Stop Panicking Over Bounce Rate: The Myths, The Math, and The Truth
Poor bounce rate, it turned into the metric that everyone learned to hate without really knowing why.
Why did we learn to hate Bounce Rate in SEO?
It’s true that, sometimes, a high bounce rate means that something needs to be improved. But, in many cases, this metric has been associated with a lot of misconceptions. Including that, many people think Google uses this data to assess the quality of a website.
What is Bounce Rate?
According to Wikipedia:
Bounce rate (sometimes confused with exit rate) is an Internet marketing term used in web traffic analysis. It represents the percentage of visitors who enter the site and “bounce” (leave the site) rather than continue viewing other pages within the same site.
The definition is quite simple. However, it’s important to analyse data in detail and consider context:
Visitors can arrive to the URL via direct visit, organic traffic, referral, etc.;
By default, Google Analytics can only calculate the time a user was on a page, when the user visits another page within the same domain;
A visit by a user can generate a bounce when, after arriving on a page, the user performs one of the following actions:
Clicks on a link that takes him to a different domain or hostname (depending how your Analytics is configured, e.g. if you want to include subdomains);
Clicks the “back button” in the browser;
Closes the browser, or open tab;
Navigates to a different URL outside the domain, or performs a new search in the browser;
Becomes inactive and the session expires (in Google Analytics, sessions last ~30 minutes);
But why did we learn to hate Bounce Rate in SEO? There are some misconceptions that help to the confusion.
Misconception #1: Bounce is a ranking factor
One of the many myths people still stubbornly believe, is that Google needs to use Google Analytics data. More specifically, bounce rate data, as ranking factor. This is totally wrong!
Google does not use, or need to use, data from a service whose data can be manipulated. Yes, you can manipulate Google Analytics data so it appears to decrease or increase whatever you want. If Google used this data, it would have a very low trust. Not to mention, it would be entering other fields related to privacy. Data from Google Analytics exists so you can understand a little better your visitors. Google does not use or need to use this data!
Misconception #2: Bounce rate is the same as a short click
Then, there is the belief that bounce rate and short clicks are the same metric. For those who never heard of it, it’s a metric used by search engines, designed to evaluate how a user clicked on a result and how fast he immediately returned to the search results. Google doesn’t need Google Analytics to assess the quality of a website. Google can see directly the behaviour of each user right there in search results.
To be able to get an approximate view of short clicks in Google Analytics, you would have to isolate traffic coming from organic search, where the user entered the site and immediately (a matter of seconds) clicked the “back” button in the browser. Only this way, would it be possible to get a rough simulation of what search engines see through the behaviour of users in their search results pages. Considering that, short clicks and bounce rate are the same thing, is sheer lack of information. But also a reflection of what the market does very often; following what everyone says without stopping to think.
Disclaimer: This is my take and personal opinion and I don’t speak for Google regarding ranking factors and metrics.
Bounce Rate and search intent
People have different intentions when they search on a search engine. Whomever works with search engines for some time, surely knows that there are four types of search intent:
Informational: When we want some information about something, eg [speed of light] is an informational query, the person who performs this kind of search probably wants to know more about the subject;
Transactional: When we buy something, eg [Nikon D300 price], a transactional query usually includes terms like: cheap, price, online, etc;
Navigational: When we want to find a specific site, for example [amazon] or [apple];
Local: Nowadays, thanks to smartphones, we can even include a fourth type of search intent. The local search. For example [Pizzaria in San Francisco] is a query with strong local intention.
Obviously, bounce rate differs according to the type of site that is associated with the corresponding type of query. Some examples:
Portals/Social Networks: can have ~10-30% bounce;
E-commerce: can have ~20-40% bounce;
Blog/Content sites: can have ~60-70% bounce;
Landing pages: can range ~70-90% bounce;
In order to understand if bounce rate is or isn’t something to be concerned about, we must look carefully and analyse the type of search and the type of site the user landed in.
Why does a Blog have an average bounce rate higher than a Portal or an E-Commerce site?
To answer this question, we must first understand what is the behaviour of a user who visits a Blog. Or the type of queries associated with the content of the Blog.
Blogs are, essentially, informational websites. They tend to host specific — and often niche — information. Their content often answers questions and seeks users interaction. Very rarely a blog is a website where you sell something, like in an E-Commerce, or has the content diversity of a Portal. Thus, blogs tend to appear more prominently in search results for informational queries.
The typical behaviour of the user who wants a piece of information is:
the user searches;
the user enters the site;
consumes the information (reads the post);
the user leaves to continue doing whatever he was doing.
Bam! Bounce!
Bounce rate tends to be lower where people spend their time
Portals and Social Networks, usually, have an immense amount of information. These are mainly visited by people who don’t quite know what they want, or have some recreational/leisure time to spend. Portals respond, primarily, to navigational queries or direct traffic (if you want to go to Yahoo or AOL that’s what you do). When a person enters a portal, he/she doesn’t want just to check one page and leave right away. Inside these kinds of websites, people try to discover content, browse for news, weather, check email, etc… So, naturally, a Portal or a Social Network is in average a site with a low bounce rate.
High bounce rate is not necessarily bad
You must understand, once and for all, that bounce rate doesn’t always mean your content is bad. Well, sometimes it is! But we must be able to distinguish between the person who entered the Blog — read the content, was very happy with the quality but left the site without visiting any other pages — and generated a bounce, from the person that entered a Blog — read the content, but was not satisfied and tried to browse in other pages — and eventually left frustrated. Or clicked on an advertisement, deemed more relevant than the content itself. As you may have noticed, in the case where the person navigated through the site, trying to find content, is not considered a Bounce. Although their experience was worse than the person who entered, was pleased but decided to leave.
There are several aspects that must be considered before judging bounce rate. The most important are: query intention, usability and quality. We must never make an isolated judgement, without considering all the factors at play. Analyse the data, considering what you offer and what the user intended when he visited our site.
Bounce Rate: The Simply Powerful Metric
There is a short video, where Avinash Kaushik talks about the bounce rate in a very passionate way. Again, one must keep in mind that the analysis done in the video is made from an E-Commerce and conversion perspective.
Google Documentation
Also recommend you read the documentation on Google’s Help Center about Bounce Rate, High Bounce Rate and Exit Rate vs Bounce Rate.