Forum Moderators: martinibuster
There are several patterns in there, but one of particular interest. The lines for clicks, epc and overall earnings follow a similar pattern up to one point. Namely the point where my site traffic took off big time. What is evident from the graph is that impressions and clicks go up (following the same pattern), epc goes down, and right in the middle of the diverged lines is earnings! And you've guessed it - the more clicks I get, the more the earnings line dips.
To me this indicates that Google have software that is pretty good at predicting patterns, and I do wonder if they use this to level off the epc. If only the EPC line followed the clicks and impressions lines!
I guess the other factor at work might be that even if I can deliver lots of clicks, advertiser budgets are limited, and Google limits the EPC accordingly.
the more clicks I get, the more the earnings line dips
That was also a frustrating experience for me.
2 days invested starting June for a new optimized ad placement system.
CTR 20% up, EPC 20% down
It like with a regressive tax schedule - the more you work the less you get. Although I haven't noticed it that much in my charts.
Overall though my earnings have dropped 40% in two months even with the visitors remaining consistent or increasing across those sites with Adsense still included, CTR etc.
We have one site which has double the ad impressions of only just a few months ago and is earning 50% of what it was!
Unlike many Adsensers who seem to build sites across a broad spectrum of things they know absolutely nothing about but manage to raise an income from several of them, we only build sites for our niche markets which makes it doubly painful when tring to get more traffic etc.
I'm just wondering at which point I will cry enough and remove Adsense altogether or shall I just sit here and say "Well, a couple of thousand Dollars per month is better than zilch and I'm not going to bother optimising any more and just accept what I receive without giving Google a thought since they don't seem to be bothered about me whatsoever no matter how much I ask them for advice or opinions".
Consider these numbers: Page views were 10% of my best day, back in April, and the lowest, by far, this year. Clicks were the lowest of the year, and, get this, a mere 2.5% of my best day. This is a decline of 97.5% from just two months ago, while I made no substantive changes. In fact, I've added pages.
Here's where it gets really odd. While my CTR was at the bottom of the range, EPC was roughly 3X the average.
I've had my doubts about the veracity of Google's numbers and the reliablity of their stats for a long time. Now, with numbers that continually trend in one way - down - I am no longer concerned with the earnings and am going to make a case that they be investigated, either by my state attorney general or the federal government. Their numbers are so scewed as to not be believable.
Adsense, as far as I'm concerned is a totally manipulated system designed to keep publishers in the dark and in a state of relative contentment. I think publishers who have seen their earnings rise, plateau, and then drop off the map, need to step forward and start complaining to legal agencies or speak to their attorneys.
While it may be difficult to pin criminal charges on these manipulators at the very least they need to be exposed. Google has been operating one of the oldest scams in the book - bait and switch.
LOL, I love winning arguments and saying "I told you so" to opponents.
Anyway, back on topic... I've also seen similar trends developing. Spikes in pageviews and traffic met with lower epc, and there are tons of advertisers in my niches. We have been looking to cj and other affiliate programs to replace adsense on a lot of our popular, high ctr pages. We look at the ads displayed on these pages, find similar affiliate products and replace the adblocks.
Results are promising so far, although not enough time has passed to get accurate data, but we're testing.
K
My recent traffic has increased quite dramatically the last couple of months, so I'm trying out something. I'm wondering if I were to reduce the amount of impressions and clicks, the epc might improve, and the earnings may then be higher. What I'm trying is to load adsense ads 2 out of 3 (approx) page loads for the banner that makes me the majority of my income.
If I could just work out what percentage of adsense blocks to other blocks in order to keep the impressions, clicks and ctr at just the right level to maximise the epc................
I'll report back in a while to say how it goes.
What does that prove? Nothing, because so many variables come into play.
One thing you've got to remember is that different topics attract different ads and users. Seasonality can have an effect, too.
On my editorial travel site, for example, I'm getting more traffic on my travel-planning pages than I was during the two months before Google's Bourbon update, when a disproportionate amount of my traffic was on image-gallery pages.
Also, most schoolchildren have gone home for the summer, so buyers of travel services probably make up a higher percentage of my readership than they do during the school year (when pupils use my articles, link directories, and photos to research school reports). And because people visiting a travel-planning site at this time of year are more likely to be "bookers" than "lookers," it's likely that advertisers are willing to bid higher on many of my keyphrases and are getting better conversion rates (with the latter having a favorable impact on "smart pricing" from my point of view). All of these different factors work together to improve my traffic, EPC, CTR, and total earnings--at least until the season is over.
The bottom line is that I'm making more from AdSense by every measure than I was in the preceding months. By November, seasonal patterns will have shifted and I'll be doing more poorly by every measure (even if everything on Google's end remains the same).
I've managed to get about 40 topics (yes Opti I am knowledgable in each).
Some topics have fairly consistent EPC from day to day - meaning up or down only 50%.
Other topics have swings with a factor of 10 and many have a 100 fold increase/decrease from day to day.
Some cycle in days and some in weeks. Some go down and never come back.
So my recommendation is to have as many topics and therefore advertisers as you can. That way everything tends to even out.
10 topic-sites with 1000 views/day
than
1 topic-site with 10,000 views/day
both add up to 10,000 page views per day - but the 10 sites would on average earn more money?
Is this the general consensus?
As far as CTR growth and smart-pricing goes (looking at only one possible aspect of smart-pricing):
an advertiser's ads go on my site - they're reasonably close to the top so the advertiser pays premium for his click. Now as this ad gets clicked more - it is deemed to be more successful and the advertiser then pays less for his click. If the CTR on my adblock is high then presumably some ads will be getting clicked multiple times - therefore giving them a performance history and possibly reducing the cost of click to the advertiser - and the value of the click to myself the publisher and Google.
Does this sound reasonable? If it does then it means you are almost always going to be the victim of your own success when you raise CTR on an adblock that consistently shows the same rotation of ads for a period of time.
If you extrapolate this point even further then a possible solution when you see that EPC has declined below whatever comfort point you set is to knock-off the ads that have been circulating so that you are attracting newer advertisers onto your pages - but there is a possiblity that this is ultimately self-defeating as you will be attracting lower paid ads.
Adding new ad-blocks to your pages might also raise your EPC in the short-term as you will be expanding your pool of ads.
Another alternative of course is to ad entirely new pages to your site so that you are attracting different ads. Lots of people have attested time and time again that this has been their approach to maintaining a consistent level of earnings.
make sense at all?
I've noticed on my sites that very often the highest click values are from the worst performing blocks CTR wise.
I think your theory is the best one going right now, that the EPC is somehow tied to specific advertisers. The more ads you're displaying for them, the less they pay per click.
I am now limiting impressions (and hopefully lowering the number of clicks) in an attempt to hover somewhere just below the point where smartpricing starts penalising you for getting clicks. It might take a while of experimenting to find the best level, but it's got to be worth a shot!
Saturday is traditionally a bad day for me, but by passing a third of the traffic over to other advertisers I did suceed in lowering the clicks. I earned just a fraction under yesterday than the previous two saturdays, despite reduced clicks. The CPC is an improvement over the same days, as is the eCPM.
I know that it's too short a time period to judge in - the real test will be in the next couple of weeks, but I'm hopeful the effect will overall be positive. Also, seeing different ads from different providors might have a positive effect.
Obviously this has been pure supposition and very simplistic - there are probably a whole bunch of algorithms or at least hundreds of factors within whatever algorithm works between adwords and adsense, the ad auction process, the process of ad matching, displaying ads and factoring smart pricing. However I suspect that if this is the case then there is very probably a sweet spot.
Whether its possible to find this and maintain it with the parameters we have control of as publishers is a different question entirely.
Its either a wild goose chase or the quest for the holy grail (gosh that sounds dramatic).
Our stats agree with your theory, now can we get anyone on the "inside" to confirm it?
I did remove quite a few ad blocks from less well performing sites this week and, as would be expected, the EPC, CTR & eCPM have all gone up for the moment...fingers crossed that this may have resolved the overall earnings issue.
it is deemed to be more successful and the advertiser then pays less for his click. If the CTR on my adblock is high then presumably some ads will be getting clicked multiple times - therefore giving them a performance history and possibly reducing the cost of click to the advertiser - and the value of the click to myself the publisher and Google.
maybe not - as ever just trying to make the facts fit after the event with only part of the information at hand.
OptiRex wrote:I did remove quite a few ad blocks from less well performing sites this week and, as would be expected, the EPC, CTR & eCPM have all gone up for the moment...fingers crossed that this may have resolved the overall earnings issue.
An interesting related thread:
Removing Adsense From Non-Performing Pages
Best advice I've gotten to date
[webmasterworld.com...]
both add up to 10,000 page views per day - but the 10 sites would on average earn more money? Is this true?
