Modern CMSs (Content Management Systems) facilitate the publication and operation of large websites, but facilitate drop-mining for web-site duplication.
A particular page under one domain may exist in the form of many different urls. For example:
Url with tracking parameter: https://www.intersport.no/winter/langren/skisko?utm_source=newsletter
Url with filtering parameter (by size): https://www.intersport.no/winter/longhorn/skisco?order=price&dir=asc
Print-friendly page: http://www.oslo-ethickel.no/print_bike.php?id=44
So to use the shoe page of Intersport as an example, here we have 7 filter options. As these can be combined completely at their own discretion, we get thousands of different versions of the same page, but marginally different content. When to search engines crawler Your pages will be automatically linked, filter options and browse the pages (page 2, page 3). Here you risk Google indexing thousands of versions of the same page, and it complicateds the process of retrieving the correct page for a particular search for Google.
In most cases it is desirable to: one version of a particular page should be indexed by Google!
When users link to different versions of the same page, and when user and authority signals are watered out across the many versions, the real SEO value of the page will sink dramatically. If all of your pages have duplication problems, they rarely achieve the potential needed to rank on Google's first page. When between 20 and 60% of web pages' average traffic comes from organic traffic, this means that traffic and sales are largely lost.
What is the solution? Canonical is a function contained in the source code. In short, this Google tells which version of the page this is considered to be.
When cannonical is correctly implemented, and Google comes over a version of the Intersport ski page with several enabled filter options (intersport.no/winter/longren/skisko?nivaa-564=racing&dir=asc&order=price) (either through trawling or someone linked to it) Google will understand that this is only a duplicate version of the “parent side” (intersport.no/winter/longhorn/skishoe).
Example of proper implementation of the canonical ski side with enabled filter options:
Exceptions and special cases: Before you activate this feature, or ask your developer to handle it, some operating rules should be set and some balancing actions made:
If the “mark” is a filter choice on a category, you need to consider whether the label page of a particular category should exist in Google’s eyes, such as “Alpina skis”. If you want Google to index it, canonical must tell that the page is not another version of the shoe page, but a version of the shoe+alpina page.
A central question when deciding which pages should be indexed or one is whether the characteristics and content of the filter options represent actual searchable phrases. Based on intuitive assumptions, and insight from Google Keywords Planner, we know that “Alpina skis” is a widely used search phrase (390 average monthly search).
NB: When the selection of tags becomes more common, such as the Alpine + Alpha, these should not be indexed since very few people are looking for ‘alpine and alpha shoes’. The same applies to sizes – few people googler to specific sizes.