I really don't understand that mechanism that's supposed to be at work here.

People often assume that Google has infinite resources, based on the search engine side of the business, where mind-boggling amounts of data are collected and mined.

But make no mistake, the AdSense side of things is not the search side of things. Why is everything stuck on a per-account basis, rather than a per-website basis? Why is the number of channels limited to such a small amount? Why do they (unlike the search side) regularly have to shut things down for "maintenance"? Limited resources, and a high cost of changing an existing database schema are plausible explanations for all these behaviors.

So it's entirely possible that AdSense is essentially restricted to the channel structure that publishers choose for obtaining per-page data for Google to analyze. Especially since ideas like SmartPricing may have come along after the existing (highly limited) database schema was already set in stone and difficult to change.

>>>I am not talking about witch craft here<<<

lol... yes, you are.

you are gathering data at the same time that adwords has altered their program standards, which means that the current data cannot be used for comparison purposes with older data.

in other words, there are two seperate sets of control factors being used in your experiment, so your conclusions can't possibly be valid.

the premise that adsense uses channels to alter publisher payouts can't be determined at this time... or probably any other time, for that matter.

So it's entirely possible that AdSense is essentially restricted to the channel structure that publishers choose for obtaining per-page data for Google to analyze.

Sorry but no, that's not possible.

There's no reason to even postulate that channels affect anything, and this is certainly not strong evidence that they do.

danimal,
give me one stable without changes month with adsense and I'll give you control factors

There are more factors than I can count that affect my site's traffic and visitors behavior, but at very high traffic and many years, I can predict almost everything I am seeing to a science.

AdSense on the other hand is the weirdest unpredictable experience in my whole life (and trust me I have seen many years). Empirical cannot be dismissed, hell it is the only thing to go by sometimes!

I asked Google for clarification as soon as I saw this thread. The reply from AdSense support:

Please be assured that using channels doesn't
affect your AdSense earnings. We appreciate your understanding.

(Sorry if someone has already posted a similar reply from Google. I did look to see if this had already been reported, but may have missed it.)

So one of three things is happening:

• support is lying to us;
  
• this effect is an unintended consequence of the Smart Pricing code;
  
• the numbers you are seeing, ann, are caused by something other than your change to your channel setup.
  

I'm in no position to guess which it is.

You'd think they'd tell you truthfully if it did? Or perhaps they don't think it does, out of ignorance of the algorithm or not given correct information themselves.

How's everybodys earnings today? Because I, after deleting 30 unused channels last night, am having the worst day in 2 months :(

and you think ann's suggestion has something to do with it?

[edited by: Hobbs at 9:47 pm (utc) on July 18, 2006]

No I'm not suggesting that at all. The channels were unused after all and really needed to go. It is too early to freak out yet but today's a write off (pending maybe a massive click dump).

This has been a really bad month for my site. Maybe 1/3 of the normal.

Because I, after deleting 30 unused channels last night, am having the worst day in 2 months

I might experiment. After this thread started, I realized that I've been watching my channels and really DON'T KNOW what I'm watching. There is NO pattern emerging after one year of watching. I can have high clicks in one particular channel one day and the next day, it's the worst! So to quote Hobbs....

AdSense on the other hand is the weirdest unpredictable experience in my whole life

ditto... it's voodooooo

I'll be watching this thread closely. :)

Some of you still don't understand the way to go about it. It is all or nothing at all, there is no inbetween or half way measures.

Here is what I did:

1. Delete ALL Custom Channels (keep your URL channels if you wish)

2. find the pages that were being tracked and CHANGE the ad to a non channel ad.

3. On those pages use just one ad

4. Give your site at least a week to see a trend.

5. add more one ad pages as you can (giving more page views to enhance your sites productivity)

6. If traffic is low then concentrate on getting visitors.

But remember, do it all, steps 1 through 4, or don't bother trying as this method works ONLY with the steps above.

(still seeing good numbers today)

Ann

I appreciate you revealing your theory Anne.

Initially I was sceptical.....but then who would have believed 1 badly converting site could smart price a whole AdSense account with multiple sites :)

Foxtunes, Same here. :)

Ann

Thanks for sharing, Ann. I removed my unused channels two days ago and yesterday I had nearly the highest eCPM in over 2 years. Anecdotal evidence for sure, but no one here has offered any proof that it doesn't have an effect.

Your theory is certainly no more illogical than the idea that removing adsense from entire websites can increase income overall, but I think this is generally accepted as true now.

The only element in your advice that makes me question your results is that removing all but 1 ad block on pages has been known to increase revenue, (another theory which defies logic) so part of your increase could be due to this.

We need to maintain a list of adsense advice which may defy logic but has seemed to help some webmasters. Your advice would certainly make that list.

>>We need to maintain a list

dataguy,
See post #:3013043

These eCPM hikes mentioned by percentages, are we talking 200% jump from $1 to $2, OR are we saying $10 became a $20? I'm curious to see if this strategy works across the board or to a certain layer within the food chain... Going from 1 to 2 most likely wont raise any flags, however, going from 10 to 20 I would think will sound off the the air raid horns. I have nothing to hide, but damn I hate hitting the G radar.

Dataguy. NO NO leave your good ads be! if you have the extra real estate use it...give them as many page views as possible.

I did every suggestion I read about in here and each time the results hung around about 2 to 3 days EXCEPT for this time, 18 days good and counting. Not every day higher than the other but enough that I can SEE how the numbers did what they did and the pattern of growth they have begun. LOL

Yippie, if the site is clean and goes strictly by the TOS then don't worry about red flags. Yes I had a 400% increase at the start then for a week daily increases of 2 to 8 dollars which has now slowed also but the trend upward has begun. I was so low by the time the test was started I was ready to give up on adsense so I really don't sweat the "flags" as long as I am "legal". :)

Before the disaster (Disaster = I discovered the wonderful world of custom channels) I was a very high earner with a steady pattern of growth. I am seeing some of that coming back. Just call me Ripley, believe it or not. :)

Ann

Ann, I'm joining your experiment. I have a major site redesign underway, so it's a good time.

I have always been very skeptical when google tells us something is good for us. Given their track record, I always assume that it will do more for them than us.

My adsense earnings are not a major portion of website revenue ... yet. So I'm really not risking much, and there may be an upside.

I leave on vacation tomorrow (but with wifi enabled notebook in the backpack) so it will take time to implement.

lol this is my favourite WebmasterWorld thread in months. Wishing you the best of luck Jeepers.

The side effect to following Ann's advice may be even more valuable.

If I stop using custom channels, I will greatly reduce the amount of time I waste messing around with them!

And that means more time spend on more productive things.

3. On those pages use just one ad

If you did all these things simultaneously, how do you know the whole effect isn't due to #3? Sorry if you've already answered this, but I got dizzy scrolling when trying to catch up with the whole thread. It's pretty well accepted in some circles that content sites benefit from only showing one ad unit - on our site, we usually place that unit at the bottom of the content.

I took the leap yeaterday, so it's too early to tell, especially in one of the slowest times of the year. The truth of the matter is that I had about 15 custom channels that overlapped, some that were hardly used and I rearely even looked at the data. When I did, it made little difference.

Sometimes, the simpler way is better. I, for one, would prefer to know roughly what I am going to make if I take a certain action. With Adsense, it's always more like a roll of the dice. At least they always pay on time...

[quote]Sometimes, the simpler way is better. I, for one, would prefer to know roughly what I am going to make if I take a certain action. With Adsense, it's always more like a roll of the dice. At least they always pay on time... [\quote]

That's how I feel as well fearlessrick. My new strategy is to have really good content, put one leaderboard at the top of the page, below the header and above the start of the content, and a search box at the bottom. That's it.

I have never really used the competitive filter, except when ads were really off base (trying to sell parrot tulips, getting ads for grey parrots).

I'm rearranging my site though, so I won't really be able to say that this change did much of anything, except it will cut down on time spent looking at channels.

Ann wrote:

3. On those pages use just one ad

Ann, if I can ask a question about this truly intriguing test, do you mean that you didn't increase the number of ad units per page -keeping 1 ad unit as before-, or that you reduced the ad units -from 2 or 3- to 1 per page, at the same time of removing the custom channels?

Thank you for sharing your results with all of us.

[edited by: Juan_G at 8:56 pm (utc) on July 19, 2006]

I put one adlinks on the pages that were formerly in channels.

The others are all staying the same but I am going to add more pages and use just one ad on each of them. This isn't difficult to understand...just that a lot of people expect Google and adsense to be complicated so they may think this is too...really, it is simple to do..the hard part is the waiting! :)

Ann

Ann,
So you are tracking by URL now only, and you added Adlinks - How do you know the adlinks aren't increasing your revenue w/out custom channels?

because I have been using adlinks quite a bit over the past year.

Ann

Guys this discussion is extremely interesting. Please report back if you try Ann's Theory.

Ann your theory has certainly a merit. Thank you verymuch.

Take care of your health. Where are you located. Do you have access to good medicine?