Forum Moderators: martinibuster
Yes, for you. But not for everybody. So if it's a "trend" or a "pattern," it's a trend or pattern only for your account and other accounts that share those same characteristics.
EFV you are an excellent foil for discussion since your posts are well thought out and coherent, but your arguments are often constructed back from your desired conclusion rather than by a process of deduction from the premises. If you disagree with a premise you will call it or the integrity of the poster into question rather than plug it into a hypothetical so we might learn something. I do think though that by keeping that in mind as a proviso your contributions are valuable. But countering your arguments can be an exercise in intellectual wheel-spinning and not necessarily productive in learning about the topic.
Humans are *very bad* natural statisticians.
We're programmed to spot unusual data points and extrapolate from them.
In the *huge* chaotic system that is AS/AW such instincts are likely to be dead wrong.
Do you want to run a successful business or an unsuccessful witch-hunt? I at least try for the former, so in this case I try very very hard not to be distracted and misled by the extraordinary individual data items, CTRs and eCPM swings, etc. And I have to accept that my experience with AS/AW may well not be in any way representative of anyone else's. I suggest that you do the same, if possible.
Threads like this are not pointless, but the conspiracy theories and personal attacks are worse than pointless.
Rgds
Damon
[edited by: DamonHD at 6:09 pm (utc) on Nov. 19, 2007]
it does not help to make ad hominem attacks...
I think you need to consider the difference between an ad hominem attack and ad argumentum. You are attacking me, that would be ad hominem.
Glitches can and do happen...Maybe software changes are tried out on one or two servers before being propagated across the board?I think that's a fair statement to make. One has only to look into some of the threads on the Google search forum to see examples of glitches affecting webmasters in such a way. Google maintains many datacentres for search, so why wouldn't the same apply for Awords or Adsense?
Agreed. New Adsense features often don't roll out system-wide at the same time (some publishers see them much sooner than others) and some beta-tested features may only be seen by some publishers, so if there was a glitch or eCPM adjustment, it need not have been system-wide to have been real.
To play devils' advocate, I can offer an explanation here that shows that the conclusions are all in the minds of the beholder.
One could explains the data you mention as follows: The marketplace has been steadily declining in terms of advertiser spend (quite likely if we look at history), but Google decided NOT to transfer those lower bids to partners such as yourself, both to reward you for your loyalty but also to continue to keep publishers tied in for market share reasons.
Then more recently, they decided to finally pass on the lost revenues to you. So, the market adjustment was buffered by Google, and you were getting too much before, and now you're properly in line. Seems to explain the data just as well as any other speculation.
Do I believe the above? Nope. Is it as likely as the glitch theory, or any other one? It could very well be, although the bias against the giant g is so strong, nobodies even suggested this possibility.
We're programmed to spot unusual data points and extrapolate from them.
Yes, but this is not my point.
My point is that we are programmed to only spot unusual data points (because these stick out) while we tend to ignore small and smallest changes (because these do not stick out). E.g. we know and appreciate that the sun shines longer every day when summer approaches (in the Northern Hemisphere), but we are completely unable to tell "today the sun was shining 2 minutes longer than yesterday".
My point is that long-term eCPM is being managed by Google, with only tiny changes applied every day. This change can only be spotted if you compare longer periods of data.
But yes, I agree that this may not be seen by everyone (and I never claimed that this is a "trend" or "universally" seen). Thus I will keep out of this thread from now on. I know that some will respond pointing to whatever other reasons there may be (which may be entirely true).
Just don't take my not-responding as agreement, please.
Two peculiarities of my site:
1) The change in clickable area has happened, but the server editable ads are not yet available.
2) I have had fairly good growth over the last two months. Yesterday and today (so far) were particularly interesting because I had higher traffic, but a sharp shift in the geographical mix with a lot more UK (my target region: the site is a .co.uk) and a lot less US.
I doubt the lack of US traffic is due to thanksgiving because it did not affect me until Wednesday last year.
Interestingly though my impressions and clicks are relatively static, but my eCPM has like so many posting here plummeted 20-30%.
So don't use that rope yet, it may get better.
In that case Google is correct in its changing of the rules, good on them.
All those sites where the navigation is close to the adverts, or the adverts are so blended that they accidentally get clicked on, are finally seeing the true level of clicks they should receive.
That very well could be the case. I think google has made some strides in click validity.
true level of clicks they should receive.
true level of clicks they should receive.
I don't think I have any sympathy for those relying on trickery to generate income
And what do you say to those who have never employed such tactics yet are still seeing huge CTR reductions or massive daily fluctuations far outside of anything ever experienced in the past 4+ years?
You are making blase accusations with no proof and seriously ought to re-consider your presumptions because if this were to happen to yourselves I feel fairly confident you would be questioning what bad data push had occurred or what experiment on which server(s) they are performing.
And what do you say to those who have never employed such tactics yet are still seeing huge CTR reductions or massive daily fluctuations far outside of anything ever experienced in the past 4+ years?
Exactly. If there were any rationality or intelligence behind these changes there wouldn't be such wild and reckless fluctuations, the dishonest would just be cut without mercy as they should be.
No, this is not art, it is incompetence and a graceless disregard for long time partners. I would be more than understanding if they leveled with me about their problems, but no I'm treated like a stranger and emailed a phony form letter.
eCPM back in the pan, i.e 50% of the yearly avergae until that October maintenance weekend, EPC well below the bottom of normal trading range, clicks have all but disappeared, I actually earned more on Sunday than on my BUSIEST day of the week.
I thought they were slowly tweaking it back in-line however obviously not. Whatever it is that whomsoever has done really needs a good thrashing and I'll be first there to administer it.
I've taken adsense off a few sites with a low CTR to see if that makes any difference after Thursday (the day that my eCPM seems to be recalculated on).
Doing A/B testing on some of the blending formats that I've never used on reportedly getting really high CTR (over 20%), the old methods now seem to lead to even 70% lower results in some cases, even as the advice on these methods is not based on getting accidental clicks.
There is going to be a new wave of ad optimization techniques based on the latest changes.
There is going to be a new wave of ad optimization techniques based on the latest changes.
My ads haven't changed at all apart from the central text area not being clickable.
My leaderboard ad is obviously an ad and the clickable links are still the traditional blue and the non-clickable still black therefore I cannot believe that this introduction could possibly do this.
There has to be some other explanation for my reductions however what it could possibly be I have absolutely no idea and I get the feeling I shall never know.
As several have mentioned before "Welcome to the new Google roller coaster" albeit at much lower levels for some of us...some days!
I don't think the "clickable area" change has had too much effect, because my ads aren't blended and most users who click on them probably know what they're doing.
I don't think the "clickable area" change has had too much effect, because my ads aren't blended and most users who click on them probably know what they're doing.
I just checked some of my pages and found identical ad units, some on the same page where some are clickable all over while others are clickable only on the text link and url, so the roll-out is not complete yet.
edited spelling
[edited by: OnlyToday at 3:48 pm (utc) on Nov. 21, 2007]
yesterday was the worst ecpm since i joined in 2004, f*ckin ecpm dropped almost 40%, this new ads format is killing me
My graphical analysis shows a steady increase in traffic, but a steady decrease in all stats that relate to income - all this year.
I see no cut-off, no abrupt change in slope, no glitch. It may look like it while scanning numbers, but if you actually make the effort of producing a chart... you'll see what I mean.
It's the economy. And maybe Google is dealing with tighter economic realities as well.
I haven't removed my Google ads yet; but I will recode my websites and solicit my own advertisers in the future - as soon as next week. I already have a few advertisers lined up, and the response is positive so far.
It's a good change, in fact. I was really ashamed that my sites were used to peddle scrapers and MFA's with only a smattering of real advertisers. I plan to deal with the advertisers directly, and manage actual ad campaigns that my site visitors will really be interested in. Meanwhile, the advertisers were limited to tiny text ads - I can give them rotating images for the same money, better targetted (I know my audience).
It's bad news for Google, more than it is for me. Because once I get in the groove of finding my own advertisers... I doubt I'll ever go back to AdSense. My content network really rocks. Google will lose out, not me. I will make more money, not less.
P.S. With that said, I agree with those of you who remarked that volatility of the kind we've seen in the last month can not possibly be a good business model for Google.
My graphical analysis shows a steady increase in traffic, but a steady decrease in all stats that relate to income - all this year.
It's incredible to think that a publisher's graphical analysis could actually reveal the economical health of an industry.
I also see a slow downtrend of EPC since last christmas. Of course, I'd rather think this is due to an economical factor rather that my sites suck ;)
My graphical analysis shows a steady increase in traffic, but a steady decrease in all stats that relate to income - all this year.I see no cut-off, no abrupt change in slope, no glitch. It may look like it while scanning numbers, but if you actually make the effort of producing a chart... you'll see what I mean.
On the other hand I see no clear long term trend, but I do see a recent drop. Daily numbers were far too noisy so I looked at a rolling average as well.
The graph did help put the drop into context, because it made it clear just how common these sorts of fluctuations are. There have been several sudden jumps in my earnings, CTR, ECPM and earnings per click over the last year, most of which reversed (apart from those linked to changes to my site).
