in 2013). To see what happens, I mixed in a couple of valuable pages.

After the initial crawl and index, it took Google two weeks to slap the entire domain. I verified the test with another domain and this time, it took Google only six days. With my third test domain I put at least SOME effort to it to bypass the most obvious detection methods (cloaking and nofollowing outgoing links, no standard manufactory product data, (somewhat) unique product titles along with a few other things). This one was actually quite lucrative in terms of revenue. Not that it was good, it was just another spam site, set up mainly for testing purposes, but it hasn’t been caught by Panda at that time. As it reached about 800 daily visitors, it eventually got caught by Googles manual webspam team though, which was actually good, so I could test how long it would take me to get a “Pure Spam Domain” like this lifted from its penalty through several reconsideration requests (to make it short, after deletion of everything and instead putting a dozen helpful articles on it, the penalty was lifted on my fifth reconsideration request).

How to tell if you got hit by Panda

If you are not sure about the quality of content on your site, answer the following questions:

Do you use a Content Management System (CMS) like WordPress, Joomla, Drupal or Typo3 and let Google index everything including feed, author archive, tags, categories, keyword searches, variables and does your theme represent the full article instead of an excerpt on some of these pages?
Do you have duplicate or overlapping content without proper pagination?
Is the content on your site typically very short (<300 words) AND does it really answer the topic described in the headline? Is it possible to combine the content of some pages? Longer articles tend to do better these days.
Do you have lots of pages with only minor keyword differences?
Do you or your authors write with great care considering all sorts of grammatical errors?
Do you accept guest posts and / or is most of the site’s content from a large number of creators? What have you done to prevent mass-produced spun content?
Do you use large text blocks on a template basis (for most or all of your pages)?
Do you steal or use duplicate content from larger sites and just rewrite or spin it a little?
Do you have more than three Adspaces?
For e-commerce websites: Do all product pages have standard manufacturer product data? What about the title? What have you done to make both title and content at least somewhat unique? If you only use manufacturer data, how is your site better than sites like Amazon? (This is actually important, because Google went on a killing spree for ecommerce sites (especially affiliate sites) that can’t compete at a high level since mid 2013.)

Affected by Panda? What you can do about it

Block all auto-generated and low quality content. It is not enough to noindex those pages. Block the crawling with your robots.txt, delete it and return 410 instead of 404 for faster processing or update the content to improve the overall quality.
Be patient. Google never completely forgets URLs, even if you removed those years ago. They might show up once in a while in Google Webmaster Tools (GWT). Depending on how many pages need to be recrawled and updated, this may take several months. According to my own tests, I came to the conclusion, that you simply can’t get out of a Panda filter as long as Google refuses to push a new update directly in their datacenters, or your Panda punishment time expires. A datacenter update happens irregularly. In the last couple of years you could notice it through a pagerank update or an otherwise major update to the existing search algorithm. I have yet to see one single domain, that had a reasonable amount of pages (>500) affected get out of the Panda filter on a “regular” day.
If you got hit, you won’t receive a message in Google Webmaster Tools and yes,
- there is no way to file a reconsideration request, even if you make serious changes to both your code and pages.
- it is faster and less trouble to get a second shot on a domain that was previously declared even as “Pure Spam”.

What can we learn from Panda? Time (and Google) should be our friend, not our enemy. We should focus our future efforts more on content quality instead of fast results.

Penguin (launched 04/2012)

Goal: Decrease rankings of websites using Black Hat SEO techniques, especially different kinds of linkbuilding, in order to fight webspam.

When is a website affected by Penguin?

Link manipulation through large amounts of repeating external anchor text keywords (internal keyword links have no effect on Penguin. It is even advised to use rich internal keyword link terms heavily. Also brandlinks are less restrictive).

Unnatural example: 1000 backlinks from different domains, all of them point to the exact same page with the keyword “cheap car insurance”.
Natural example: 1000 backlinks from different domains pointing to different pages with different keywords including no keyword at all, stop words and generic links like “here”, “click here”, “this page” in addition to brand links like “domain”, “domain.com” and “www.domain.com”. Anchor text diversity along with an assortment of link types is key.

There is no rule like: “We allow a maximum of 30% of the same external keyword anchor text, if you have more than 20 backlinks to the same page. If you have 31%, we will slap you”. In fact, Google needs to define rulesets in order to trigger the Penguin filter, but these behave different for every branch. It is likely that Google looks at average branch values and if your site has a much higher same-text-keyword-anchor-density without a really good reason (e.g. a widely used, highly trusted plugin), you will likely be affected by Penguin.
Paid links are a clear violation of Google’s policy. But how does Penguin decide, if an external link is paid for or is a real recommendation? Simply put, the algorithm can’t tell for sure, but it can assume if a link is legitimate or not. Important factors are thematic relevance (see below), the used link text and the trust level difference of the linked domain to the linking domain. In Googles eyes it seems unlikely, that, for example, a PR5 page on a PR8 domain links to a PR0 page on a PR1 domain directly. The other way around seems much more likely.
To determine if a backlink is legitimate or not, the thematic relevance of the linking page to the linked page is also an important factor. If you talk about “cheap car insurance” and link to a porn site, you might want to reconsider whether this is the best way to go.
If you do it, at least build links at a constant rate. Google remembers the date of each external link when it first sees it. If you placed 100 links per month over the last three months, then zero the following three months and 10.000 the month after that, you’ll likely trigger an unnatural links penalty.
Another aspect is the quality of your links. If the majority of your backlinks are the result from easy-to-come-by kind of links like blog commenting, directories, link lists, profiles and news archives, you might want to rethink your strategy, because these kind of links are devalued by Google. In fact, Google announced in January 2014 that this is also true for backlinks coming from guestblogging and article directories. In short, almost every type of easy-to-get backlink can, if acquired in large quantities, be harmful for your website. On the other hand, if you link to extremely good content, you’ll get a small boost for your site as well (it’s a quality signal, so do it).

How to tell, if you got hit by Penguin

Like with Panda, you do not get an email when Penguin negatively impacts your site. But if you receive an “unnatural links penalty” in GWT, then consider yourself lucky, because you can at least do something about it and you are not completely left alone in the darkness. This happens mainly, if Google thinks you are the one who placed the backlinks and thus violated Googles Quality Guidelines.
If you detect a huge visibility penalty starting on a specific date, check if Google rolled out a Penguin update that day. The graph might look something like the image below.

Should Penguin be your primary concern, try to keep a holistic view on the subject. The individual signals are just parts of a puzzle. If you don’t send multiple Black Hat signals, you can come by with a much higher rate of, let’s say, specific keyword anchor text. But if you have a high amount of non thematic, keyword rich links from low quality sources, you might trigger a penalty, even if the overall keyword anchor text is lower than the ones from your competitors. In addition to this, highly trusted and extremely big sites like Amazon or Adobe can have theoretically (almost) unlimited Black Hat signals without triggering a Penguin penalty, because, on the other hand, they have so many more quality signals, that it doesn’t really matter anymore. This is why most frustrated SEOs these days speak about an unfair “free pass” for big companies. The truth is it is unlikely that Google gives them any special treatment. But because of their size and very high number of quality backlinks to their individual pages, they must rank high for almost anything and in turn can do lots of crazy stuff, an average webmaster simply can’t do anymore. At least not, without the risk of getting wiped out by Google. (For example, good luck ranking high nowadays with non-unique content in combination with a HUGE site-wide footer that has an awful lot of followed links in it. Yes, even external ones.) Whether things like that are considered fair or not is up to your judgment.

Affected by Penguin? What you can do about it

Status quo doesn’t help. Clean your backlink profile by thoroughly analyzing it and disavow bad links. Before disavowing, you might want to contact the link sources and ask them to remove the links (if they ask for money, ignore it and just disavow). Keep in mind, that disavowed links can’t harm (but also can’t help) you anymore. Investigating and analyzing a backlink profile is a time-consuming process. If you are not sure what to do, hire a professional to look at your profile.
It might be a good idea to use the disavow tool from time to time to get rid of crappy, auto generated links that might harm your site. The problem is it’s up to your judgment what to disavow. Unless you get an unnatural links penalty, Google doesn’t provide examples of which links are considered bad. So by using it, there is also a chance to reduce your SERP (search engine results page) visibility even further. On the other hand, suspicious low quality links don’t pass a huge amount of link juice anyway.
Be patient. It takes a while to notice the effects. Unless you have a manual links penalty, you simply can’t get out of the Penguin filter as long as Google refuses to push a new big update directly in their datacenters or your Penguin filter penalty expires, which can take several months, maybe even over a year. In the last couple of years you could notice a datacenter push through a publicly visible pagerank update or an otherwise major change to the existing search algorithm (like another Penguin update). People sometimes post about recovery stories, but from my experience they are lacking important information or have other stuff like additional penalties going on, which got revoked, so I don’t buy it. I have yet to see one single domain that had a reasonable amount of backlinks affected get out of the Penguin filter on a “regular” day. Even if it sounds odd, but you have much better chances of lifting your Penguin algorithmic penalty early, if Google also catches your domain for violation of their Quality Guidelines and you can revoke the penalty through a successful reconsideration request, because this also resets all other filters.

Why are webmasters still afraid of a Penguin update?

Most SEOs are afraid of Penguin, because the quality of backlinks is still the most important ranking factor. Good luck trying to rank in a competitive niche without some kind of link building strategy and content, which can’t attract lots of links in a natural way. Most people don’t realize that not all topics and branches can attract serious amounts of natural backlinks, no matter how valuable the given information is for the user and how structured it is represented. The hard truth is, if you work in such an environment and you don’t build backlinks, you simply get steamrolled by your competition. And if you do, you might get punished in a future Penguin update. This is the world we live in today. So the decision is yours. But whatever you do, please hear this advice: If you build backlinks, don’t feel bad about it, but do it in a meaningful way. A low amount of quality links can serve you much better than a high amount of easy-to-get-links. Creating good content and help it reach people with some active quality link building (e.g. links, that receive real clicks) is a legitimate approach. Social media marketing is only one way of doing it.

Hummingbird (launched 08/2013)

Goal: Increase the quality of search results by trying to better understand the intent of the users search query. If you search for a whole sentence, Hummingbird tries to understand the meaning of the sentence, rather than just looking at it word for word. Pages that match the meaning of the sentence will do better than pages matching just a couple of words. This was an important improvement, because every year, more and more people use natural language voice inputs. If you search for only one or two word phrases, you might not notice anything at all. But long tail searches are heavily affected by Hummingbird.

What you can do to make your site Hummingbird proof

Write for your users, not for search engines. For example, stop stuffing keywords and also stop creating multiple similar keyword rich headlines (<h1>how to make coffee</h1> | <h2>how to make coffee the right way</h2> | <h3>how do I make coffee?</h3> | <h4>coffeemaking explained</h4>) in the same article. Those days belong to the past. Google understands, that your post is about <h1>the best way to make coffee</h1> or in case of the headline of this paragraph: <h4>How to optimize a website for Hummingbird</h4>. No need to repeat variations of it multiple times anymore.

Possibilities, risks and benefits

If you have read this far, you probably want one or more of the following questions answered:

“How can I do better in Google?”
“How can I rank number one?”
“How can I get back the rankings I lost?”

Every case is different but you are not alone. These kind of questions are constantly asked by more than 70% of caring webmasters in the SEO world. The times, where you could easily manipulate Googles SERPS by using shady methods without much risk are mainly over. In general, the probably best advice (and yes, most people don’t want to hear it, because it is not the answer they were seeking for) is to focus on great content that can potentially attract links and grow slowly over time. If this can’t be done (because you are working in a niche that has no way of acquiring links, yes, that is not only totally possible, but also gets overlooked way too often) and your mind set is like “I need much more links to succeed or else I’m finished”, then you’re setting yourself up for lots of sleepless nights. Instead of watching the doomsday coming closer, it might be a good idea to invest more of your time in offline strategies to promote whatever you want to achieve. Rather think about “What can I do besides the internet?” instead of “How can I win the internet?” or “How can I please Google?”. That said we are not (yet) in a state, where every online effort you make gets punished right away. So let’s talk about possibilities, risks and benefits.

High risk, low benefits

Link building tools

Using link building tools like SEnuke, GSA, Magic Submitter or Ultimate Demon on non high authority sites. You may still get good results utilizing tools like these for optimizing YouTube video urls, but using them for tiered link building or blasting links directly to your site, will likely do you more harm than good.

Comment Spam

Why comment spam doesn’t work anymore, should be pretty obvious. Links from comments are in general nofollowed and if you put a keyword anchor text instead of your name while comment spamming, Penguin is even more likely to knock on your door to say hi.

High risk, high benefits

Buying links: Still an option?

In these days, you might want to think twice if you intend to buy (direct) links, especially if they are either site-wide or blog links in the typical “three external links per post” style. Those two are the most commonly bought type of links (not counting large quantities of useless spam links on spam pages with hundreds of external links). They are quite easy to detect, so you might want to avoid it. I’m not here to judge someone for buying links, but if you really want to know what kind of bought links are (for the time being) still quite safe, read on.

Check the current trust level of your own domain(s). Googles Pagerank indicator was quite good for this in the past, but as you already know, at least the publicly visible part doesn’t get updated frequently anymore, so use Majestic (or something similar) instead.
Avoid buying links from sources that are way out of your league. This looks suspicious (and it also represents a risk for the link seller). Instead, look for sources that are only slightly better than what you have right now and if it also has a thematic relevance, it’s even better. As long as the link seller doesn’t get caught for selling links in large quantities or you are using a Black Hat link network, the chances of successfully getting away with it are still pretty good. Yes, paid links are risky, but the reward is still quite high. This may change in the future.

Utilizing your own link network with extensive, keyword rich site-wide footer links

This is still very effective, but if you don’t have high authority sites with lots of quality signals, chances are, it won’t do you any good in the long run. Why? Should one of your sites get caught by one of Googles quality raters, your entire network is at risk. Some really big companies have done this for years in order to gain an additional advantage over their competition, fully knowing, that their entire network is rock solid, so they can interlink all sites on a page level for maximum link juice flow.

Low risk, low benefits

Create large amounts of mediocre content and living in hopes, that this solves all your problems. A large amount of medium quality content alone is not king. Sure, if it’s still somewhat useful, you won’t get penalized for it, but without some kind of marketing, it won’t get you anywhere, as long as you can’t acquire natural backlinks from it.

Low risk, high benefits

Create excellent content that can truly earn backlinks over time.
Reclaim links that you already have, by using GWT to discover broken links to no longer existing URLs and fix them via 301 redirect.
Try to get links only from quality sources that require a manual review with editorial approval and are in general hard to get. By submitting outstanding content that is truly unique, helpful and insightful, you can drastically increase your chances. But keep in mind, since it has been officially doomed, you can’t acquire lots of links from guestblogging anymore, so you are better off trying to build human relations with other webmasters.
Generally speaking, each quality link is more valuable than ever. Over time, in most cases quality outranks quantity. So focus your efforts based on that.

A word on social media and social signals as a ranking factor

There is no doubt that incorporating social media as an alternative way of generating free and highly targeted traffic to your site is a good thing. That said, social signals (e.g. number of likes, shares, +1s) as a ranking factor for search engines are hopelessly overrated. To get straight to the point, they don’t do a thing. Even having thousands of likes and tweets that don’t result in real backlinks from your visitors, won’t move you up one single place in the SERPs. They can only help you, if you acquire additional backlinks through them. This may change in the future, but I find it unlikely, because these signals are way too easy to manipulate and Google is most certainly not interested in relying on things, they can’t reliably get under their control and protect it from spam abuse at the same time. This is also true for Googles own +1 button.

If you don’t believe me, because you’ve heard different things elsewhere, then go for it and try it yourself. Create a mediocre page on a non-authoritative domain, give it a few thousand likes, tweets, pins etc. and watch it not rank a single bit as long as you like.

On the other hand, Googles author rank is real, so you are well advised to join Google+.

Having an author image show up in the SERPs for a given search term will increase your click throughs.
If you are already a reputable author on a given subject on Google+, you might indeed rank higher than others. The author rank is implemented as a clear quality signal in Googles current algorithm.

Other generally useful tips

Make your site mobile friendly with a responsive design. Every year, more people use mobile devices. Even if it’s not strictly for SEO purposes, a user who can’t use your site is a visitor, who you can’t potentially make business with.
Use pictures and if possible videos. Both are indeed SEO quality signals, so use them. Don’t just present plain text.
Depending on your niche, get local! Google tries hard to boost sites in your language, your country and especially your town. You might be able to use this for your advantage.
Content quality is evaluated in three ways:
- Richness: How insightful is your content? What does this mean? It’s whether your article as a whole is interpreted as “shallow” (e.g. writing lots of stuff without actually saying anything or contributing to a possible solution) or “helpful” (e.g. providing precise answers to thematically related questions).
- Uniqueness: How unique is your content compared to other thematically similar sources?
- Freshness: Google likes new and fresh things. More recent posts may do better than older ones on a similar topic.
Try to always use correct meta data. A poor set of meta data can screw up your entire site.
Unless you are running a churn and burn business, think long-term (at least two years in advance) and focus your SEO efforts based on a corresponding timeline. If you are only after fast results, you are likely going to negatively impact your long-term goals. In two years from now, no one is interested whether you where once the number one or not. Much more important is the question: “Does future Google still like, what I do today?”
Speed is obviously important. A very slow loading website can’t compete with lightning fast ones. In most cases, caching of recurring objects, using a content delivery network (CDN) for static files, reducing PHP database requests, enabling browser cache and minification of Javascript, HTML and CSS results in a performance boost. The caching method greatly depends on whether you have root access to your hosting server (VServer, Dedicated > All, including APC, Memcached, XCache…) or not (Shared > HDD, Database).
Backlink relevancy matters. This is especially true for foreign languages. It’s always best to acquire links from domains in your target country.

Recommendation: Be a White Hat, but keep testing behind the scenes

Succeeding is not about ranking as high as possible in a short amount of time (which is more suited for churn and burn sites) and not just about marketing. It is about having something to market in the first place: A valuable product or service, well presented, with excellent content, based on code that gets indexed smoothly.

Over time, you will build a certain reputation (and trust) in form of freely given links or references. If you can keep it up, you might be able to attract more visitors. It all depends on what kind of niche you are trying to succeed. Why is this important?

It might be hard to accept, but in the long run, the best links are not the ones passing the most link juice today. The best links are freely given references from people who like your site. These links might not get you straight to where you want to rank, but with Googles current anti-spam-direction, it seems to be only a matter of time, until these links are the only ones left, that could possibly get you to rank high in the long run.

In the past, the term SEO had a lot in common with ways to trick or beat Googles algorithm, because it has been extremely profitable. Since then, SEO has undergone a complete overhaul, so don’t be one of those who fail to adapt: The question is not “How can I cheat Google?” anymore. Instead associate the term SEO more with “How can I improve the overall quality (code and content) of my site in order to make it easier for Googles employees to do their job right?”

That said, I would still recommend you to keep testing stuff! Even if it sounds a bit contradictory to the paragraph above, don’t stop testing new and crazy ideas. If you don’t try a lot of different things (and mostly fail with it), you always have to rely on second hand experience, which, in many cases, do not provide you with the same knowledge and flexibility you get, when you experience something first-hand.

In general, I would listen more closely to a person talking about SEO that got a bunch of domains completely killed by Google in different kinds of ways over time (on purpose or not) than someone, who did everything “right” by avoiding all possible risks and mistakes since the beginning. The first person knows exactly why he is doing something the way he does, because he experienced the “why” first-hand and experience is worth a lot. Also, the most efficient methods keep constantly changing while only the basis (e.g. acquire quality backlinks, write excellent content) remains relatively untouched. So if you got caught in the past, don’t feel bad about it. Learn from it, move on and do it better next time. And if not, don’t be afraid and test awesome things.

That’s it for now. Thanks for reading and have a nice day!

Webmasterhelp James Star