I've noticed that "no referral" now accounts for about 60% of the traffic on my site. Last year at this time it was about 40%. How to explain? Is less of the traffic getting tagged correctly and more just getting dumped into this catch-all bucket? I have a hard time believing this is due to any change in customer behavior (ie. 50% increase in bookmarking and fewer customers using search engines to get to my site). Anyone else seeing this? It's making Urchin even less useful.

After checking my stats this morning, I'm putting it a support ticket so they can check it. For yesterday alone, my non-referral percentage has gone up to 87%. It does make the stats totally useless.

Hey Autotradastores and Barbiro.
Excellent question. I've seen this in my stats too (for a long time) and just took for granted that it was correct.
Barbiro: When you get your ticket addressed, can you please post the details or info you got back?
Thanks!

MartiniGuy, this is the replied I received from :



So, I guess, they're basically saying that 87% of my viewers have my brand new site bookmarked? Hmmm..... I really don't believe that is accurate.

Maybe if we all put in trouble tickets it will get someone's attention. Or we can e-mail Ryan or Steph. I'm sure they could find out and tell us what's going on. I agree it pretty much makes the stats useless as we're looking at less than half a loaf now.

You're right and it's been recent - 2 weeks ago we were at 43.64%, now we're at 70.09%

thanks for the reminder to check!

I just put in a ticket as well and sent an e-mail about it to Steph.

This just in from tech support:

"What no referrals mean is that the visitor knows your website name directly and are going to it not through the search engines but by typing in your website name in their browser. Unforturnately, MC does not have any control over what is tracked with Urchin. But I don't believe any thing is wrong here, just that users are recognizing your site name and are going to it directly. Thanks!"

Ummm, I don't think so.

You know, I wondered about our "no referrals" being so high as well. I thought we were doing pretty well by having so many folks know who we were that they'd go directly to our site! Duh. Maybe there is an issue here that needs to be researched a bit.

Sorry, but I don't think so either. I just checked my stats for 2 other web sites I own and the non-referrals are 27.44% on one, and 17.64% on the other.

But do any of you show any results from Overture?

I don't.

I assume that since I have so many Overture listings that they are coming from there.

I know for a fact that thousands of people each day are not reaching us through bookmarked pages. Our site's not that hot.

The non-refs could be bots/spiders hitting your pages. See it the increase in non-refs relates to high bot/sprider traffic. I bet it does.

I got the same reply from tech support, but it doesn't make any sense. Essentially what I'm looking at is an additional 451 sessions, but no additional orders. If it was repeat customers, I'd assume we'd have additional orders. The fact is we sell boring printers & toners so it's not like they're shopping around getting holiday ideas

As far as the bots, I'm at a lose how to really determine if that's what this could be. I see the MSN was all over us last week, but google was way down. Why would urchin suddenly not be able to recognize google if that's what it was? Any idea? Thanks!

Has anyone else noticed an improvement in these stats this week? Ours have come back down to the 50% mark which is a big improvement

s, was any research done on this to see if there's a problem? My non-referral is averaging over the 50% mark, which is really dissapointing not being able to determine where the visits are coming from.

Date: November 3, 2004 = 58.72%

not to take over the thread, but what the hell kind of stats do i have here?

no referral 64%
google 14%
everything else under 2%

Where is the yahoo and msn traffic? How can you get a number of how many unique visitors are on a site?

Wow...your non-referrals are even higher than mine

I'll bet you're right on that! We only have 2 referrals in the past week that appear that they've come from Overt*re..and I know our ppc costs doesn't reflect just 2 charges Good catch!

BTW - our nonreferrals are up to 63% again

When I switched to , it was around 50%, now its up. Are those referrals from other pages within the site, thereby making them a useless statistic and inflating the overall numbers?


Breaking off topic again, speaking of overture - have you found success with it?

You are breaking off topic again

but

that's okay.

I'm signing up with Overture this upcoming week and I'm curious about it as well.

Clicks from your member profile on M C--where do they show up in Urchin?? Just curious

jvautoart, my non-referrals have topped yours

88.56%

, is anyone researching this?

Im holding at 64.95%!

, which Urchin software version do we have?

Question with regard to Tech Support's response of:

"What no referrals mean is that the visitor knows your website name directly and are going to it not through the search engines but by typing in your website name in their browser. Unforturnately, MC does not have any control over what is tracked with Urchin. But I don't believe any thing is wrong here, just that users are recognizing your site name and are going to it directly. Thanks!"

So if someone has our site bookmarked, the Urchin stats do NOT pick it up? Does that fall as a non-referral? I'm just failing to see how 85% of my viewers directly type my site name in their browser.

Barbiro,

I know this may not you much, but this is from the Urchin website :


Why do I have so many 'No Referral' entries in my Referrals report? (#1465)

2004-07-05 v.5.x

The "No Referral" entry in the Urchin Referral report is typically do to one or more of the following conditions:

Your web server log files are not formatted correctly
Visitors are accessing the site directly (by typing in the URL or book marking the site)
Visitors are accessing the site directly by clicking on a link contained within an email.
Search Engine Web spiders are identifying the web pages
Internal referrals resulting from cached page requests (not applicable when the Urchin Traffic Monitor (UTM) is properly configured.
Log Format:
In order for Urchin to report Referrals, the log files must be configured to include a Referral field. See the following document for more detail on configuring Apache or IIS log files. Logging - Apache and IIS
Other log file format information is available under the 'Documentaion-> Urchin Administration-> Log Files' section at http://help.urchin.com

Direct Address:
If a visitor types your website address into an explorer address bar, there will not be any referral information passed to the web server. Every file request -or "hit"- to a web server is logged in a file, and each entry has a referral field which may or may not contain data. When an address is accessed manually as described above, no information is recorded in the referral field because the visitor was not referred by a link on another site.

Email link:
When a visitor clicks a link in an email message (such as an email advertising campaign) there may not be any referral information passed. It is possible to create email links that pass referral information to the web server the site resides on. But, this must be setup by the individual that creates the email content.

Web spiders:
Robots crawlers and spiders spawned by search engines usually account for the majority of "No Referral" hits. these are all programs that visit Web sites and read their pages and other information in order to create entries for a search engine index.

Internal Referrals:
Urchin does not include any internal referrals in its reports if they are configured correctly. An internal referral is a referral from the same website that the first requested page belongs to. If an internal referral is discovered in the referral field, it will be included in the report as a "No Referral" hit. It is uncommon for an internal referral to appear in the initial request for a web page. But, there is one specific circumstance in which such referrals are common.

Many large ISPs and Search Engine sites use cache servers to capture web site content. By doing so, they can reduce the time and network traffic necessary to serve these pages to visitors faster. But, when a visitor requests your web page and a cache server responds to that request, your web server log has no knowledge of that request. As a result, your web server log file will not contain any information about the request either. The net result of this action is that a new visitor will often not be seen by your web server until a second or third page request is made. By that time, the pages he or she requests will be accessed from links on other pages on your web site. Therefore, the referral field in the log file will display your own pages as the referral. This results in the internal referral hits that urchin adds to the "No Referral" report.

Unless you are able to keep all of your web pages from being cached remotely, there is nothing you can do to remedy this situation with the Urchin 3 product. However, both Urchin 4 and Urchin 5 can reclaim the missing log file data if you configure the Urchin Traffic Monitor (UTM). More information about the UTM is available from our documentation center under the "Visitor Tracking" menu.

\
from
http://help.urchin.com/index.cgi?&id=1465

you know, in trying to answer the last question, I think I created a new one for myself.

If "no referrals" can be from search engine robots, that means robot sessions are counted as part of the overall sessions and thereby void the chance of finding the number of human sessions on the site? Is there any way to get a fairly accurate number of unique users on the site per day/week/etc.?

Thank you for that lengthy explanation. It would be nice if we had Urchin 6 instead of Urchin 3 (I think that's what we have; I'm not sure). It seems to take care of a lot of the problems.

, is an upgrade in the works?

my urchin is v. 5.7.01

ok, im going to stop trying to answer your questions that you are directing towards .

, you hear that?

Thanks for the post that you made just above this as well! This obviously has been bugging me for a few weeks as well. Lo' & behold - when I was checking the stats this morning...our sessions had more than doubled, and our (no-referral) had shot up to 84%!

Since there are Robots by Hits Drilldown, I've been having a hard time coming to terms that No Referrals could be including robots. What I think has finally sunk in is that Robots are recognized - spiders are the no referrals...could this be right? Imagine this...here comes the MSN Robot Bus (that counts in the Hits Drilldown) but then 99 spiders get off the bus & crawl all over town...Urchin only recognized the bus - not the tourists that got off....When I look at the length of session & see that over 1/2 have been less than 10 seconds, I have to come to this conclusion.


Yeah, it's time for another pot of coffee...strong!

Thank you, Cookie.

I've been really upset about the "less than 10 seconds" thing.

I was really thinking that there were massive numbers of fraudulent clicks, but your musings make much more sense.

I need to stop worrying so much about the Urchin stats and concentrate on other things.

Hey guys,

Here are a few things to keep in mind.

1) All stat servers are running v5.7.01

2) No referrals can happen from a variety of causes such as robots hitting the site, you visiting the site, current customers hitting your site (maybe clicking on a link in an email they got or just typing in the address directly), and former customers hitting the site.

I will display a standard day in my stats for example.

Sunday I had 204 no referrals which totals 58.62%. However I had a grand total of 441 robot hits. Now it is important to realize that all of the robot hits are not sessions, but still I'd wager more than 25% of them are their own unique session as treated by Urchin. Combine that with the numerous times I visited my site throughout the day (I'm sure at least 10 if were to count) and that my customers may have been checking up on their orders, and repeat visitors that have come to my site before and are coming back, I'm sure that number is just about right on the money.

It is unfortunate that robot traffic is included in the stats since it does skew the numbers a bit.

Urchin 6 (at least right now) is a completely different service. Urchin on demand... where the solution is hosted on Urchin and you have lots of great reporting goodies. However their standard price is $499/month. It's totally worth it if you do a good amount of business and need to know everything about your traffic. I find Urchin 5 to fit my needs great though (saying as a site owner) and I can track everything I need to with a little bit of effort (such as making every single ad of mine have it's own tracking URL). I've used other stats packages in the past (AWStats, Webalizer, AnaLog, IndexTools, etc) and this is the most bang for your buck I think you'll find.

Just some food for thought.

Also from what I've heard with some Microsoft updates and other security suites your referrer can be blocked.

I know when I updated my home machine a couple weeks ago my security settings were set really high and I had to turn them all down to do anything effective.

Don't have solid proof on that one yet, but something to keep in mind

We use a completely seperate stats program along with urchin and find that if you take your search engine sessions and add about 10 to 15% depending on your type of business that will give you a fairly accurate # of total users for your site. The sessions urchin reports does indeed include not only all the bots but also any internal work you are doing to your site behind the scenes. If your urchin stats says 1000 session you are likely only receiving about 300 total users for that day. Keep that in mind when you are determining your conversion ration as well. The stats package we use is www.hitslink.com and I feel overall it is better in certain areas for accuracy, although the log files from urchin can be better in other areas. Hitslink is only 15 to 40 bucks a month depending on your traffic and it is real time stats which also helps.

woo hoo im under 50%!

, my non-referrals are up to 92% -- that is really overdoing it! It's totally useless to have stats when I can't figure out where 92% of my traffic is coming from.

Can someone please look into this? I'll open up a support ticket.

Actually, for yesterday alone, here's the reading:

(no referral) 755 Sessions 98.18%

Just wanted to let you know I am checking with Urchin currently and trying to find out why this might be occuring.

I have the exact same problem and Urchin gave me the same excuses but if I duplicate one of my profiles and then run the new empty profile on the exact same log file all the referrals show up.

It seems like it is a problem with urchin profiles getting corrupted. On my sites sometimes it happens within days and sometimes it takes months but once it happens the no referral number jumps above 50% and stays there.

It is frustrating that Urchin does not address or even admit the problem.

It is possible version 6.0 would fix this problem but when I bought the software in September they promised they would release 6.0 in the fourth quarter of 2004. When that date passed they then said it would be released in late January or early February.

Wait, so you're saying that its a known bug in Urchin that is causing an error in reporting - and that the data is really sitting in the log files? That's a relief.

I was considering just pulling the log files and using an app that I'm familiar with (Webtrends) but then MC rolled in Urchin as a part of the package I'm paying for and figured since I'm paying for it might as well use it... but with all the probs I've heard with it - it seems like a waste of my time to even bother with Urchin. I'm back to "PlanA".