Brilliant... never even thought of it...

http://www.captchaad.com/

this is a german start-up doing the same thing.

You mean... brillia-NOT?! The main use of CAPTCHAs is to distinguish between human input and that of a (potential spammer's posting) robot. If the textitem to be entered is posted in graphically-unskewed form, and more or less in recurring space in ad-plates from easily detectable server(s), and that technology gains any foothold, it will be easy to crack by mechanical means (as already has been done even to graphically "disturbed" text strings). So where's the alleged brilliance of that.

ianf: The brilliance is coming up with the business model: make a customer acknowledge the brand in order to get some kind of "reward". Just think, this business possibility has existed as long as captchas have. If any of us had thought of it a few years ago, we could have commercialized it.

As to the easier crackability, it's probably not an issue. Perhaps over 90% of captchas are protecting low-value targets and can use an easier-to-crack mechanism. Obviously, high-value targets won't use captcha advertising. But lots of sites (including this one) don't need to use the most secure captcha possible.

Next target in Google's shopping list?

andy, srssly

Roger Brown said:
"If any of us had thought of it a few years ago, we could have commercialized it."

That is not true. I thought of a much better way many years ago. I have never commercialized it.

I wonder who owns the patents on this.

> I thought of a much better way many years ago.
> I have never commercialized it.

Well tell me your "much better way" and if I agree it's much better, I'll commercialize it!

If I had thought of adapting captchas so that they would simultaneously build brand awareness or convey marketing messages, I'd have commercialized that for sure.

So this is already a "business model" ready to be monetized? Don't you think it's too little to suffice for that? And what about the downsides. Here's one I can think of straight away.
http://blogoscoped.com/files/captcha-ad.png
Quite apart from a string like that "$180 savings" being too long, thus too prone to typing errors, I think any attempt to make users effectively parrot something to build a brand will backfire. Or would you still be comfortable with a CAPTCHA like, say, "Brand-X rules"?

I am trying to avoid revealing details. If it is commercializable, why should I give it away rather than save it for future development?

If Roger can demonstrate he has sufficient resources, he can send me a proposal through Philipp. If he represents reCaptcha, then he probably has sufficent resources, and I may be able to meet him in person.

If I thought of this, probably many others have also. There are many reasons to not commercialize, publish, or even patent an idea. It does not matter if Roger agrees whether it is better. His claim had no justification.

Does anyone know of a web page that uses the AdCopy captcha?

reCAPTCHA says they serve more than 30 million captchas per day.
They also have several demonstrations and examples.
http://recaptcha.net/

Regarding reCAPTCHA, there's a dramatic demonstration of how this has improved the transcription accuracy of two example documents:
http://recaptcha.net/digitizing.html

The below quote is from the re CAPTCHA site.
http://recaptcha.net/learnmore.html

"About 200 million CAPTCHAs are solved by humans around the world every day. In each case, roughly ten seconds of human time are being spent. Individually, that's not a lot of time, but in aggregate these little puzzles consume more than 150,000 hours[sic] of work each day."

The idea is appealing, but I think that ultimately it will fail because brands are not supposed to be associated with a chore. Imagine being a regular commenter at a blog, and as it often happens you get repeated advertising. You end up having a negative brand experience.

Build the thing in February 2008 and called it captchatising.com.

Guess I was ahead of my time. ;-)

Proof:
http://bomega.com/2008/02/13/new-projects-wordpress-plug-in-captchatisingcom/

Boris: much props to you! And your version even had an advantage for the end user by letting them use their existing brand recognition to make the captcha easier for them to solve.

George R:
> I thought of a much better way many years ago ... If it is
> commercializable, why should I give it away rather than
> save it for future development?
It doesn't work like that. The internet moves too fast. Ideas "saved for future development" just rot, either because they become dated and no longer relevant, or because someone else recognizes the potential and patents or develops them first.

> Roger ... can send me a proposal...
Not me. I have nothing to propose. I don't have the imagination to come up with new ideas like this.

The concept is not entirely new indeed. I would refer to Boris Veldhuijzen van Zanten also, he beat me :-). In 2008 the idea was introduced.

For reference:

http://www.websonic.nl/nieuws/022008/captchatising_adverteren_in_capthas.php

http://webwereld.nl/nieuws/49934/veldhuizen-van-zanten-zet-in-op-captchatising.html

Boris: By the way. What happened with Captchatising? Does not longer exist?

http://www.adcaptcher.com/ is the same thing :)

The problem with advertisement captcha example given above (with the Dish Network ad) is that bots could easily use optical character recognition read the text "$180 savings" and pass the CAPTCHA. The text is not obfuscated in any way. Additionally, if the CAPTCHA only displays one ad at a time, then the advertiser or the website would need to come up with millions of different ads to display, in order for it to be secure against brute force attacks or random guessing from bots. Likewise, if you're relying on that advertiser to come up with different phrases to be entered for the security check, you're not going to have enough different options to make it secure against bots. Having a handful of different ads or a handful of different phrases is simply not secure.

A better option is to for a website or advertiser to place branded photos within a grid of photos like in this captcha: http://www.confidenttechnologies.com/products/confident_CAPTCHA.php

Because it is a grid of randomly generated photos that change every time, there are many more combinations possible, making it much more secure against brute force attacks from bots. The website or advertiser could have a few of their branded photos (or photos of their products or services) included among the other photos on the grid. They get their advertising in front of the website visitor, and the visitor could be required to find the advertiser's image among those on the grid and click on it, but at the same time it's more secure against bots. Personally, I would also be more willing to simply identify a photo of a product and click on it, than be forced to type a cheesy slogan.