Sur d’autres sites (8046)

• How to analyse 404 pages

  1er juillet 2019, par Matomo Core Team — Development, Plugins

  How to analyse “not found” pages (404) in digital analytics

  Have you ever sent out a newsletter and one link wasn’t active yet ? Would you like to know how many users get affected when this happens ? Would you like to know if your visitors are encountering 404 pages ? 

  In this article we’re describing an easy way to analyse “not found” pages on your website with Matomo to increase your visitors’ user experience, user acquisition, and SEO (search engine optimization).

  How to know the number of 404s on my website ?

  There are different ways to get this information. Depending on how your website is built, you may or may not collect this data.

  The easiest way to answer this question is to fire a 404 page on your website, you do this by accessing a wrong url :

  

  As you can see here, in our case, the page title starts with “Page non trouvée” which stands for “Page not found” when translated in English (as the website we are considering here is in French) :

  

  In this example 19 page views have been fired and it generated a bounce rate of 67%. As a result ⅔ of the visits ended here.

  In some cases, the information related to a “not found” page can be found either within the title or within the URL, as some websites redirect you to a specific web page when a page can’t be found.

  If you can’t identify “not found” pages via a page title or a page URL, we strongly advise you to use this specific tracking code method on your 404 page : “How to track error pages in Matomo ?”. 

  You can easily set it with Matomo Tag Manager with a custom HTML tag :

  

  where the trigger is the following :

  

  You will however, have to define this trigger as an exclusion for all the other tags which may conflict with it (here below is the new trigger defined for the generic Matomo tags we are inserting on all pages) :

  

  Once this specific tracking is set, you will be able to track the source of the 404 and will gather all the “not found” pages in a specific group within your Page Title report :

  

  Here, for example, you can identify that the homepage of this website had a link pointing to a 404, in our case it was https://www.webassoc.org/pro-du-web.

  Note that this is just one technique. You could also create a custom dimension report and decide to send the 404 there also.

  How to get notified when a 404 page is visited ?

  Trust us, you’re not going to check everyday whether a 404 page has been visited. In order to avoid checking it manually, you can define custom alerts.

  There are three possible scenarios when “not found” pages can be fired :

  • internal 404 : one link within your website is pointing to a wrong url on the same website.
  • external 404 : someone from an external website made a link to yours and the link is not correct.
  • direct access 404 : someone access directly to a not found page on your website.

  You can define all those three within Matomo, but in your case, you will only have to focus on the first two only. In fact, you can’t really fix the third scenario. That’s the reason why we’re not focusing on it. It would result in irrelevant alerts.

  Custom alert for internal 404

  An internal 404 is defined from a 404 where the source is an internal web page. As a result, it will look like the following in your report :

  

  In this example, we’re using this specific custom implementation, the title of the page will contain “From = https://www.webassoc.org/”. So set our custom alert accordingly :

  

  Now every time a 404 page will be fired from an internal page, you’ll be notified by email.

  Note that you can also decide to not receive any email and track the evolution of alerts with the History of triggered alerts feature.

  Custom alert for external 404

  External 404 is almost the same setup. The only thing you need to keep in mind is that we want to exclude the 404 where the source is not indicated. As a result, your configuration will look like the following :

  

  Here your regular expression pattern is the following one :

  404/URL = .*From = (?!https://www.webassoc.org)[^\s]+

  as you’ll want to have any referrer coming from a website which is not Matomo and not a direct 404.

   

  You can now be notified every time that a 404 is fired from any link.

  Note that this configuration may slightly differ from website to website. So always double check your tracking code and the way the values are sent to your reports. Also try to trigger those alerts first before validating them.

  How to follow the evolution of your 404 over time ?

  It may be interesting to know how good or how bad you are performing in terms of 404.

  In order to check this information, you can click on the evolution icon near the 404 title :

  

  But you may be interested in accessing this information more regularly without having to create this report each time.

  So, one way to analyse the evolution of your 404 is to create a segment such as :

  

  and to click after that on evolution icon :

  

  As you can see below the number of “not found” pages is quite low in general, but we can also notice that a period received an increase in terms of 404 not found pages on May 27. It may be interesting to investigate it :

  

  You can start from the overview of referrers :

  

  As you can notice here the main source of 404 is coming from direct entries which is the most difficult channel to analyse as we don’t really know where the visitors are coming from.

  How to perform your analysis even faster ?

  As you can see analysing reports in Matomo in order to detect 404 pages is a time-consuming activity. In order to make it faster, you can already create a report about it within the Email reports feature with the following settings :

  • Segment : 404
  • Email schedule : never.
  • Visits summary and Page titles as selected report.

  You will then end up with a saved report listing all the URLs concerned :

  

  You can also have a look at the “Custom reports” premium feature.

  It will provide you with more flexibility. You will then be able to focus on the most important thing : the cause of 404.

  Good luck and happy analytics !

• Today we celebrate Data Privacy Day 2019

  28 janvier 2019, par Jake Thornton — Privacy
  Today we celebrate Data Privacy Day 2019 !!!

  What is Data Privacy Day ?

  Wikipedia tells us that : The purpose of Data Privacy Day is to raise awareness and promote privacy and data protection best practices.

  Our personal data is our online identity. When you think what personal data means – our phone records, credit card transactions, GPS position, IP addresses, browsing history and so much more. All so valuable and personal to us as human beings.

  That’s why we cannot take our personal data online for granted. We have a right to know which websites collect our data and how it’s then used, something that’s often not visible or easily recognisable when browsing.

  What Data Privacy Day means to Matomo

  Every year the team at Matomo uses this day as a chance to reflect on how far the Matomo (formerly Piwik) project has come. But then also reflect how far we still have to go in spreading the message that our data and personal information online matters.

  2018 saw the introduction of the EU General Data Protection Regulation (GDPR) to protect people’s data online. As a team, Matomo was at the forefront of this development in the analytics space and have since built a GDPR Manager to ensure our users can be fully compliant with the GDPR.

  With every new release of Matomo, we are ensuring that security continues to be at the highest standard and we will continue to be committed to our bug bounty program. Our most recent release of Matomo 3.8.0 alone added a Two Factor Authentication (2FA) feature and a password brute force prevention.

  What next for Matomo and data privacy ?

  As always, security is a top priority for every new release of Matomo and continues to only get better and better. We have a duty to spread our message further that the protection of personal data matters and today is a vital reminder of that. We are, and forever will be, the #1 open-source (and free to use) web analytics platform in the world that fully respects user privacy and gives our users 100% data ownership.

  In 2018 we changed our name, we updated our logo and website, and advanced our platform to compete with the most powerful web analytics tools in the world, all so we can spread our message further and continue our mission.

  Come with us on this exciting journey. Now is the time to take back control of your data and let’s continue creating a safer web for everyone.

  Please help us spread this message.

• Node.js asynchronous video conversion slow

  9 décembre 2016, par lukstei

  I wrote a little website/service, which can download a video from a website (currently Youtube) and converts it on the fly to an mp3 file and sends this file back as the response.

  For example, you when you request http://localhost:8000/v=http ://www.youtube.com/watch?v=HhoewflkQu0, then it will download this video and response the audio layer encoded in MP3.

  This all works very well, my problem is that this is very slow and I can’t figure out why.

  ---

  Simplified the script behaves like this :

  Download the video and write it to the stdin of ffmpeg, and the stdout goes to the response.
  Video (MP4, FLV) -> FFMPEG -> MP3

  I used curl to figure out how fast the script is :

  $ curl http://localhost:8000/v=http://www.youtube.com/watch?v=HhoewflkQu0

  I get only about 5-10k.

  So why is this so slow ?

  1. The server, from which I am downloading the video is slow.
  2. The conversion is slow (because of a slow CPU).
  3. The data transfer between node.js -> FFMPEG is slow.

  I tried to download the video in a normal download manager, and i got about 320k, which is my normal download speed, so the first point isn’t the bottleneck.

  To point 2 and 3, I tried to write a local file to the stdin, and I got about 600k so that isn’t it either.

  So why is my script so slow, and what can I do to make it faster ?

  https://gist.github.com/1304637

  Thanks in advance.