SEO: Technical introduction to SEO
This article mainly concerns Google.
What does Google want?
Show the most relevant answer possible for a given query, it is a search engine more than a search engine.
Some definitions
The crawl and indexing are two distinct concepts and yet confused by many developers not familiar with the field of SEO.
The crawl
The crawl is the process by which the robots of the engines roam a site, reading the code of the pages. Roughly, it's as if you were "viewing the source" of each page of your site, and then you follow each link on the page. The pages browsed by robots are normally controllable by the robots.txt file at the root of your site.
Tough to reproduce by hand, there are many tools to allow you to automate this step: thanks to its availability on Linux, I use Screaming Frog , but BeamUsUp is also available, and I also recommend the Botify online crawler. You will discover how the engines technically perceive your site and you will raise a number of problems associated with the "technical" part below. Incidentally, you'll find that to ensure the way that Google perceives your site, many tools make the analysis of logs rather than the standard crawl.
Warning: the engines often allocate a fixed time to the course of your site, hence the importance of performance, so that your site is browsed more in-depth during the passage of the robot (see Crawl Optimization for the details of this statement) .
indexing
The index is simply the fact that a page appears in the engine index. We can "control" this indexing by the META Robots, which can take for example the value "NOINDEX" to ask the engines not to display this page in their search results.
The difference
Everything is possible in SEO: crawling a page and not indexing it (META ROBOTS NOINDEX), but also not crawling a page (blocked by the robots.txt file) and yet indexing it, because for example many links external links to this page. You will need to remove this page from the robots.txt and put a META ROBOTS NOINDEX so that the robot goes through your page, and deindexes it.
The basics of SEO
The basics of SEO are often presented as a pyramid (type pyramid SEO in Google Images, you will not be disappointed), but not always in the same order, so I'm just going to talk about several poles on which to focus: technique, content, and popularity (or links).
The technique
The technique is what will allow search engines to understand your content. Entirely based on the HTML language, some tags will allow your site to give specific information to search engines, but also to social networks, about a page of your site.
For example, the "TITLE" tag in the "HEAD" of the HTML represents the main text (the one that contains the link) displayed in the search engines in response to a request made. This is also the text that often appears in the header bar of your browser (or the current tab). This is often the same as the main title of the page (tag H1), but nothing requires it.
The meta tag "Description" represents the text displayed next in the search results (the famous "SERP": "Search Engine Result Pages").
To see what a page of your site looks like in search results, just type "site: your-url" in Google.
META tags
There are many meta tags. I have already talked about the most common (Description, Robots), but there are many others, some specific to engines, others to social networks.
For example, the "twitter: card" meta tags allow Twitter to display additional information about a given tweet.
Facebook and Google+ use their OpenGraph METAs, starting with og (like OpenGraph). Incidentally, know that you can refresh the Facebook cache on a page of your site by using the open graph debugger page.
We note here that behind the term "generic" SEO are actually many techniques to increase the number of visitors to your site, sometimes even unrelated to search engines (sometimes referred to as Social Media Optimization ). As everything revolves around the presentation of your content for robots, these areas are however related.
Duplicate content and CANONICAL tag
There are cases where you will not be able to prevent duplicate content on your site. In this case, Google's recommendation is to use a Canonical tag that tells Google the "main" version of the current content. You will find on https://en.onpage.org/blog/canonical-guide-dos-and-donts a summary of what to do or not with this tag.
Attention, all this is only an indication for engines that treat the tag as they see fit (as often in SEO).
The sitemap.xml file
There is often one (or more) sitemap.xml file (s) at the root of the site whose function is to list the pages of your site to guide the crawl of the engines. Again, this file is only an indication and the engines do what they want.
In general, I share the conclusions of this article and consider that for a "small" site (less than 10 000 pages), this file is useless and robots are more efficient, provided that your site is correctly designed in terms of 'tree. Beyond that, the file can add value by targeting the robot's work.
Finally, sitemap files can also be useful by differentiating the content they contain (one for news, one for videos, ...).
The performance
Performance is important for the search engine (crawling optimization), but also for visitors to your site (which will come back more naturally if they do not have to wait too long for each page load). It would be too long and impossible to cover all the points related to the performance of a site (moreover, the subject is very technical), so refer to our articles on the subject to identify points to work or ask we conduct a performance audit if needed.
The contents
"Following best practices often does not work for new and emerging sites." ( Rand Fishkin in 5 seo problems )
If you follow all the good practices described in the previous section, you will have a site properly understood by the search engines. These best practices are being followed by most CMS today, including Drupal and Plone, which we use on a daily basis.
However, without interesting content, your site will not necessarily be positioned in the first positions of the engines (unless you are on a sector not at all competitive or a clean site will suffice). Indeed, it is the interest of your content that will (re) come the visitors and will generate external links of quality to your site, and therefore opportunities to be properly positioned.
In addition to the quality of the content, the organization of this content in the site is also important: you can "push" some pages of your site by adopting appropriate site hierarchical structures . An internal structure that is much talked about these days is the semantic cocoon , evolution of the silo concept .
Without going further in the "content" part of your site (which is more marketing and for which I am not a specialist), the consistency in the choice of keywords used in your text is a first step to increase your chances SEO on a term, but today we are also playing on semantic proximities, and the writing of content becomes a real science.
There are also analytical tools to evaluate which keywords you are positioned on: alyze , semrush , .... All this will help you, but nothing replaces quality content (both for search engines, but especially for your visitors ;-)).
Popularity
There, no secret, you need links (and good ones). It also requires a lot of work, domains that make links to your site have criteria of trust by the engines that make all links are not equal. But without popularity of your content, there is little chance to shine there too. There would also be a lot to say here too, but we are leaving purely technical criteria and I will not go further (and again, the quality of the content is paramount).
Several types of answers
80% of the clicks are caused on the first 3 links returned by Google: more or less 60% on the first, 15% on the second, and it goes down quickly then (hence the importance of being in the first results ).
But according to the requests, Google returns different types of answers, sometimes including videos, slideshows, a map. It is therefore possible to position yourself differently than on the first 3 "classic" results.
Comments
Post a Comment