As a user of Google Analytics, you may be aware that a recent change moved all users to Universal Analytics. Any time a change happens, it is highly likely that users do not understand all of the nuances that the change brings immediately. One of the results of this most recent change is that Universal Analytics tracks site sessions by default in a somewhat different manner than Classic Analytics did. There is an excellent article on Distilled that explains all of the changes that have been introduced with Universal Analytics. However, the site session change can have your metrics off by approximately 10% to 20% when compared to the previous method of recording. Here is an explanation and how to fix your tracking for better metrics.
A site session is when a user arrives on your site, then goes dormant. Site sessions default to 30 minutes unless you change the setting. Later on, the same user may come back to the site. Under Classic Analytics the 30-minute site session would record the source site of the user. If the user timed out at 30 minutes, then returned to the site later on, the same source location was used during the second session. Universal Analytics changes this method of recording. If a user times out after 30 minutes, they are tracked as coming from your site when they come back the second time.
Since most site owners are primarily concerned with traffic and conversions from organic search, the change in how site sessions are recorded may skew your numbers. For instance, under Universal Analytics your site is now being attributed for traffic that may be coming from search or social media. So instead of having the correct number of search referrals, some are being attributed to self-referrals.
To put your numbers back in the proper order to track your SEO sources, you can implement a simple fix. You simply need to add your domain to the Referral Exclusion List.
Once this fix is implemented and your code is updated, you will have a much more accurate picture of your referrals and conversions.