People believe that the lower the bouncerate, the better, but as the old saying goes, the lower the bouncerate may not be as good as we expected. BounceRate is one of the basic metrics of website analysis. I believe everyone is familiar with it. Bouncerate seems to represent
People believe that the lower the bounce rate, the better, but as the old saying goes, the lower the bounce rate may not be as good as we expected.
Bounce Rate, one of the basic metrics of website analysis, I believe everyone is familiar with it. Bounce rate seems to represent the "magic challenge" given to us by some kind of website analysis. if we have the power to uniform or control it, we seem to have achieved something remarkable.
Therefore, most people believe that the lower the bounce rate, the better, but as the old saying goes, "Too Late", the ultra-low bounce rate may not be as good as we expected.
Monitoring technical problems with ultra-low bounce rate
Either way, if the bounce rate of a website (or some major landing pages) is lower than 5% or even less than 10%, there may be monitoring technical problems. There are at least two possible causes of low bounce rate.
First, if monitoring code is added to both the parent and child pages of the Framework page, in most cases, when the page is loaded on the browser, the monitoring code is run twice in a short time. For website analysis tools such as Google Analytics, two different pages will be opened (because the URLs of the parent and child pages are definitely different ), therefore, this situation is obviously considered not a bounce, even if the user does not click any link on this page to browse other pages. In my example, in this case, the bounce rate is lower than 5%.
Second, if two identical monitoring codes are accidentally added to the same page, or for some reason, the same monitoring code runs twice. When a page is loaded, Google Analytics does not calculate this situation as bounce as it is refreshed once. In this case, the bounce rate is even lower than the first case.
Of course, there may be a third case, namely, traffic cheating. In the case of poor cheating, the robot accesses the website according to a preset program, which seriously reduces the bounce rate. This situation is undoubtedly not worthwhile.
Other negative information with low bounce rate
If the bounce rate is unreasonable, we can determine that there is a technical problem in this case. But if it is very low, and it does not reach the very low range, can we just breathe a sigh of relief?
I did not dare to come to the conclusion easily. in a sense, the low bounce rate may also convey some other negative information. The biggest problem is that the low bounce rate may mean the lack of fresh traffic.
There is a conventional rule in website analysis, that is, for commercial websites, the bounce rate value is often linearly related to the new access ratio, and some or even the two have an approximate relationship of, which is quite interesting. We can think that in order to get more new traffic, we tend to sacrifice a better bounce rate, which is completely understandable. It is the correspondence between the bounce rate and the new access ratio of nearly 50 websites I have monitored, with obvious linear correlation. This means a "paradox" that we don't expect but actually exist. if we want bounce rate to decrease, we should try our best to save old users, but the proportion of old users is too large, the website traffic situation may be less optimistic.
Of course, it is not absolute. the bounce rate is not only determined by the proportion of new access. For example, we can also see that there are some points in the lower left corner that significantly deviate from the trend line, showing the "less beautiful" side-a very low proportion of new access (10% ), but there is still nearly 50% bounce rate. In addition, the overall trend of bounce rate and new access rate also shows that when the new access rate is low, the performance of bounce rate is very discrete. We seem to be able to come to the conclusion that a too high proportion of new access will almost certainly affect the bounce rate, which also tends to be high; but when the proportion of new access is low, the bounce rate is affected by more other factors.
Low Bounce rate also means another possibility, especially when it suddenly gets low, we believe it is caused by abnormal guidance or temptation. If this guidance or temptation is intentional, it is beyond Review. However, unexpected situations may mislead website operators. One of the examples I have ever encountered is that the bounce rate of a website suddenly drops by about two or more in a day, which is encouraging. However, looking at the data carefully, we found that the sales volume of the website has not increased any more, which is not in line with common sense. We further discovered that the station's art sales channel went online one or two weeks before the bounce rate fell, and we further found that on the day of the decline in bounce rate, the website has a lot of traffic from the search keyword "human art" or "human Photography", and the landing page of these traffic is the art sales channel. Interestingly, there are indeed some paintings or photos with few or no clothes on this channel, and they do get better clicks.
From the perspective of actual operations, bounce rate is indeed more pleasant than others, but be careful to ensure that our bounce rate reflects the real situation rather than false prosperity.
Troubleshooting of abnormal bounce rate
In turn, if the bounce rate is really high, it is not necessarily a big problem, or we should at least ensure that the high bounce rate is not caused by inadequate monitoring.
The most important non-location monitoring is prone to occur on pages with many non-HTML interaction elements, such as a page with many links consisting of Adobe Flash or JavaScript.
Both Google Analytics and Adobe SiteCatalyst require additional code processing for non-HTML interaction elements, such as tagging these elements using the Event Tracking or Virtual Page monitoring feature of GA, in this way, the user clicks on these elements. otherwise, all interactive website analysis tools related to these elements will not be able to perceive them. For example, on a page, all the circled buttons are expected to be clicked by users, but they are all made by Flash, the new page opened after clicking is not your own (but jump to another person's page, which means that the page opened after clicking cannot be monitored)-that is, if no additional monitoring settings are available, when this page is used as a landing page, even if each visitor directly clicks these buttons on this page, the bounce rate of this page will be very high. This does not reflect the real situation.
In my opinion, if the bounce rate is really bad, do not be confused. first, eliminate possible monitoring problems. When you are facing a bounce rate of up to 95%, you should do this: Step 1, ensure that all user click behaviors can be monitored, especially for non-HTML elements; step 2, check whether the proportion of new access to the page (or website) is as high as 90%. If you have checked and ruled out these two steps, then you need to really worry about the bounce rate of up to 95%.