# How can I have more bounces than exits in Google Analytics?

According to Google Analytics, a particular page has 97710 page views. This page has an exit rate of 57.82% and a bounce rate of 64.4%. According to the Navigation Summary page, 91.76% of page visits were entrances, and 8.24% came from previous pages.

From the page views and entrance rate, I calculate the number of direct entrances as:

`.9176 * 97710 = 89659`

and from that, I calculate the number of bounces as:

`.644 * 89659 = 57740`

This seems reasonable. Then, from the exit rate and the number of page views, I calculate the number of exits as:

`.5782 * 97710 = 56496`

This too seems like a reasonable number...but then I saw that the number of bounces (57740) > the number of total exits (56496)!

I've read about what exactly bounce rate and exit rate are, and I've looked at a few examples online where people showed examples of how to calculate bounce and exit rates. Does anyone see anything wrong with my math? Or do I misunderstand how Analytics works?

the data is correct, your math is too. the only difference on why you get a discrepancy of 1244 is because you are calculating with only 2 decimal numbers.

if for example your "real" bounce rate is 91.7653% (and not 91.76% flat) that would result in 89659 direct entrances

``````.917653 * 97710 = 89663
.644353 * 89663 = 57774
and so forth
``````

also your total exit rate can be smaller than your bounce rate. for example visitors could time out on your site and hence a new session is made without accounting for an exit ...

• Hey, no problem. Glad to have helped ... Aug 4, 2011 at 11:20
• This doesn't make sense, increasing the decimal places doesn't account for that big a discrepancy. The highest the exit rate can be would be about 56,500.
– cjk
Jun 13, 2012 at 9:05
• well analytics cannot be used for exact number crunching. there are many factors to consider: overlapping sessions in his measured timeframe, timeouts, performance issues, etc. his calculation is correct and the small discrepancy of several hundred is too small to account for a faulty implementation or mathematical error. also a bounce and an exit are different metrics and exit rates can in fact be smaller than bounce rates, due to timeouts, etc. Jun 14, 2012 at 11:04