Ignored by most businesses for some time now, web analytics continues to gain more and more power and interest in the eyes of business managers who want to optimize their web marketing performances. While a number of solutions (free or otherwise) are available to analyze your website’s data, the terms used and their definitions may vary, creating certain misunderstandings. This article aims to explain the basic concepts and misunderstandings that can lead to certain wrongful interpretations.
Web analytics: bounce rate is an incomplete tool
Remember that the bounce rate of a website represents the percentage of your visitors who arrive on your site and then leave without visiting your other pages. This indicator is regularly used to determine if the landing page is relevant to visitor queries on search engines.
Indeed, if you have a bounce rate of 95% this may mean that your visitors arrive on your website to almost immediately leave the page as the content you offer doesn’t correspond to their searches. Therefore, it’s your job to identify the pertinent relationship between your landing page and Internet user search results.
On the other hand, it can also mean totally the opposite! Let me explain. If your bounce rate is 95% but the average time per visit is 2 minutes, it means that your visitors find exactly what they are looking for and so do not need to click elsewhere on your site. As I’m sure you can understand, analyzing only the bounce rate is not enough to draw a good conclusion regarding your website’s performance. So make sure to add another parameter such as the average duration per visit or percentage of your pages that have been browsed.
It is quite possible to analyze all the actions carried out by your visitors on a given page. Some (paid) tools such as Crazy Egg or ClickTale allow you to see all the clicks made by your visitors, including ” dead clicks ” (clicks made on a zone without a hyperlink). You can also set up Google Analytics to generate an “event” to alert you each time a visitor has browsed 10%, 20%, 50% (etc.) of your site pages.
The colored sections above represent user clicks that have been registered. Each color represents a type of action performed by your visitors. This analysis is a great way to test different page scenarios with links that are placed at specific locations. You will also know which links are clicked on most by users.
A page view is not always a page viewed
Again, misinterpretation may cause confusion if you have not defined the terms analyzed beforehand. Since the number of pages viewed may vary from one analyst to another, the results announced can sometimes be different and this may cause misunderstandings. For example, it is quite possible to consider a pop-up that appears before the display of a page or the images as a page view. A page with a pop-up and five images will thus be able to count as six page views (the pop-up and the five images) while other analysts may consider this to be one page viewed.
Another example: A visitor who arrives on your site clicks on a link then returns to the previous page. Should one or two page views be considered? Again, this will depend as the results announced may vary from one to two.
Also note that the loading time of your pages will also influence your results. Indeed, the tool will not necessarily take into account the page view if the visitor leaves it without it having had time to fully load.
The difference between a page view and a visit
It is not uncommon for people to confuse visits and page views. A visit corresponds to one or more actions carried out by the visitor on the website (clicks, page views, making contact). It is fairly common to define a time interval more or less equal to 30 minutes beyond which the visit will be considered complete if the visitor has remained inactive. A page view is simply a page that has been visited by the visitor and has had time to load completely.
As explained above, a visit can have several pages viewed and all of your page views will not necessarily appear in your reports if the loading time is too long.
A visitor is not a person
The classic error made by most beginners in analytics is to assimilate their number of visitors to the number of people who have visited their website. In this case, it is not uncommon for some managers to say that their site attracted more than 800,000 unique visitors, which is equivalent to half the city of Montreal (for example). Why is this statement a big mistake? Simply because you should not compare what is not comparable.
Indeed, many people can visit a website by sharing the same screen and no tool will tell you how many people actually saw your content during this visit. Conversely, the same visitor can easily consult your website using a computer and then a smartphone and then a tablet.
Your tool will therefore give you three unique visitors for this situation, when in reality it is actually the same person just on different devices. Finally, if a person uses multiple browsers (Google Chrome, Internet Explorer, Mozilla etc.), your analytics tool (Google Analytics or other) will also consider each browser as a new visitor.
As you have discovered by reading this article, it is very easy to make the numbers speak to your advantage and greatly inflate the performance of a website. Therefore, you must be sure to define each analytic term/notion before you share your results; also, try to use the right terms when you talk with your analyst or client. You will avoid misunderstandings and surprises that arise from a lack of clarity.
If you liked this article, you’ll probably like this one too: Program events on Google Analytics