Click Here To Infect Your PC!

Email me for FREE viruses writes "Just how many people would click an ad saying "Is your PC virus-free? Get it infected here!"? According to the security researcher who ran that very ad on Google for 6 months, 0.16% (409 of 259,723) would click on it. 98% of those people were running Windows. The Google Adwords campaign cost $23 in total, which works out to $0.06 per infection had the site actually been malicious."

Comments

0,16%

[JanneM]

At a click rate of 0,16% - about one in 600 - I have to wonder if not a fairly large portion is simple click errors. You intend to click on some other link nearby on the page but by mistake click that one instead. There's several kind of interaction slips just like that that we do in other circumstances after all.

Re:0,16%

[dour power]

Even those who deliberately clicked on the link could have easily read the text as, "Get it inspected here!" Not an excuse, but certainly understandable. How many /. article postings contain at least one sincere reply of the form, "Am I the only one who read that as...?"

Hmmm

[gordgekko]

It's news that at least 0.16% of people are idiots? Actually I'm shocked the number was this low. This is actually good news.

Re:Hmmm

[julesh]

Consider that click through rates to a relevant ad are typically less than 3%. This represents 5% of people who would normally click on an advert.

Not exactly.

[SolitaryMan]

The Google Adwords campaign cost $23 in total, which works out to $0.06 per infection had the site actually been malicious."

Not exactly.

$0.06 per infection attempt, which is obviously not the same thing.

Re:0,16% Mac/Linux users

[nyctopterus]

And of that tiny percentage how many were Windows users taking the fairly safe bet that the ad didn't do what it said?

Doesn't really mean much

[gazbo]

Hell, if I saw that link I'd click on it for sure. Well, I might drop to Cygwin and use lynx "just in case", but there's no way I'd not investigate such a link.

Why does it matter what OS they were using?

[Torodung]

I once explained that browser security is almost entirely determined by the user. This proves it. I wouldn't trust that 0.16% with a pocket calculator, let alone a computer!

You can't write code or design software that will secure "stupid." Firefox and Linux are certainly easier to secure, and they have a better security model, but they aren't idiot proof.

If those folks were using an abacus, they'd probably get their head stuck in it! <G>

--
Toro

Re:It's hardly a surprise

[Eivind]

The worst-case scenario are however very significantly different.

Worst-case for a virus-infected windows-machine ? Complete reinstallation. A day lost, hell make it "toss away $1000 machine".

Worst-case for std ? Death.

Not really comparable.

Re:Attempted Infection == Infection

[ajs318]

Yes, but if you didn't know what it was or whether it was safe, you wouldn't click it in your browser, would you? You'd use netcat. For example, if the link goes to http://somesite.someisp.cc/some/long/filename.ext? query_string then you'd need to do

echo -e "GET /some/long/filename.ext?query_string HTTP/1.1\nHost: somesite.someisp.cc\n" | nc somesite.someisp.cc 80

which will dump the raw HTTP response onto STDOUT. And that's safe because you can't muck anything up by printing to the screen (well, you might get unlucky and have some weirdy escape code sequence turn off echo or redefine the entire character set or beep incessantly; but the whole beauty of xterm windows is that you can always close one forcibly if you have to).

And then, if and only if it looks safe, you can use wget http://somesite.someisp.cc/some/long/filename.ext to download it for investigation.