Forum Moderators: DixonJones
I use WebTrends software to analyse our server logs.
Until now, I've assumed that PPC visitors were included in the "Visits Referred by Search Engines" (VRBSE) figure.
This afternoon, I've gone back a while and found that about a year ago, the actual number of PPC visitors was greater than the VRBSE figure! This would imply that the PPC visitors aren't included after all, which means that my true Search Engine referrals are higher than I'd thought.
I've run some test visits, looked at the logs and the log file information seems to indicate that the right information is being recorded for PPC visitors.
I use tracking urls (eg adding "?src=g&c=ms" to the target url) for all PPC campaigns, but I wouldn't expect this to prevent it registering as a Search Engine referral.
Has anyone any ideas?
As long as you use tracking strings or landing pages the numbers work out.
As I interpret it, you are saying that the Top Search Engine totals does not include the PPC click-thru's (with tracking strings) and I must add them to get a true figure.
My original understanding was that WebTrends picked out anything containing, say, 'google' as a Google referral (tracking string or not).
But the numbers suggest that this is not the case.
BTW, I have left the setup within WebTrends at default values.
This matter is of serious importance to me, because on the basis of my original assessment, PPC appeared to deliver as much traffic as the listings. If the new scenario is true, then it only delivers half as much.
Big implications for my whole strategy!
the Top Search Engine totals does not include the PPC click-thru's (with tracking strings) and I must add them to get a true figure
I wish that were completely true. In my experience there is no simple math to get the clickthroughs from webtrends. You don't seem to be able say to a client/boss "This number plus this number is our overture total" (for example). I got onto to tracking strings by talking to a goto rep long ago.
If you look at referring urls and you use a tracking string you will get a good portion of your clicks in there and they are just tacked onto the page they came in on. They don't say the engine anywhere in that url.
Even think of yahoo, yahoo shopping and yahoo search are not the same thing but the report just says yahoo. So I don't know where that leaves the data quality.
The most reliable way to track hits I've used is to have orphan pages that are only there for ppc landing pages. I then check the uniques to that page in my logs and the numbers are usually on the money.
I usually used a bit different equation for each client. I know the number of clicks from the accounts, I know their penetration/ranking. I can usually reverse engineer a suitable equation from there and it will stay true for that client.
Probably not the exact answer you are looking for but the best I can give.
kgoeres,
I am far from a master but I assume you have a method of tracking aside from webtrends. I wouldn't go with the webtrends number if you are confident that your tracking is fairly accurate. If it is the person referring them to you that says 200-300 then I would find a more reliable system to track the clicks.
We also have a huge number of "Referrers" that are our own URL. I had always assumed that these were bookmarked pages, or where the visitor had typed the url into the browser. Now I have doubts - is this the case? Seems like a whole load of what I've been doing has been based on taking things at face value : was I wrong? (If so, where's my shovel - I'll have to somehow dig my way out of providing months of false conclusions...)
I'd always qualified my work by saying "don't look at exact numbers - look at the trends" (which can sound like BS, I know) but it looks as if I've been fundamentally mis-reporting the stats, which is another ball game.
What I'd hoped to do was :
1) Get the total referrals from the server logs
2) Deduct reported PPC referrals
3) The remainder (I'd assumed) was referrals from listings
If WebTrends can't do this (and maybe it can) then all its detailed reporting on "Top Search Engines" is pretty meaningless too.
A further step step I'd wanted to take was to distinguish referrals from, say, Alta Vista web listings from its "sponsored" (PPC) listings. I'd better hold fire and get the basics right first!
Is there another reporting package that will be up to it?
[webmasterworld.com...]
There are lies, damn lies and then there are web statistics ;)
Webtrends (or any other log analyser) will give you relative figures but not absolute figures for the reasons given in that thread.
In other words if you analysed, say, monthly. Then you can compare the trend from month to month, and compare the relative figures for each search engine, but not have that much faith in the absolute figures
Just struck me, maybe that's why they call it WebTrends!
Just use the PPC tracking url as a filter in the log analyzer, for example, 123loganalyzer, you'll see all the referring domains and urls that lead to your tracking url and you'll see a lot of non-search engine sites there.
