Welcome to WebmasterWorld Guest from 126.96.36.199
So either this referral site (a modest PR 5) is somehow sweeping up 10 times what I would expect in referrals, or it is manipulating the number it is sending.
I run around 30 individual hotel sites, so have a lot of experience in the numbers to expect.
1. I have looked at source code and there are no (that I can find) pulling downs of photos off the clients site. The only links to my clients site look kosher, in other words the user would have to click the link to visit the clients site. A true referral
2. There is an odd thing in the source code. After the head tags there are serveral blank screens to scroll down to get to the normal page coding.
3. I can find no evidence of a redirect. The page given in the clients logs appear to be the page I can read the source code for.
4. I have been to the referral site (without activating any direct links to my clients site) and checked the client logs over a period of time.
The client logs do (vaguely) seem to reflect my visits to the referral site. But appear over a 2 hour period and from a motely collection of IP numbers.
If I do not visit the referral site for a day and stir around, the number of referrals drops dramatically
5. The referral site has a contact address as a PO Box number.
I cannot prove wrong doings, but I am convinced there is a scam
Has anyone any experience of trying to track down what it is?
Perhaps I am not as stupid as you imply ;) That was my first assumption, which was why I examined the situation in the first place.
I have become intrigued enough to get the other guys traffic analysis (is nothing sacred in life) ;)
Basically it does not have great traffic. 600 total unique visits a day on average to his referral site. That figure fits with what I would have expected. The site contains in total details of hundreds of hotels across all its pages
The main page with my client on gets 50 unique a day, and 15 a day of these are now being referred to my clients site. Clients logs show referrals from that page. There are 3 other pages with my clients details on and these generate a further 15 referrals a day, making 30 referals a day in total.
I have spent a lot of time trying to see if it is a traffic pipe that I have not yet worked out. I remain convinced that the balance of probability is that there is another explanation!
That is an annual rate of over 10,000 referrals a year, and in the uk hotel business is an outstanding figure. Many hotels will struggle to get 10,000 visits to their own site in total.I have access to lots of hotel logs and know what referral numbers are across the board. Anything over 1000 unique a year is good. Anything under 100 is normal
The main referral page for my client has 72 hotels on it, my clients is bang in the middle of this list, and has an ordinary entry, in other words nothing distinguishes it from the 71 other entries.
The referral page is an unordered list, and the site itself is not well positioned on serps. There are no niche terms either, their logs show that there are none.
It seems, to say the least, unlikely that 1 in 3 visitors to that page is going to click on my clients hotel url as a single choice. Statistically this suggests that each visitor to the referral page is clicking through to 20 or 30 hotels.
I remain convinced that the traffic to the referral site is unlikely to generate that number of referrals. It appears to be an ordinary site, with no special gismos, no special navigation or layout, no remarkable search engine positioning. I would have expected it to have around 100 referrals a year, it is delivering 10,000. Having looked into it, I remain convinced that it cannot have more than that in genuine referrals.
Any site that delivered 10,000 a year referrals to a hotel site would be a nice little earner (no I won't sticky anyone the url) if it was genuine. I will just keep on trucking to see if I can explain the traffic
This will be server side code executed when you hit the page. Could be absolutely anything.
I would look closely at the IPs used and the pattern of these visitors once they hit your own site to see if any of it looks automated and to check visitor quality. In any case, if they are real and converting visitors you're being sent, does it matter particularly how they got there? ;)
Thanks. Yes, that is my next line of analysis, but I have to wait till Sunday to do it. Othewise I am having to download megs of log files each time I test the effect of entering the referral site on the client site log files. (my server starts a new log file weekly on Sunday)
Doesn't reading raw log files sends you cross eyed!
>>In any case, if they are real and converting visitors you're being sent, does it matter particularly how they got there?
If it were the route to untold riches, then I would want to know how it was done ;)
My .02 is that he has some rankings somewhere that you are not privy too. Shoulder up to the guy and see what you can figure out. It is surprising what people will tell you if you just ask. (inquire about running ads on his site, but you must know the traffic source and numbers ;)
The real gems of promotion for any kw or site are buried deep in 3-5 kw word combos. I've seen some on Google produce significant traffic numbers. However, I'm also coming to the conclusion that my idea of significant and others idea of significant are quite different.
It's OK I wasn't offended (that's why I put the smiley in) - difference in US and UK senses of humour!
I will run a test on Sunday, when log files are not so big and easier to read.
If he is using a random redirect script server side, to a list of URL's then I should be able to get an idea of what is going on if I do a controlled test.
I have checked his total site traffic, and have also got the Google referral numbers and keys words used for his site (amazing what you can find out about other peoples web sites if you put your mind to it). I remain convinced that the referral traffic is in the region of a factor of 100 too high.
If its not, then the guy has the secret to untold wealth - either way it's worth my whjile finding out how it's done.
As I pointed out in my first post
>>After the head tags there are serveral blank screens to scroll down to get to the normal page coding.
and post 2 here suggested
>>This will be server side code executed when you hit the page. Could be absolutely anything.
Having found the sort of scripts that will do random redirects from random ISP (from nominated lists) it would appear that this could be the explanation for the blanks in the coding.
For those of you interested go to the fount of all knowledge and search for things like random redirects, redirectors, server side, ssi and the like. Make your hair curl what you can do with these things
Most of them dispise the "prevent right-click" code that can be placed in the page to protect the source because it also prevents other useful actions.
It has become a very common practice for them to insert a big block of whitespace before their html content so that when a visitor chooses to "view source" they see nothing.
They swear to me it works, even after I pointed out that the visitor can simply scroll down to see the source.
Thanks for that, but they have the head tags there before the acres of white space. So if you view source you can see the head tags.
If nothing else I have convinced myself that it is a SSI program that generates random visits. Most of them are from a not particularly prominent ISP (cos they are dynamic, one can only really see the ISP each visit rather that the particular user coming back again)