revenue patterns

[finer9]

Here is what I am observing, and I wonder if anyone can explain:

Some days I get a more than average number of clicks and get little revenue. Some days I get a less then average number of clicks and get a lot of revenue.

However, it seems consistent within the period of one day - aka some days the per click average is .10 and some days it is $1.50

I don't understand why this would be at all...why would all the clicks within a certain time period be less valuable?

Is it just some sort of statistical type thing, since my revenue is not that high yet?

[KenB]

I frequently see the same thing. I think the spikes are new advertisers who set too high of a spending limit and then burn their cash too quickly. Last week I had a day where the average rate per click was way above normal and then the next day I got almost nothing because Google had run out of ads for my pages.

My biggest question on this issue is how to legitimately improve the overall average CRT for AdSense banners on a site? I don't want tricks from the dark side.

[finer9]

So you think it has to do with new advertisers? I could see that I suppose.

I think it is just strange that within the period of a day, all clicks seem to be generally higher priced or lower priced. But it could be an illusion of the statistical sort...

[Jenstar]

Some advertisers pause campaigns over the weekend as well.

December also brings wild fluctuations - certain areas often reduce their monthly spends, while other areas, especially for those selling products etc that could be given as gifts, tend to increase their spend. It should settle down to a more consistent pattern in the new year.

[freeflight2]

I don't think 'advertisers fluctuations' are responsible for that... I was/am a distributor of overture's+fwht's payed search results before and their revenue was very constant... it took several days or even weeks for the revenue to go up or down 10%.

[finer9]

Interesting. But also note that Over/Fwht are auction style with few bidgap, especially on competitive listings. Google has their whole weird somewhat of a mystery pricing algo and it makes me wonder sometimes...

[freeflight2]

finer: now we should have somebody running adwords on a big, long term scale (more than 1k clicks/day) confirming if he's expenses stay about the same every day or not

[finer9]

freeflight - the problem/issue is that GoogleAdwords is fixed budget...The advertiser sets a daily budget that is exact, unlike Overture.

[markus007]

i get over 200 clicks a day and i see WILD flucuations, which is wierd considering my industry probably has the most advertisers on adsense.

[finer9]

well it makes me feel better that it is not just me then...

I get around 35-50 clicks per day and earnings range from $7 per day to $35 per day

Josh

[finer9]

Here is a recent example of what I was describing

Thursday, December 18, 2003 1,441 57 4.0% $32.46
Friday, December 19, 2003 1,377 69 5.0% $8.74

57 clicks for $32
69 clicks for $8.74

wow

[runboard]

If I had this kind of CTR I'd be driving a Porshe instead of lurking in these forums :)

[KenB]

Ya, with that kind of CRT I'd quit designing websites for others and simply focus on my content sites and travel.

EDIT
Actually, with my luck, I'd drain the advertiser pool and be back to low paying ads and empty spots.

[finer9]

Well, thank you I guess. I didn't realize my CTR was that good, as I just started my network of content sites.

The CTR is very constant - between 3% and 6%. However, as you can see the rev per click is anything but constant.

I am curious to see the patterns into the new year.

Happy Holidays!

[freerunjeff]

Regarding CTR and Rev Fluctuations.

I changed the ad units from 468 x 60 Banner to 728 x 90 Leaderboard and saw an immediate jump of around 3% CTR.

About the same time google either upped our revshare or we started consistently getting higher bidders being displayed. I'm now consistently seeing 6-8% CTR on 1100+ impressions weekdays.

[loanuniverse]

finer:

Don't know if anyone has mentioned this, but you are not supposed to disclose specifics.... While you might be anonymous in the boards, there is always a chance that you might slip and let yourself be know.... Best to keep things in general terms.

[europeforvisitors]

I think it's a mistake for a content publisher to focus too much on clickthrough rates. The only numbers that really matter from a publisher's point of view are effective CPM and bottom-line revenue.

Also, getting hung up on CTR may result in editorial coverage that's limited to a few high-paying topics or subtopics. That may be risky over the long haul. What if you focus most of your widget coverage on purple widgets and Purplewidgets.com, the major advertiser in that category, stops running content ads? Or what if you devote 500 pages of your pet site to ferrets and ferrets go out of fashion? In the long run, editorial diversity within your topic will attract more readers and give you more economic security. Unless you're looking for the quick buck and are willing to start all over again when the next fad comes along, there's a lot to be said for building a content site that will grow and serve an increasingly broad base of readers as time goes by.

[finer9]

I will take note of this - what is driving the concern?

[KenB]

The main driving concern is Google's desire for secrecy just like they are with search engine algorithms. If too many people knew too much I think Google is concerned that it would have a bad impact.

[finer9]

Hm, I am not sure I agree with this. Can any mods comment if this is the reason?

[Silver Surfer]

With a CTR like that the Porche would be my winter car.

[jomaxx]

finer, the MAIN reason not to discuss specifics is that you agreed to this when you joined AdSense.

IMO they don't want data to be too specifically known because it's easier for competitors to steal both advertisers and publishers from AdSense if they know details of bids, searches, clickthrough rates, CPM, revenue share, etc.

Also AdWords is the main revenue generator for Google, so this info is extremely proprietary and sensitive. Once Google go public, it would be quite possible for a smart person to amass a considerable amount of information about Google's revenue patterns. Run a couple of AdWords accounts and a couple of AdSense accounts, network with some friends, monitor boards such as this one, and you could be trading quite far ahead of earnings announcements.

[anallawalla]

finer9 - I think here the risk is that someone could sticky you and tempt you over to a competing program (when there is one).

Small publishers don't have a good data sample to guide them why their stats range from 0 clicks to some nice number. I have found that URL drops in strategic locations - target market forums, newsgroups, email lists etc play a big part in the variable hits. Can't do a lot in the free searches in the short term.

[finer9]

Interesting. I have to admit I the #1 thing I hate most about WebmasterWorld is everyone's hypersensivity to specifics. I know they are intelligent people and must have their reasons, but sometimes I just feel that business is all about the details!