The engine becomes increasingly punitive in its desire to fight against spam...
It goes up to penalize sites for links from other sites, although the webmaster has nothing to do with them. It seems that one expects from webmasters that they conform to a mold for which the rules are unclear and with the disadvantage that the advice given to them have little to do with what we find in the pages of results.
(See update below)
I often hear webmasters complaining that sites topped the list of results, just thanks to keywords contained in the text, with no original content. The logic is that a site that is called Blue-Danube is more relevant to talk about the Danube than if it was called, for example, MySite.
When Google itself has created a website for its search engine, it has not called it "Search Engine", they found a more original name. It did not even was afraid of the presence of spelling mistakes as the name chosen would have been Gogol.
When a user does a search on the rivers, should he go on Amazon.com? And are only sandy beaches lovers going on eBay?
It could be given to ordinary webmaster the same freedom as the creators of these sites which become big: he has the right also to choose the domain name he like.
Update onr July 27, 2015:
In a post entitled Google's handling of new top level domains the firm said that henceforth the keywords in domain names are no longer taken into account. Domains and their extensions (even the latest) are treated equally.
That is a factor that Google is struggling to decide. To vary the types of results it was decided to place only one link to the same site in the first page of results. Then they changed their mind and now result pages are populated with links to the same domain and more recently it has decided to reduce the number of links to the same site. As a user, I have never enjoyed the choice reduced and be led insistently to the same destination. It is one thing to be offered a choice of sites and another being pushed by force to the same site. This is not the same experience.
Surely there are complicated arguments to base this decision. Maybe through the promotion of links on Amazon, the engine wants to discourage buyers to search directly on the site? But that is what will happen if we continue on this path, because finding only links to big websites will lead to replace the engine with a table of icons representing these big sites.
Then there is this Panda algorithm that is supposed to distinguish between quality pages and others, which are penalized. Remains to be seen what is a page quality.
I guess Panda has produced the desired result and I search for "Search engines" (this was on the french version), in private mode so as not to influence the results with my habits. I get the usual Wikipedia page (Like I need to go to Google to access Wikipedia), then a second Wikipedia page (we never get tired) finally a page to that big French site that is on a mission to make a post on any subject on which it knows nothing, and which corresponds to the definition of a content farm. The link is to a completely hollow article with all the keywords to be in the results pages of Google. The article begins with an information that I really need:
"There is a huge amount of information on the Internet."
Apparently there is no more serious item that has ever been written to describe a search engine, a page like this, Anatomy of the Google search engine is less interesting than a page crammed with ads and popups and which provides this new trick:
"When the user performs a search, he enters in the text field the words he seeks."
How the user is he supposed to have made to come on this page?
In short, Google does not succeed to provide a useful result of this request, we have the sad impression that it forgot to tell to the algorithm what is a search engine!
According to the spokesman for Google - I do not know if it's a coincidence that it is the head of the spam that is always open to webmasters - what matters today is the user satisfaction. This is apparently the big trend. Finally when a site sees its content fully copied by another which passes it in the SERPs, this is his fault, customers prefer the copy to the original.
In the same vein, if a store is robbed of its stock of Rolex watches and they are sold on the street for 50 €, what matters is that the customer is happy.
But Google is not wrong to say that we must generate enthusiasm in the reader. This is something that happens from time to time to me. Tweets and links on Facebook and other social sites have brought me more than 30,000 visits in one day. At the same time, Google brought me three or four visits. The advantage when we generate excitement is that we no longer needs to search engines to get a wide audience!
In his quest to eliminate intermediaries - our websites - and provide direct answers to users, SERPs now display a card on the main keyword of the query. It's convenient and it shows that it is possible to skip the search engine. These data are taken from Wikipedia, the content is free and the program needed for this kind of MSDS is not very complicated, so anyone can create his own knowledge graphs.
The problem with the idea of Google to provide a direct response to user is that it does not have the answers, these are the sites that have them! This corresponds to the definition of Web scrapping: retrieving through a script the contents of a page to reuse it in another site...
It seems that the developers of the Yandex search engine have understood this problem as they offer an alternative that will appeal to webmasters: they build cards themselves with the content of their site and snippets of internal research and they propose them to Yandex so that it provides them to users. This is the Islands service. Their sites are highlighted instead of being short-circuited by Google.
Or can we replace the search itself? Rather than relying on a robot that makes the law from a remote and pervasive website, we would have our own robot which collects the information on our behalf. Since each search drive to Wikipedia, the robot itself can access the site and find the information. The same with all the major sites that monopolize the top positions of the SERPs. For the rest, it can use the Common Crawl database and for selection, to rely on the content more effectively than does the algorithm that relies solely on accessories signals and not on the content. It is easier for the user on the one hand to interact with the robot, on the other hand eliminate spam sites for future research.
Such a robot pursuing its natural evolution will go where a search engine will never go. Instead of waiting for you to ask a question to answer, it will look for information that may be useful to you at all times, produce lists or collections of data for you and be at your disposal. Eventually it will alert you when it has useful information.
The number of avenues to explore to find the information is actually unlimited. But how to make know these new tools? Why not a good conventional search engine that gives only links to sites?