| 9:59 pm on Jul 9, 2013 (gmt 0)|
manually typing a url is a direct request.
clicking a link is a referred request.
there's no way to track a direct request.
in some cases where the Referer: header is not sent with the request, you could have a visitor who clicked a link and that referred request is not trackable so it would look like a direct request in analytics and your server access log file.
| 1:37 am on Jul 10, 2013 (gmt 0)|
|Is there a way to track users whose previous page was xyz |
No. Unless, ahem, you also own page xyz. I remember when I first learned this, probably in the context of hotlinks. It's very aggravating but there's absolutely no way around it. The user-agent (whether browser or robot) does not and cannot tell you where it was or what it was doing at the time it put in the request for the page. It similarly can't tell you whether the request was a type-in or a bookmark-- another thing it would be useful and interesting to know.
| 5:37 pm on Jul 10, 2013 (gmt 0)|
Thanks for the response. Never even considered the whole bookmark issue.
| 5:30 am on Jul 11, 2013 (gmt 0)|
create a special page for visitors from ebay on your site.
Then make this URL known to ebay visitors in your ebay auctions / me-page.
When they type n this url (instead of the url they arey using now), you know they are from ebay.
| 3:15 pm on Jul 12, 2013 (gmt 0)|
besides bookmarks, direct visits, and stripped Referer: request headers, "unreferred visits" can occur when links are clicked in email clients, apps, or any type of non-web-browser referral.