There are many different definitions for the bounce rate:
- Google Analytics Bounce Rate
It represents the percentage of visitors who enter the site and then leave. Be careful that if you use tracked events, the bounce rate will be modified. For example IF you track PDF download or javascript click, user will not be counted as bounced. Here are some instructions to set up tracked events : https://developers.google.com/analytics/devguides/collection/analyticsjs/events
The problem is : you can't know if a user enter your website and read your article 10 seconds or 10 minutes. - Adjusted Bounce Rate :
In this case we modify Google Analytics Bounce Rate to reflect reality : It represents the percentage of visitors who enter the site and leave before a certain amount of time. It's far more useful because in many situation, and for Google also, if a user enter your landing page and find the answer to its query, it's perfect. But if a user enter your site, then leave it instantly, you may have Panda issues. A high adjusted bounce rate is obviously a problem, because the page does not appear to meet user needs. For further info and how to setup, see https://moz.com/blog/adjusted-bounce-rate or http://www.gsqi.com/marketing-blog/adjusted-bounce-rate-in-google-analytics/ and search for poggo sticking or dwell time.