Systematic SEO

This chapter will give you a bird’s view of everything SEO, including the user, the webproperty, Google, distribution, and traffic as well as feedback loops and targeting. All of these topics will be revisited again in detail later in this book.

The User

When in doubt, start with the user. As there is always doubt in the online world, always start with the user.

The user is a strange, mystical creature. They:

• do not have an age but an age range
• do not have a gender (but they are still never an “it”) but a gender distribution
• might use the website on a desktop or mobile (or something in-between)
• do what they want, not what we want
• might be located anywhere in the world, and even then, they could be at home on their couch, at work doing work, at work doing something else, on a bus, or even on a beach somewhere.

As a matter of fact, we don’t know much about the user until we have a closer look. The user is similar to Schrödinger’s Cat: “The User” is all of the things listed above and much more, at least until we have a closer look.

Furthermore, we can determine a lot of things about a single user after they have visited our site, but we cannot say much about users in general, and we cannot predict future users. What a bummer.

What we can say for certain is that the user is not us. We know the technical, business, and organizational reasons why our website is the way it is; the user doesn’t. We care about our business; the user doesn’t (yet). We (sometimes) read what we publish on our website; the user doesn’t.

Remember that we are users, too, on other web properties. The user is:

• efficient
• lazy
• clever

Any online business that treats its users like a mindless herd of lemming that just needs to get pushed down a funnel has already lost. The user knows what they want - but we don’t.

We must make it easy for them to do what they want, because the user is efficient and doesn’t like to waste time. On top of this, the user is also lazy and clever, and will always use the easiest means to accomplish their goal.

The Webproperty

Your webproperty might consist of one or more domains and one or more subdomains, and most of the time has a startpage and a lot of other pages. This image below is a symbolic presentation of a webproperty:

Links connect the startpage with pages, and links connect pages to other pages. There are thousands of variations and mutations of this arrangement (startpage only, single page apps, multiple startpages, and multi-language domains to name a few).

The User And The Webproperty

In a perfect world, we wouldn’t need to worry about anything. In a perfect world, a user would be interested in what we have to offer - they would simply type the URL into the address bar of the browser, and our startpage would open. They would navigate through our website using our well-thought-out navigation structure and do what we want them to do because we know what’s best for them.

Of course, this never happens. In most cases, the user does not know us, does not care about us, does not enter via the startpage and completely ignores our fancy navigation logic. And they don’t do what we want them to do. What the user does is google.

Google is a marketplace! It is by far the biggest marketplace this world has ever seen. Google takes an inventory (its index of web resources like web pages, images, videos, maps, products, locations, knowledge graph data and so on) and matches it with the user’s demand.

For all western markets, Google is the dominant search engine and therefore the dominant content marketplace. Most of the principles introduced in this book are also suitable for other search engines, but it hardly matters.

Google is the undisputed number one search engine in all western markets. Second place goes to YouTube, third to Amazon and then there are some other minor players.

Some people (mostly politicians and media companies) say that Google is a monopoly (in most markets it most likely is), and while it does make for some interesting discussion, the topic is not relevant for SEO. You have to optimize for the market that there is, not for a market you wish to exist. And for us, that market is Google.

The User And Google

The user googles, always, all the time. He or she types (or speaks) a query in the Google Search input field or Chrome omnibar (or the equivalent on the other browsers) or search widget on a phone and thousands of other places. The user does not pause to think about what to write, they just google.

A query can range from a word to multiple words to a full sentence. It’s rarely even a thought; it’s usually half of a thought, one that the user is still thinking. Quite often it’s not even spelled right or in a logical order.

But every user that googles has an intent: a “job to be done,” something they want to achieve. Everybody does it, all the time, without much thinking. They just google.

We can categorize a user’s search intent into “Do-,” “Know-” and “Go-” queries.

1. Do-queries (action intent): also known as “action” and “transaction” queries. The user wants to accomplish a goal or engage in an activity: download an app, play an online game, book a hotel, order a new coffee machine or find awesome cat videos to watch.
2. Know-queries (information intent): the user wants to find information; they want to know something for example: the best time to travel to Sri Lanka, or what the weather is like in Italy in June.
3. Go-queries (navigation intent): the user wants to navigate to an individual web page that they already know about, most of the time the startpage of a particular webproperty such as: Twitter or Facebook.

This categorization of do-, know-, and go-queries does not include all varieties of human search behavior, but it is a good starting point for our purposes.

Google takes these chaotic, fragmented and often nonsensical queries, and enriches it with everything they know about the user. Everything they know about the user can be quite a lot:

• their location
• device
• internet speed
• current search session
• previous search history
• plus lots of other things we don’t know they know, which will have an influence.

Google takes a best guess at what the specific intent is, and then matches it to their inventory, which is a huge index of everything that is online. Everything online? Well, no, not exactly.

Google matches it to their own inventory: that is, everything that Google has discovered, crawled and indexed. So if we want our pages to be included in that inventory, we need to distribute our content to Google.

Whenever you see a Google query, you must not see these words as mere “keywords.” They are much more important. They are a textual representation of the user’s search intention.

If your goal is to have sustainable traffic, then it’s your job to fulfill their intent. You can - for a short time - gain traffic with targeting “keywords,” but you can only maintain sustainable traffic by meeting the user’s search intent.

When we talk about queries or search queries in this book, I always mean the “textual representation of the user’s intent.” I don’t mean “keywords.”

Here is what search demand looks like:

In every vertical (the segment of the market you are active in, for example, music, insurances, cars, vegan, dating, poker), there are always search queries that are frequently searched. We call these high volume queries. For the instances for the verticals given above, [shakira], [car insurance], [tesla], [vegan new york] or [online poker] would be high volume queries.

Now, in every vertical, there are always queries that will be searched very infrequently, from a few times a day to once in the whole of eternity. We call these “low volume queries.” So for the instances above, examples would be:

• [shakira worst video using oil]
• [car insurance Ohio drunk driving]
• [should I buy an e-card mastercard]
• [vegetables for breakfast berghausen austria]
• [online poker strategy minimal budget 10$]

The funny thing is that for each and every vertical if you add the search demand of all the high volume queries together, and compare them to the sum of all the low volume queries, you will find that the total amount of searches for the “low volume” queries is much, much bigger than the sum of the “high volume” queries.

The sum of the low volume queries is called the long tail. The sum of the “high volume” queries is referred to as the short tail. For every online vertical, the long tail is much bigger than the short tail. As people search in more diverse ways, the long tail gets longer and longer.

To be successful at SEO means to have a strategy (or more than one strategy) to target both the long tail and the short tail. If you target just one of them, you will underperform.

Note: some people say that there are “only short tail verticals.” I tell these people that they just don’t have enough imagination and don’t look hard enough. Sometimes it’s not very obvious, but there is always a long tail.

Distribution

Distribution is the process of communicating our inventory to Google. This is so Google can fulfill its role as a content marketplace and match the user queries with suitable content, with our content as the inventory.

It can be quite technical, but at its simplest, it is a list of all your targeted page URLs. Your targeted page URLs are all the pages that you want to be found for (there will be much, much more about targeted pages later).

Most of the time, the list is in an XML format called “sitemap.xml.” You place all your targeted page URLs into a list (called stacking) and notify Google about the existence of this list. Google crawls the list, fills its discovery queue with these URLs, and then crawls your content.

If it fulfills a Google internal quality criteria, Google will even index your content (which means it will make a searchable copy of your content within their database). That’s what distribution is.

You can optimize your distribution in quite a lot of ways (again, more about this later). What is important is that you have a working, maintainable distribution logic to deliver your content to Google. If distribution does not work, SEO does not work. That’s why we have to make sure distribution works.

Note: do you really need a sitemap.xml? As Google is crawling you anyway? Simple answer is yes, you need a systematic distribution - a sitemap.xml logic. It is the only scalable way to tell if Google has crawled and indexed your content. If you don’t have it, your distribution pretty much depends on luck. And luck does not scale.

Traffic

After Google has indexed us, Google fulfills its role as a marketplace. It will then display our results on the search engine result page (S.E.R.P). The user does what we want them to do, and clicks on our result.

Yes, we get traffic. Job done! Well, not quite. This is actually where the fun (and work) starts.

The User Experience Feedback Loop

User Experience is measured over one search session. By sending traffic to our webproperty, Google has made a bet. Google bets that our content, our web page, was the right offer for the user’s search intent. And Google does check if its bet has paid off.

Google is unbelievably good in measuring their users’ user experience (UX), including the point they reach our site and beyond.

How does Google measure this? Only Google knows the exact details, but confirmed for more than a decade is the “Long Click.”

This means, if the user clicks on your search result, Google will send them to your site. However, if the user immediately returns to Google and clicks further down in the search results list, Google deems this a negative vote for your site.

Also, it is suspected that data for measuring user experience is drawn from:

• Google DNS
• Google Chrome (including desktop, Android, and iOS)
• Google Data Saver
• Android, ChromeOS, …

In its terms and conditions of the now deprecated Google Toolbar, it was even stated very visible that they track their users’ online behavior if a certain feature was enabled. Google has denied that Google Analytics and Google Adsense data is used to determine user experience signals for Google Search.

Fact is that we don’t know exactly how Google collects UX data, but they do, and they are magnificent at it.

It is surprisingly easy to create a website that attracts lots of traffic for a short time. It is much harder to create a webproperty that maintains this traffic and manages to improve on it. If the user experience is negative from Google’s point of view, you will not keep your traffic. Plus, there is also another kind of feedback loop.

The Brand Experience Feedback Loop

Positive brand experience from an SEO perspective means that:

• when results of your webproperty show up in Google, users prefer to click on them
• users search directly for your brand (i.e.: your domain name)
• users search for your brand together with some generic words (your domain and something else).

If you fulfill these criteria, then congratulations, you have successfully branded your users. User experience looks for the behavior of the user over only one single search session, whereas brand experience measures the impact on the user over time.

If users search directly for you or prefer to click results of your webproperty over the competition, then this is a very valuable vote for you. On the other hand, if users never search for your brand, or even worse, actively avoid clicking on you in search results, then this is a negative vote for you.

You have to “brand” the user as soon as they visit your site. It is crucial that:

• the goal of the users search (“job to be done”) was fulfilled
• they had a good user experience
• and they remember your brand (most of the time the name of your website and your domain) and want to visit you again.

It is quite easy for any halfway sophisticated SEO to create a website that attracts lots of traffic for a short amount of time. It takes hard work to create a webproperty that maintains this traffic and increases it.

If your brand experience is less than desirable and the users do not begin actively searching for you and clicking on your results, you will not keep nor increase your traffic in the long term.

Targeting

You will only be found for a certain user query if you have a page that wants to be found for that particular user query. It is a two-step process, and therefore you have to:

1. determine/guess/make a hypothesis about what the users might search for
2. create targeted pages for these searches

This process is called “targeting.” However, you should know there is a variation for when it is done editorially, and that is commonly referred to as “Keyword Research.”

We target a certain user query and the intent behind it, and create a “targeted” page for that user query and the intent behind it. Targeting can be a manual process or an automatic process, or something in between.

Regardless of how we do it, the goal is always the same: to create pages for stuff that people actually search for, might search for or will search for in the future.

Marketing

As we discussed earlier, it is important to remember that the term “marketing” should be used in the broadest way possible. It is everyone and everything that communicates your website to the public. With this in mind, we are now going to add two more arrows to our diagram: branding and inlinks.

Inlinks

Inlinks (also called inbound links, backlinks or simply links) are links pointing to your webproperty from other webproperties. They are still a factor for Google to determine site and page relevance. That is unfortunate, because it places a dependency on external factors, and must be considered a resource black hole.

Nevertheless, the main thing to remember is this: although you need inlinks, you don’t need a fixed number - not 1,000, not 10,000 nor 100,000. You need a healthy growth over time.

That means a variety of different domains must link - in a sensemaking way - to pages on your site. Never mind if it’s ten websites this month and only four the next.

What matters is growth over time and that many different websites link to yours. A site without a growth of inlinks will always underperform in the long term.

Link Building

The truth about link building is that 75% of all the resources you invest into it will be wasted. The pie chart below shows what happens to your link building investments.

You can see from the pie chart that 50% of everything you do will lead to nothing. That’s 50% of your crafted content that will not get any form of recognition in the form of links.

That’s 50% of all your emails that will go unanswered, 50% of all your pitches that will be ignored and 50% of your link bait stunts (that is, creating pages or web apps uniquely for the purpose of getting linked) that will not work at all.

Of the remaining 50%, half will lead to links that will have no positive impact on your website’s SEO performance. If you do it very badly (and the links are spammy), they might even have a negative impact. Therefore, you are left with 25%.

Two percent of that will be the “links you tell your grandmother about,” so for instance when:

• an international magazine covers you
• you were mentioned on the BBC (with a link of course)
• or the leading blog of your niche wrote a review about you.

These are highly valuable links - links that will have a direct impact on your website’s traffic. Today you may be performing ok, then the very next day you get a “link you tell your grandmother about” from a major source which will make your site perform much better.

The remaining 23% are the links that will have some positive impact on your site and help your overall growth over time.

But no matter how you look at it, 75% of all your invested resources into link building will be wasted.

There are tools out there that promise to tell the difference between “no or negative impact links and positive impact links.” They are useful for identifying those big impact “links you told your grandmother about” (which you already told your grandmother about). For classifying the remaining inlinks into links that have some value vs. those that don’t, these tools only work half of the time, so you might as well flip a coin.

So, what should you do? If you do link building for the sake of link building, the quality of your links will deteriorate quite quickly. At first, they will just not be sensemaking. Then they will become spammy, and then just plain spam. So don’t do link building! Do marketing, do PR, do business development as normal, but with links in mind!

You don’t need a link building team. Instead, aim to fill the wasted 75% of your resources with additional value. Make sure that every person that communicates in the name of your company is thinking about links too.

Public relations should think about links for example; all press releases should link to valuable, and relevant targeted pages. Don’t get me wrong; I do not recommend writing a press release for sharing links, but if you have a story to tell that might fit the format of a classic press release, make sure that you link back to the correct parts of your site.

Marketing, done right, has immense value even without any links (for example, building brand awareness). Therefore it’s great when you can boost its value by making sure there is at least something to be linked to.

Make sure your team is keeping an eye out for any opportunity to link to your site. In sales, your clients can link to you, ideally with an awesome recommendation. Even in business development or customer support, there will be opportunities if you are looking for them.

Making links part of the routine activity of business divisions, with their own quality standards, will ensure that the links remain “sensemaking” and are of a high quality.

Even with the business integrated approach I have suggested above, half of all the links you get will have no impact. But at least you will not squander your invested efforts as you will still receive value with these activities. Anyway, compared to what really matters (Online Brand Building, see next section) linkbuilding is just an annoying nuisance.

Brand Building

Any form of marketing can be used to push brand awareness. It is unbelievably important for sustainable SEO success that users start:

• to google your brand
• to google your brand with generic terms
• to prefer to click on your pages if they see them in the SERPs.

These points above can all be achieved by providing a good user experience (that is, fulfilling the user’s search intent and communicating your value clearly) once they’re on your site. But there’s more you can do.

It is quite common that following offline campaigns (for example, extensive TV ads) not only does the branded search traffic increase, but also the unbranded organic search traffic. After the advertising campaign is over, unbranded organic search traffic stays high while the branded search traffic goes down again. The same effect can also be seen to a lesser extent in radio, print and billboard campaigns.

Google has repeatedly stated that paid search listings - Adwords - do not have a direct positive impact on organic, unpaid traffic.

However, as any good SEA (Search Engine Advertising person) will tell you, it is “possible” that if you have an extensive AdWords campaign your organic traffic - assuming you have an SEO optimized platform - also rises. After having observed and tested this myself multiple times, my conclusion is that this is due to the brand effect.

Similarly, Google states that their ad network, AdSense, does not have a direct positive impact on organic search performance. But if you use AdSense as a retargeting channel, so basically stalking the users with your brand throughout the web after they have visited you once, it is an excellent way to remind users of your brand.

However, be very careful, because retargeting can quickly become very annoying (and this, in turn, is negative branding).

Also, social media is a great way to build a brand. It can be another great channel for businesses. If you are just starting out, it’s sometimes even faster and easier to gain users via social media rather than SEO.

I strongly recommend for everyone to have a social media strategy as well as a search strategy, but social media is not SEO! Social media is a great way to build a fanbase and position your brand.

If they like you - and follow you - on Facebook, there is a higher possibility that they will click on your site when they see it in their search results. They might actually google your website once in a while.

However, whenever you do social media - and you should - don’t do it just because of SEO; social media is a channel in its own right.

What You Need To Focus On

It’s essential you focus on things that are within your direct control. Let’s discuss the diagram below.

Everything within the trapezoid is within your control. That includes:

• the user experience and brand experience once the user reaches your site
• the pages of your site and how they interlink with each other
• what pages you have and what pages you don’t have
• what queries you target with what pages
• the distribution until the sitemap.xml

My recommendation is to focus strongly on the things within your control. 100% of everything you invest in the area within the trapezoid will be 100% efficient, minus only your own internal overhead.

Google is not within your control. You cannot tell Google what to do. So don’t focus on Google. Google is a black box; you cannot 100% understand Google nor can you reverse engineer Google.

What you can focus on is what and how you distribute to Google (quantity and quality wise). You can make educated estimates on what the users demand (as in, search demand).

The user is also not within your control. You cannot force the user to use the words you want them to use. Yes, it’s true you can influence them via brand building, but that is far from controlling the user. However, the queries you choose to target are within your control.

Inlinks are not within your direct control. That is work that somebody else, another webproperty, has to do. You can influence it, but you can’t do it yourself.

(Technically, you could build some kind of link farm, with millions of pages with hundreds of different domains within your control - all interlinking each other. But let’s just say it doesn’t work anymore and will get you kicked out of Google in no time).

Feedback

Now that you know what you can control, remember that getting feedback is critical to success. Previously, I mentioned two major things I want you to take away from this book. One of them is: if you do something be sure to get some feedback.

We can also turn this sentence around: if there is no form of feedback available, then don’t do it.

Thankfully there are a lot of avenues to access SEO feedback, and it’s all readily attainable with free tools. Let’s look at a few of the major feedback and analysis tools that will help you optimize your pages and strategies.

Google Analytics (GA) or another easily accessible analytics solution is essential for your webproperty. Easily accessible means that you - or anyone who has anything to do with SEO (development, marketing, content) can access the data without too much hassle and can easily make a meaningful analysis of this data, even without in-depth training. If you have an analytics solution that

1. does not work
2. only a select few can access
3. that people hate

it does not count as “easily accessible analytics solution.” Google Analytics is free and easy to configure, has a huge, and mostly helpful community and I highly recommend implementing it.

Google Search Console (in short: GSC) can unearth 80% of everything that is wrong (or right) from an SEO perspective with your webproperty. GSC does not need installation; you just need to verify it on your site. (Simply google [Google Search Console] and follow the instructions on how to verify it.) No paid tools can come close to the insights available in GSC.

Important note: Add all protocol domain variations to GSC. That means if your site is https://www.example.com you should add the following to GSC:

• https://www.example.com
• https://example.com
• http://www.example.com
• http://example.com

Protocol and (Sub)Domain SEO issues are quite common (for example, you get pages of the www-subdomain indexed on the non-www-subdomain version), and GSC provides an easy way to find these errors. GSC is the most important SEO feedback tool.

I will often refer to GSC, but I will not explain each and every metric as this would be a book on its own (If “Google Search Console - The Missing Manual” existed, I would buy it). Until then, it’s up to you to check out GSC and whenever you have a question, just R.T.F.M. (it’s usually hidden in GSC reports behind a circled “?”).

Google Search Console has various metrics that relate to distribution, for example:

• submitted pages via sitemap.xml
• indexed pages from sitemap.xml
• indexed pages overall
• crawling stats, and more.

If you want to know if your distribution is working, just check GSC. What’s not in there is the logic that you use to decide what you list in the sitemap.xml. That must be controlled via some internal logic (for example, custom logfiles or just looking into the xml files).

I.e.: what queries you show up for in Google (otherwise known as impressions) and what pages you show up for (top pages) is listed in GSC. The impressions that lead to clicks are also listed in GSC. You can filter those impressions and clicks for countries, devices, search type and more.

The user experience for users coming from search can be measured via Google Analytics. What the users do on the page and if their search goal is fulfilled can also be analyzed with GA. If people actively google your brand and preferably click on your results, GSC covers you.

A lot of onpage feedback can be found in GSC, as well as a lot of onsite stuff (for example, most internal linked pages, duplicate title tags, and so on). You can check in GSC how many links your site has and also get some great link data via GA referrers.

If you want to identify the terms the user searches for and for which you do not show up yet, you won’t have much luck with GA and GSC but try:

• Google Autocomplete / Google Suggest
• Google Trends
• Google Adwords Keyword Planner

The data from the tools mentioned above can help you identify what new targeted pages you should create.

If not everything of what I mentioned here makes sense yet, don’t worry. It will - by the end of this book. Right now, the only thing I want to make clear, is that there is always data and the possibility to check the impact of your actions. And you really need to look. Not regularly looking at GA and GSC is like driving with your eyes closed.

It’s A Model!

The diagram above is not SEO. It’s a model, a framework, a way to think about SEO. And it’s damn useful! (SEO is what you literally do.) It’s the way I propose you think about SEO for your business. Why? It’s measurable and manageable.

It shows what impacts what and gives you a lot of leverage points. It’s relatively straightforward (all models are simplifications). Everything that has to do with SEO has a place somewhere in this diagram. If you cannot pinpoint it in there, it doesn’t have anything to do with SEO.

Note: If you disagree, please come up with a better model. That is called progress, and I’m looking forward to it.

Common Mistake: Reverse Engineering Google

Do not try to reverse engineer Google. You will not succeed. Even if we were to take everything we know about Google and make estimated guesses about everything we are aware that we do not know, we would still not have anything even approaching the complexity of Google. There is just too much that we don’t know that we don’t know! And it continuously grows.

Let’s go even further and imagine we get our hands on the source code of Google, and we implement what we found, we would still not have Google. We do not have the same underlying data, nor the same index, nor the same user search history as Google does. We will never 100% understand Google.

You cannot understand Google, but you can understand search, and if you understand search you can understand SEO.

There are a multitude of blogs and “experts” who claim they figured out:

• Mayday
• Caffeine
• Panda
• Panda 2.0
• Magic Unicorn
• Venice
• Penguin
• or any other stupid animal name Google uses for their publicly announced algorithm updates.

These so-called “SEO experts” don’t know what they don’t know. Additionally, nearly all SEO algorithm updates are not directly actionable for webmasters as we cannot change how Google works. So more or less just ignore them.

I say more or less because each event (algorithm update) might be part of a trend. Sometimes such a trend is quite a good reason to change how we interact with Google and the users (for example, mobile, speed, HTTPS). But if Google does this, Google will communicate this, so again, you can (must!) ignore all the SEO blog posts by “SEO experts.”

Tip: don‘t look for stories or explanations, look for trends.

I recommend that you read every new announcement, mostly on

• developers.google.com
• support.google.com > search
• webmasters.googleblog.com

Then:

• For heaven’s sake, do not read any SEO blogs about the same topic
• Do nothing! Don’t overreact based on a single event
• Look for trends (in your traffic, in SERPs, in future or past announcements)
• Think about what you could do better based on the trends and the new information
• Do better.

I could go on about this topic for hours and hours, or pages and pages. It would be an entertaining read but not very valuable for you. You just need to remember:

• Google is a marketplace
• You cannot reverse engineer the marketplace
• You can, however, look at what goes in (search demand, your inventory) and what comes out (SERPs, traffic)
• Algorithm updates are not immediately actionable
• Look for trends, and…

Don’t read SEO blogs!