Bounce rate, time and uniqueness
A free video tutorial from Pavel Brecik
Web Analyst and Data Evangelist
4.4 instructor rating • 3 courses • 14,128 students
Learn more from the full courseUltimate Google Analytics course + 50 practical examples
Practical Google Analytics course based on real experience including 50 practical examples + 100 quiz questions
04:40:45 of on-demand video • Updated February 2021
- Understand how the tool works, what data is there and the most important find the business value in it
- Ability to make the right data decisions
- Know which are important Google Analytics reports
- Among others we'll go through 50 practical examples!
- You'll practice achieved knowledge with 100 quiz questions
- How to think in context and not act hastily
English There are a couple of metrics used very frequently when it comes to evaluation: Bounce rate, time and something that contains unique in its name. They all have to be perceived in context, otherwise it's very easy to misinterpret them and make a wrong decision. The first one is Bounce rate. One of the most common metrics in general. I personally love it. And again I have to stress, it has to be used in context. It shouldn't be used as the KPI. Let's look how the Bounce rate is defined. It's simply the percentage of single page sessions. It means that it's the number of sessions during which only one page was viewed regardless how much time you've spent there. The most common usage is when we look at the number with individual page or set of similar pages like all part of details or all category pages. A lot of marketers still consider high Bounce rate as a wrong thing. The first thing we have to mention that the metric itself doesn't tell us anything, at all. Let's have a look at the page with contact information. Let's assume a simple scenario. Your potential customer is looking where your offline store is located. He found the information and left the website, which is OK. It's natural that this contact information page will have a high Bounce rate. It's quite obvious that there are pages with bounce rate very close to 100% and it's completely OK. OK. On the other hand, if we have a product detail page with let's say 1 000 page views and a very high Bounce rate then it's definitely worth to do further examination. In a better case your measurement is broken but you're not affecting your business directly. In the worst case there's really something wrong with particular product detail page. Time measurement. Very commonly used by many marketers favorite metric, even though not many people really understand how time measurement imperfect is. It's based on timestamps and let's have a look on simple illustration how as obvious thing as time is measured in GA. So time measurement is based on the timestamps and timestamp is just the time format. The way Google Analytics works is that it subtracts the following time stamps for particular pages. We're going to look on three examples and see what time will be calculated in Google Analytics. So we have example number one. We have a session consisting of four page views. We started a visit at nine o'clock and we exited a website 9:25. Now we're going to show how Google Analytics will subtract timestamps for particular pages. So we start with page 1. We have to subtract five past nine minus nine o'clock. It's five minutes. And this is the time measured for Page 1. For page two we subtract 9:10 minus 9:05 which is 5 minutes. And this is the time measured for Page 2. For page 3 9:20 minus 9:10 It's 10 minutes. And this is the time recorded for page 3. And then we spend five more minutes on page 4. But because there is no further timestamp the time measured for Page 4 will be exactly 0 minutes and 0 seconds. So, to sum it up we have the first timestamp which is at 9:00 and the last timestamp 9:20 which gives us session that lasted 20 minutes. Example number 2. Let's assume a scenario: user started a session at 9:00. Then he viewed the second page at 9:05, the third page at 9:45 and then he left at 9:55. Now as we know how inactivity window works we see that between Page 2 and page 3 there's a 40 minutes window, which we know is the moment when a new session starts. So we actually have here two sessions. The first one started at 9:00 and ended at 9:05, which gives us a five minutes session duration and then there is a third session that started at 9:45 and ended at 9:55. We only here have one timestamp at 9:45 which gives us a session that lasted exactly 0 minutes and 0 seconds. The simple rule worth remembering is that we don't measure the last page. And now as you know how the Bounce rate is defined, just check for how many sessions your average session duration is exactly 0 minutes and 0 seconds Plus in the non-bounced sessions as mentioned the last page is never measured. Yes, it's probably unfair but this is how GA works. So I guess from now you will rely as much on this metric any more. There are much better ways to track content than time. Let's say events, with the information about to what percentage have users scrolled. This requires of course slightly advanced measurement implementation and you'll find a tutorial for it in the description of this chapter. There are a couple of metrics in GA that have a word unique in their names. For example page views and unique page views. Uniqueness in GA means that something occurred once per session. It's not per user or something else it's per session. So if you view one page 3 times during one session, it'll be counted into page views metric 3 times and into unique page views just once.