Forum Moderators: martinibuster
The CTR has remained consistent and the EPC slightly up.
I have written to Google about these discrepancies and, of course, all I get are canned responses.
Either Google is fiddling me, which I do not believe, or there is a bug in their reporting systems counting pages and ad unit impressions.
Something is wrong and no one is listening. Is anyone else seeing such a difference?
Looking at my site stats and the Google stats are like looking at two completely different things. I have Adsense set up in my forum so it is appearing on every page view. My server shows 98,000 page views for yesterday and my Google stats show 34,000.
Maybe this has something to do with their graphic ads. I have been seeing a lot more take up the leaderboard that I have up. In fact yesterday I was seeing a LOT of Chinese gibberish in the banner ad. Just what my audience base wants... NOT!
if you want your numbers to match Google's with any degree of accuracy.
That is not the issue since the Adsense stats and me did tally until 05/05/05 plus on three days since then we have also tallied, however for Adsense to be 20% less on all the other days is a serious difference and especially so when those were the only three days whereby my earnings were actually above the average daily earnings since the 5th.
Wouldn't you be questioning if you saw, for example's sake, Google's 25,000 ad unit impressions per day reduce to 20,000, Google's 10,000 page impressions per day reduce to 8,000, meanwhile you knew from your own statistics that the traffic had not only remained consistent but had actually risen a little?
Now you can probably appreciate why I am asking whether anyone else has seen such a variation in the Adsense reports and their own.
If my server is not accurate, why does it tally exactly with Google's reports on some days, but not on others?
Because some spiders can be identified from their behaviour, such as not downloading images or not providing referrer info. But other spiders cannot be identified so easily, and thus your stats will be incorrect to an unknown degree each day unless you have implemented proper tracking.
(Your tracking of humans will also be off for similar reasons, but IMO the most likely explanation for day-to-day swings is spidering).
This is the running-faster-to-stay-in-the-same-place phenomenon cited by quite a few webmasters. Add lots of quality content and raise the number of visitors and, quite coincidentally, epc goes down.
I know, it could be due to a whole bunch of factors (all unknown, as there is no transparency) as defenders
of the holy adsense faith will say time and time and time again.
I wonder if this could account for the discrepancies?
If you have 100,000 page views, but AdSense failed to display 20% of the time, wouldn't Google show 80,000 impressions as opposed to your 100,000?
If you have 100,000 page views, but AdSense failed to display 20% of the time, wouldn't Google show 80,000 impressions as opposed to your 100,000?
A page view is a page view, an ad unit impression is an ad unit impression. Two entirely different things.
Quite often a new ad unit is not displayed simply because there are no new ads to be shown etc; however I would have expected that Google's page views would match up within a few percentage with my page view figures.
Until 05/05/05 they never varied more than 2-3%, now they are 20% less.
I am not saying Google is fiddling me, far from it since if they are being genuine then are they fiddling themselves or has someone found a magic key at the Googleplex and is syphoning off from certain accounts?
It's the old bank teller's halfpenny trick, it's not legal tender by itself however accumulate enough off everyone's accounts and you have a nice, tidy sum.
My figures are looking a tad better today and since I'm away all next week on business I shall just check in a couple of times a day to see what happens, but the situation still remains that I am reporting more page views and Google is reporting less.
I fully agree with your findings and I'm actually exchanging emails with Google about this issue
Hi elfred, are you getting anywhere with them about it though?
All I have received are canned responses such as:
"When tracking revenue trends using your AdSense reports, we suggest comparing similar days of the week. As traffic volume and type can fluctuate from day to day, this method will help you to generate the most accurate reports."
Not one response has addressed my queries to them and responses such as this enrage me even more since it is plainly obvious that whomsoever has received my e-mails have not read them because they would know that what they are suggesting I had already done.
They are treating us as idiots and implying we have no idea of how to track trends, read statistics and comprehend them. I've news for them...I used to teach this methodology to sales staff so that they could analyse their own and competitors statistics!
After a reasonable start to the day yesterday it again fell flat on its face and I ended up with what I would have said only a few weeks ago "a typical weekend's earnings" which are usually 70% of an average week's earnings.
I have made the decision this morning that on 1st June I shall be using another advertiser across my second biggest earning site and will let it run for a couple of weeks to assess the impact.
Of course this has two advantages:
1. To see if earnings rise or fall both on this site and site wide across all the other sites still using Adsense.
2. To see how many pages they report and if they agree with Google or are substantially different.
I hate having to do this however I now see no alternative since corporate G ceased listening to us all a long time ago.
However, today, my page impressions are the lowest they've been since I've been serious about Adsense (my big mistake ;-), and I've added lots of pages since the start, back in Novermber of 2004. Page impressions are going to end the day nearly off by 65% from my best day 2 days ago, but, CTR is nearly double from the 19th, at a rate I've never seen before. eCPM is pretty high, but only by about 25% higher.
So, should I assume that the people who no longer are visiting my site were click slackers, and those who hit the site today are eager clickers? Just because there is less traffic, there is no direct correlation to higher CTR, but that's what my figures ar showing.
I would permit that this suggests Google's page impression stats are way off, at least today. I'll check my site stats tomorrow monring and see what they say, as they are fairly close, relatively, to what G reports.
But could it simply be eurovision song contest/big football match (in uk) etc causing it? Everyone I know seemed to be watching one or the other...
My site has been hammered down to nothing - hits are down to about 20% of what they were a week ago - by the same company that wants me to host their ads and make both of us money. Doesn't make much sense to me, especially business sense.
I have a very hard time believing Google actually knows what they're doing.
I have a very hard time believing Google actually knows what they're doing.
It really does make one wonder how many people are tweaking whichever part of their algorithms and which departments are working in tandem with their so-called work colleauges?
Yesterday, Saturday, I had my highest ever CTR and eCPM, the actual EPC was up more than 10% on a week ago and the actual earnings were nearly a normal Saturday for me, and this with a Google page impression and individual ad count of 10% less than the last 3 weekends and 25% less than 4 and 5 weekends ago.
With a little luck I may just earn the same this month as I did last month however it is going to be a very close run thing.
My page counts show a completely normal weekend so I have absolutely no idea what they are doing except driving us all barmy!
Andi
Yesterday, Saturday, I had my highest ever CTR and eCPM
This is in line with what I predicted. Somehow, I think that Google is reporting less impressions than they actually are. This causes CTR to increase. If Google is protecting against impression fraud, they might simply have added a computation when showing impressions. They might be reporting 75% of the real figure.
I will have to give it a few days to see if I can see if this trend continues and if so, I will somehow have to look at the situation and try and resolve it.
Any opinions on if I am right in thinking along the lines of smart pricing hitting my site.
Thanks in advance
The total earnings for May are down 1/3 in this last hour
We had the same issues - with earnings CTR and eCPM dropping really really low - lots of untargeted ads... i started blocking the crappy ads - then gave up after awhile...
by leaving thing's alone it took a few days and things are as right as rain now... and even better earnings than beginning of the month... though still not as good last month...
i think the new system involves Google evaluating more ad performances for particular pages on your site with their wider keyword matching algo - and once the best ads are found for your pages earnings increase again... well it appears that way to us now... it's difficult not to get crappy or panicky when things drop dramatically by 70-80% - i normally freak a bit - but learning ride it out and leave your site alone does good - it should improve after a couple to a few days...