[ANN] a-ads.com: Bitcoin advertising network. Advertise now!

[arsenische]

Because when people get hit by captchas on outgoing links, they blame it on the publishing site, not the advertising one. Sort of like when sites use interstitials before you view an article. The assumption is that the publisher is who set up the ad/captcha/pay wall, and therefore it drops their readerbase, while having no negative effect on the advertisers. Essentially, what happens is this:

1) User visits site
2) User clicks link/ad
3) User gets hit by captcha
4) User says "fuck this site" and doesn't bookmark it
5) User just visits the advertiser site directly or Googles it if they were really interested

Thank you very much for the detailed explanation, but how likely is this scenario?

Let's say (X*100)% of your audience doesn't use AdBlock, (Y*100)% of them click the ad, (Z*100)% of them sees the captcha and (W*100)% of them gets annoyed in such a manner that it really damages your site (e. g. they planned to bookmark it, but due to the easy-to-pass captcha opened in the new browser tab upon their ad click they decided not to do so).

So publishers should expect to get R times higher revenue from advertisers at the cost of C=(X*Y*Z*W)*100.0% of annoyed customers.

What do you think are the realistic values of X, Y, Z, W?

According to my estimations, R is multiple orders of magnitude higher than C, doesn't it worth it? Advertisers can put their own captcha wall on their sites or advertise intersistial ads anyway (and some of them do).

PS: Advertisers that set captcha rate to 100% are potentially loosing (W*100)% of their conversions (or more, since some people won't get annoyed in a damaging way, but will just close the browser tab). Also the visitors might attribute the captcha to their sites as well (since it is opened in the new browser tab upon click on their ad).

[1713584220]

1713584220

[ranlo]

Because when people get hit by captchas on outgoing links, they blame it on the publishing site, not the advertising one. Sort of like when sites use interstitials before you view an article. The assumption is that the publisher is who set up the ad/captcha/pay wall, and therefore it drops their readerbase, while having no negative effect on the advertisers. Essentially, what happens is this:

1) User visits site
2) User clicks link/ad
3) User gets hit by captcha
4) User says "fuck this site" and doesn't bookmark it
5) User just visits the advertiser site directly or Googles it if they were really interested

Thank you very much for the detailed explanation, but how likely is this scenario?

Let's say (X*100)% of your audience doesn't use AdBlock, (Y*100)% of them click the ad, (Z*100)% of them sees the captcha and (W*100)% of them gets annoyed in such a manner that it really damages your site (e. g. they planned to bookmark it, but due to the easy-to-pass captcha opened in the new browser tab upon their ad click they decided not to do so).

So publishers should expect to get R times higher revenue from advertisers at the cost of C=(X*Y*Z*W)*100.0% of annoyed customers.

What do you think are the realistic values of X, Y, Z, W?

According to my estimations, R is multiple orders of magnitude higher than C, doesn't it worth it? Advertisers can put their own captcha wall on their sites or advertise intersistial ads anyway (and some of them do).

PS: Advertisers that set captcha rate to 100% are potentially loosing (W*100)% of their conversions (or more, since some people won't get such annoyed, but won't pass the captcha either) and the visitors might attribute the captcha to their sites as well (since it is opened in the new browser tab upon click on their ad).

This is a hard one. I'm mostly judging based on the whole CoinURL fiasco, where tons of people won't use sites that use CoinURL URLs because they're beyond irritating.

I think a case study may be in order here, and could give valuable data down the line. I'm not 100% sure how to quantify it, though, without doing a survey (which has response bias), as a straight A/B test wouldn't prove too valuable as I don't think variables could be locked on that would substantiate the claim one way or the other.

[arsenische]

How much can i earn by a website with 10,000 visitors daily ?

Oh, sorry for missing your message. It is hard to say in advance how much you can earn. Much depends on the category of your site, and whether we've seen IP addresses of the audience of your site during the last 24 hours.

Our network-wide averages with CPM estimations are here: https://a-ads.com/stats -- but your earnings could be much more or much less than that. I'd suggest to try us our for several days and then decide whether it worths it.

This is a hard one. I'm mostly judging based on the whole CoinURL fiasco, where tons of people won't use sites that use CoinURL URLs because they're beyond irritating.

What went wrong with CoinURL? I don't know their story.

I think a case study may be in order here, and could give valuable data down the line. I'm not 100% sure how to quantify it, though, without doing a survey (which has response bias), as a straight A/B test wouldn't prove too valuable as I don't think variables could be locked on that would substantiate the claim one way or the other.

To be frank with you, I don't know how to conduct a case study. It sounds hard, perhaps the easier solution could be to let publishers to disable captchas for their ad units. If they do so, they won't get money from the advertisers who want 100% human traffic. On the other hand other advertisers might benefit from it.

[winspiral]

How much can i earn by a website with 10,000 visitors daily ?

Oh, sorry for missing your message. It is hard to say in advance how much you can earn. Much depends on the category of your site, and whether we've seen IP addresses of the audience of your site during the last 24 hours.

Our network-wide averages with CPM estimations are here: https://a-ads.com/stats -- but your earnings could be much more or much less than that. I'd suggest to try us our for several days and then decide whether it worths it.

This is a hard one. I'm mostly judging based on the whole CoinURL fiasco, where tons of people won't use sites that use CoinURL URLs because they're beyond irritating.

What went wrong with CoinURL? I don't know their story.

I think a case study may be in order here, and could give valuable data down the line. I'm not 100% sure how to quantify it, though, without doing a survey (which has response bias), as a straight A/B test wouldn't prove too valuable as I don't think variables could be locked on that would substantiate the claim one way or the other.

To be frank with you, I don't know how to conduct a case study. It sounds hard, perhaps the easier solution could be to let publishers to disable captchas for their ad units. If they do so, they won't get money from the advertisers who want 100% human traffic. On the other hand other advertisers might benefit from it.

Why do you want the choice from the publisher?
i do not undertand why?

The publisher offer space...
the advertiser pays for this space...
if the advertiser pays,then it is normal that he can chose the kind of click he want...it's not the domaine of the publisher.
Why should my site only send human clcks to an advertiser...i do not know what the advertiser want...
if i see an ad...saying for a faucet...if I have to solve a captcha just for going on a faucet,I prefer go away and go on sites without captcha.But this is the advertiser problem not the one of the publisher.

[Brad Harrison]

Exactly this new feature is just plain dumb and only benefits advertisers, publishers will be screwed in the ass and visitors to those publishers sites will get very annoyed having to worry about captcha when clicking an ad, reminds me of the 90's, either you implement a better system to detect bot clicks or remove this annoying feature, stop only thinking about advertisers and worry about your publishers too! because if it's not for the publishers than you would have no advertisers either

PS. You say clicks don't count then who cares if a bot clicks it or a real person, once again you only care about the advertisers! You are only doing this so advertisers come back and spend more money but in the end you will just drive publishers away

[arsenische]

Exactly this new feature is just plain dumb and only benefits advertisers, publishers will be screwed in the ass and visitors to those publishers sites will get very annoyed having to worry about captcha when clicking an ad

How often do you think an average visitor clicks the ad? Do you think that clicking the checkbox is that annoying? Why do you think that it will have significant impact on your user loyalty?

reminds me of the 90's, either you implement a better system to detect bot clicks or remove this annoying feature, stop only thinking about advertisers and worry about your publishers too! because if it's not for the publishers than you would have no advertisers either

PS. You say clicks don't count then who cares if a bot clicks it or a real person, once again you only care about the advertisers! You are only doing this so advertisers come back and spend more money but in the end you will just drive publishers away

That's true, our publishers are the key component of our business. We want to make them happy and we appreciate their loyalty despite the low revenue. We think that they deserve to earn more than they currently do.  Or do you think we are trying to solve non-existent problem? Publishers, are you happy with your current level of income from a-ads?

Publishers' earnings depend on amounts of money spent by advertisers. That's why we want to make our service look good for advertisers.

When advertisers realize that we have 50M impressions from 1M unique addresses and 2M clicks, many of them just leave because they understand that most of this traffic is fake and it doesn't worth it. Yes, they don't pay per click, but not all of them understand that. And they don't know how many real clicks they should expect to get when they set their advertising budgets. The numbers we showed so far (like hundreds of clicks for 0.01$ in btc) were too good to be true.

Also in our system fake traffic makes more damage to legit publishers than to advertisers because it reduces their share of the advertising budgets. So even if this feature doesn't increase advertising budgets, it would allow us to reduce the income of the fraudulent publishers and increase the earnings of the honest ones.

Why do you worry about captchas so much? We don't plan to display it every time user clicks an ad. Less than 1% of your visitors are likely to ever see it, and most of them will probably pass it by just clicking the checkbox. The negative effects are likely to be negligible in comparison with the possible gains.

Yes, that's somewhat similar to 90s except the no-captcha recaptcha is much more convenient and easy to solve.

An alternative solution would be to invade visitors' privacy with cookies and javascript. Perhaps we could change the money distribution principles and switch to some form of RTB network. I feel that it would be damaging to the original idea of user-agnostic service, but perhaps it would be a better experience and better monetization.

Do you think it would be better? Or is it better to stay in 90s with no javascript and no cookies?

We need to gather more feedback from publishers and to see whether the captchas influence advertising budgets and traffic conversion. If it doesn't solve the problems it is intended to solve - we'll probably rollback this change.

Are there any publishers that find my arguments reasonable, or is there a consensus that this solution is really, really bad?

[ranlo]

//stuff here

I think your proposal of allowing publishers to choose whether or not to allow this feature is a good solution. This at least gives control back to the publishers, allowing them to evaluate the two and see what works best for them. As long as this is a possibility, I have no issues with it.

[eternalgloom]

Exactly this new feature is just plain dumb and only benefits advertisers, publishers will be screwed in the ass and visitors to those publishers sites will get very annoyed having to worry about captcha when clicking an ad

How often do you think an average visitor clicks the ad? Do you think that clicking the checkbox is that annoying? Why do you think that it will have significant impact on your user loyalty?

reminds me of the 90's, either you implement a better system to detect bot clicks or remove this annoying feature, stop only thinking about advertisers and worry about your publishers too! because if it's not for the publishers than you would have no advertisers either

PS. You say clicks don't count then who cares if a bot clicks it or a real person, once again you only care about the advertisers! You are only doing this so advertisers come back and spend more money but in the end you will just drive publishers away

That's true, our publishers are the key component of our business. We want to make them happy and we appreciate their loyalty despite the low revenue. We think that they deserve to earn more than they currently do.  Or do you think we are trying to solve non-existent problem? Publishers, are you happy with your current level of income from a-ads?

Publishers' earnings depend on amounts of money spent by advertisers. That's why we want to make our service look good for advertisers.

When advertisers realize that we have 50M impressions from 1M unique addresses and 2M clicks, many of them just leave because they understand that most of this traffic is fake and it doesn't worth it. Yes, they don't pay per click, but not all of them understand that. And they don't know how many real clicks they should expect to get when they set their advertising budgets. The numbers we showed so far (like hundreds of clicks for 0.01$ in btc) were too good to be true.

Also in our system fake traffic makes more damage to legit publishers than to advertisers because it reduces their share of the advertising budgets. So even if this feature doesn't increase advertising budgets, it would allow us to reduce the income of the fraudulent publishers and increase the earnings of the honest ones.

Why do you worry about captchas so much? We don't plan to display it every time user clicks an ad. Less than 1% of your visitors are likely to ever see it, and most of them will probably pass it by just clicking the checkbox. The negative effects are likely to be negligible in comparison with the possible gains.

Yes, that's somewhat similar to 90s except the no-captcha recaptcha is much more convenient and easy to solve.

An alternative solution would be to invade visitors' privacy with cookies and javascript. Perhaps we could change the money distribution principles and switch to some form of RTB network. I feel that it would be damaging to the original idea of user-agnostic service, but perhaps it would be a better experience and better monetization.

Do you think it would be better? Or is it better to stay in 90s with no javascript and no cookies?

We need to gather more feedback from publishers and to see whether the captchas influence advertising budgets and traffic conversion. If it doesn't solve the problems it is intended to solve - we'll probably rollback this change.

Are there any publishers that find my arguments reasonable, or is there a consensus that this solution is really, really bad?

But using captchas in ads is unheard of, and with good reason, I would never do a captcha that appeared in an ad.
There must be other ways to determine that the traffic is real or not.

[arsenische]

But using captchas in ads is unheard of, and with good reason, I would never do a captcha that appeared in an ad.

It doesn't appear in an ad, it appears in new browser tab with some probability after you click it (the probability depenends on past captcha performance of this ad unit and on campaign settings). So the network-wide click stats you see now should be more adequate (there are ~5K, not ~1M of real clicks per day).

If significant amount of users refuse to pass the captcha, advertisers' conversions should drop. If there are any advertisers in this thread - please let us know whether that is the case. I received 1 complaint from an advertiser from Russia, who noticed the decreased amount of clicks. But I'd be thankful for feedback from other advertisers as well (preferably English-speaking).

If conversions suffer significantly then we'll need to roll-back the change. Otherwise I think we should allow publishers to specify max captcha rate for their ad units. Publishers will be able to set it to 0, but if campaign's captcha rate is higher than ad unit's, then ad unit won't earn from that campaign. It may take couple of days to develop, test and deploy it.

[arsenische]

I queried the database to see what kind of statistics we got and found out that over ~1K ad units generate > 0 captcha requests per day (daily estimates are calculated as: today_fraction*today_value_of_metric+(1-today_fraction)*yesterday_value_of_metric).

An average ad unit in this set has:
  captcha display rate = 47%
  captcha pass rate = 46%

If we assume that all ad units generate only real traffic, then 47%*(100%-46%)= 25.4% of visitors who clicked the ad on the average ad unit were stopped by captcha. That is the most pessimistic expectation.  Actual stats should look much nicer since we display captchas more often to ad units that rarely pass them and also many ad units are probably generating only bot visits thus distorting the general picture.

I've split this set of ad units into 5 groups:
A) ~42% with captcha pass rate = 0%
B) ~10% with captcha pass rate in (0..50%) range
C) ~5% with captcha pass rate = 50%
D) ~12% with captcha pass rate in (50..100%) range
E) ~31% with captcha pass rate = 100%

I think ad units of the group D are the most interesting since most of their traffic is confirmed to be human and they have more than 2 captcha requests per day.

An average ad unit in this group has:
  captcha display rate = ~1%
  captcha pass rate = ~75%

If we assume that all ad units in group D generate only real traffic, then 1%*(100%-75%)=0.25% of visitors who clicked the ad were stopped by captchas. This is less significant than daily fluctuations of traffic and our CPM (that we aim to improve by improving the quality of traffic our advertisers receive).

Some more stats (just to make more sense of the data):

~49% of all ad units that have click attempts, have captcha display rate of ~1%

~84% of all ad units that have successful captcha passes, have captcha display rate of ~1%

Ad units with > 10 daily unique clicks and captcha pass rate > 30%, have average captcha pass rate of ~74% and average captcha display rate of ~2%.

[upsidedown75]

Just a beginner question. i set up an account and created a banner campaign of size 120x600. Now i want to create another campaign with banner size 250x250 but i can not find a way to do it with my existing account. do i need to create another account for another banner size? :/

[ranlo]

Just a beginner question. i set up an account and created a banner campaign of size 120x600. Now i want to create another campaign with banner size 250x250 but i can not find a way to do it with my existing account. do i need to create another account for another banner size? :/

A-Ads doesn't use accounts, so all you're doing is setting up a new banner. Not sure where you came up with this "account" thing from.

[arsenische]

Just a beginner question. i set up an account and created a banner campaign of size 120x600. Now i want to create another campaign with banner size 250x250 but i can not find a way to do it with my existing account. do i need to create another account for another banner size? :/

If you want to add a new banner size to the existing campaign then you need to create a new ad, upload all the banner sizes to it and then link that ad to your campaign (you can do it in campaign's settings).

If you want to add a new advertising campaign, then you can use campaign creation wizard (it will create an ad and a campaign at the same time).

Ranlo is right, we don't have user accounts, but you can use the same refund address for your campaigns and then find them using our site-wide search. If you sign in with that bitcoin address, you will see them too.

[arsenische]

@arsenische Hi i see some difference between some site. Some site send huge amount of unique hit and  some site send small amount of unique hit but small amount of unique hit site earn good amount of money. I don't understand what is the difference between those site. [Suspicious link removed] (Ad Unit #67832) this site send only one country and small amount of visitor but his cpm is so high and other site freebitmoney.com (Ad Unit #10539) this site send huge amount of unique hit and it comes from different country's but this site earn is so less. @arsenische can you tell me where is the difference this two site and "please can you tell me what is the algorithm you use for cpm calculate". sorry for my bad English.

Hi, thanks for your question. There are several factors that influence. The algorithm for calculation of CPM is pretty standard (except we calculate it based on globally unique impressions): 1000.0*<ad unit's income>/<ad unit's globally unique clicks>. The algorithm for money distribution is not trivial. Ad unit's income depends on a) targeting and budgets of advertising campaigns; b) competition for advertising budgets c) the relative volume and quality of ad unit's traffic (the 'quality' part is tricky and I am not going to explain all the nuances here).

Thanks for picking good examples, ad units #10539 and #67832 are quite similar. They both belong to Bitcoin and Earning online categories, they both have about the same amount of traffic, and I would also say that #10539 is embedded on a nicer site. Why does it earn ~7 times less than #67832? Let's have a closer look.

1) See the list of advertising campaigns that advertise on ad unit #67832:

(https://a-ads.com/ad_units/67832/campaigns)

Almost half of its income is generated by campaign #6468 that targets English-speaking non-US traffic.

Since ad unit #10539 has significant share of US traffic, it doesn't fit the targeting criteria of that campaign and earns significantly less.

2) See the stats of ad unit #10539:

(https://a-ads.com/ad_units/10539/stats)

You can see that it had ~1200 non-unique clicks (~400 unique clicks) a day until we deployed our solution to filter out bot clicks. When visitors click the banner then with a certain probability we display captcha and count the click only if the captcha is passed. If captcha is not passed, then the probability of captcha for this ad unit is increased.

At the moment ad unit #67832 displays captcha with probability of 1% since most of the clicks are made by humans, that's how it is supposed to work for every ad unit.

However ad unit #10539 displays captcha with probability of 99.9% since most clicks are probably made by bots. So the bot clicks are not counted and the traffic quality is considered low.

I hope 1) and 2) explain why ad unit #10539 earns much less than #67832 despite the fact that it has about the same amount of traffic and a nicer site.

[Brad Harrison]

a-ads pays extremely low! not worth it at all running them

[arsenische]

But same as again ad unit #10539 not gain more than #67832 and can you tell me how is that possible ad unit #67832 views and unique hit same without bot traffic. I say it because if visitor is real there is no way unique hit and views are same. For ex= 28/2015  ad unit #67832 send unique hit 4304 and views is 4360.

You are right. Such a little site gets so many impressions from a single country that are all unique... It looks suspicious indeed. We'll investigate it and try to improve our quality evaluation algorithm so that sources of good traffic earn more (and Brad Harrison is happy ).

[winspiral]

a-ads pays extremely low! not worth it at all running them

I agree with you absolutly.

But do not forget that it is an auction system.
if it is low for publishers then it is as well low for advertisers.

The day i have something to promote...i will chose a-ads because it is very interesting for advertisers.

I have a long experience in clicks
if you have to much clicks uniques or not for 1000 views then something is "un-natural"

of course here in faucet's country...it's a little different because many visitors click on faucets banners or faucet's rotator banners each time they see them.

But for exemple if 10 same visitors click 10 times on an ad talking about a specific car or a specific car insurance,then something is wrong.

The big problem with ads advertising is that if i click on banners about a specific faucet then winner in the story is not necessary the banner owner.
Or you see 100 times a banner on  few sites and you click suddently by an new publisher for advertisers who have paid 100 times for showing the banner.

it's why the unique view system by a-ads is not the best for every one.But all system have advantages and disadvantages for publishers and advertisers.

Publishers have choice to propose ads spaces and advertisers to bid a price...what can be better?

[EXtremeAEX]

Yeah I agree. Also the expected daily income turns out to be 0 everytime, and affiliate rewards are also 0, is it because there are no advertisers? Maybe you should really start of with a 'deep cleaning' on sites with fake traffic, or inactive sites, might be better for advertisers as well. The categories or the ratings of the site should be better rated, as they might not be accurate. Im not sure about how the system works but a-ads is not a bad advertising site, so keep it up, and make it better!

[Brad Harrison]

Don't get me wrong I love a-ads and I use them sometimes to advertise but when it comes to publishers I think it's one of the most unfair advertising systems. For a publisher to send over 2000 unique impressions and make literally a penny from it is unheard of and quite frankly a huge rip-off, so this tells me one thing, the advertiser is getting a huge advantage because there ad is showed to over 2000 people and I make a penny from it. Tell me how your system is fair for publishers?

Also the affiliate ads only benefit the advertiser since the publishers only get rewarded if an action is completed, ie. signup, survey, etc! So this means the advertisers will probably make anywhere from $1 and up per signup while the publisher on a-ads is rewarded with pennies

Another thing like the user above mentioned is go through all the sites and delete junk ones so the more legit publishers have a chance to earn more

[Brad Harrison]

Any changes made yet?