Building Better Spam
henbane writes "Cringely is plugging a new method of advertising from Dr. Jim Kowalick and Mario Fantoni. Their book entitled 'E-Mailing Your Way to Sales With
the Taguchi Approach' is out in the autumn. What could be worse than a method which increases the returns on spam?"
...this would give anti-spam developers insight as to how to improve spam blocking techniques.
The simple truth is that interstellar distances will not fit into the human imagination
- Douglas Adams
This improvement isn't about targeting specific gullible who are more likely to respond. It is about an easy, rigorous way to fine tune what the spam says to better play on one's gullibility. And the analysis is quite cheap, so it is well worth the effort.
Or in your analogy, they are still hitting as many people with their better carpet bomb, but sustain more fatalities.
If Taguchi works as well on spam we can just about forget another spam control methods!
(This sig intentionally left blank)
As an advertiser in search engines and other mediums, this would be a great way for me in increase my conversion rate. With a tool like ConversionLogic Keyword Tracking one can now use the methods described, and accurately measure the worth of a search or affiliate campaign based on different versions of ad copy produced.
But yes, spammers will be reading this with interest as well :-)
Newsfollow.com
If spam is illegal is certain jurisdictions, wouldn't sale of this book in those jurisdictions be akin to inciting criminal behavior? What would be the financial liabilities of this? (Obviously IANAL.)
Bob Pease wrote a detailed refutation of a voltage
regulator circuit design which was optimized with
the Taguchi method and published in Electronic
Design magazine in the late 80s. The resistance
values in the circuit just looked fishy and his
analysis revealed that the circuit would not work.
The input voltage would track the output voltage.
The author had made certain the performance was
independent of the quality of the parts alright. A
fair argument could be made that the author did
not properly apply the Taguchi method. Bob's point was
the output has to depend on something. In this case, it
depended on a zener diode. The author thought he was
accomplishing something by making the output
independednt of the components. He didn't consider
that the circuit wouldn't work then. So be very
careful with this Taguchi stuff.
Wansu, th' chinese sailor
But consider the worm. There is some fairly cool technology in there to get them to work right. Right now that technology is being used mostly for evil. In the future the technology may mature to the point where we will wonder how we lived without it. As an example, instead of a slashdot web site, there could be slashdot worms that find you and deliver your content you want in real time to a computer or handheld device instead of you needing to query for it. It would simply be there on your computer. No more world-wide-wait. Don't laugh it could happen (or maybe not).
Similarly spam is using technologies in creative ways for evil. Yet the same techology is being used here on slash dot to notify me when one of you wonderful people comments on my comments. As the technology matures we may see new and wonderful things grow out of what we call spam; again maybe not.
However, both of these evils are responsible for a whole new set of technologies the thwart them. As the spam filters get better, the spam improves to overcome the latest improvements. The spam filters react and create new technologies which the spammers respond to and so forth. Worm and virus protection works the same way.
Each iteration requires using some existing techology in a new way or creating an altogether new techology. These technologies start to be applicable in other domains, thereby increasing and improving the general level of technology in the world at large.
As I gaze into the crytal ball, I see that the current war of spam and virus, like all other wars, will create new techology that was not considered and would not have been conceived had it not for the necessity of the times.
See James Burke "Connections" for a similar view of history, only much better.
Your friend and well-wisher
m0smithslash
http://www.ferociousflirting.com
They also rely on the orthogonal test cases to be reasonably linear, at least between the starting point and the optimal solution. If the system is non-linear, it may find a locally optimal solution, separated from the global optimum by a "saddle point". Determining whether the system is locally linear may not even be possible in open-ended systems, such as human behavior.
In this particular case, the history of ad presentation (user fatigue) is a troublesome variable. The classic VW (old) Beetle ads were very effective when they first came out, but became less so when their originality wore off. Tastes and styles change. The Taguchi experiments should be ongoing, to track these social tides.
Taguchi methods are well suited to optimizing familar problems with known inputs. It is not the answer for unexplored new ground. It is the tool for the maintainer of the status quo, not the changer of pair o' dimes.
Here is what companies need to do if they wany any positive return from me:
1. Make sure that I want to read e-mail that contains ads. For example, take $10 off my net bill every month and I'll opt-in to receive some offers that interest me. Limit the qunatity of such offers to a reasonable number.
2. Make professional e-mail ads. Please no heavy graphics, javascript, foreign characters, random subject lines that read "!!!!!!! Women will beg for more!!!!!!!," etc. Also, make sure that the companies that you represent are responsible for what they sell.
3. Provide effective methods for cancelling the service. If I want to unsubscribe, I must be able to do so without clicking through thousands of pages and user agreements.
This method has worked for me in the past. I was shopping for a leather jacket last fall and my girlfriend signed up for notifications from J.Crew. She received an ad about a hidden online sale, I went to the site and purchased a leather jacket for a reasobly low price (it was cheaper than anywhere I shopped in the area). I had no problems doing it because I knew that it was from a legid company that offered me a good price on what I wanted to buy.
Employ a fleet of people, working from their home like Google Answers or Expert Exchange experts, to browse usenet, blogs, web journals, forums, etc, and write individualized emails from personal accounts selling products. These people would have a product database of products from companies who are willing to pay for the promotion.
In other words:
- Jack blogs about this segway recall and comments that he isn't going to return his.
- Jill, a PersoSpam promoter reads this while browsing the web.
- Jill's PersoSpam Firebird extension automatically matches the text of Jack's blog with SegwayPaniers.com, a business in the product database that sells paniers for Segway HTs.
- Jill decipher's the Jack's obfuscated email address (jackDOESNTLIKESPAMrepressedanger.com) and writes an innocent sounding email to him about how she was just reading his weblog, and she just got these cool paniers, and Jack should totally check them out.
- Jill BCC's outgoing@persospam.com
- PersoSpam deposits $0.50 in Jill's account, bills SegwayPaniers.com for $0.60 and keeps the extra.
- ???
- Profit!
Might work, eh?Erik
The reason spam is so annoying is that we get so much of it, for things we aren't interested in. If we only got email about things we were actually interested in, not only would the senders get a much better response rate, we probably wouldn't even consider it to be spam.
I'm surprised that I haven't seen any discussion on /. about the new trend in spamming described in this wired article:1 282,60564, 00.html
http://www.wired.com/news/technology/0,
After I read it the day it was posted at wired.com, I didn't quite know how much it was prevalent, since I had never seen one appear on my home computer.
That is, until I had to switch off my firewall for a short time to try solving a problem. Almost immediately a pop up ad exactly of the same type described in the article appeared on my monitor.
The problem I was having with my firewall was with the game Earth & Beyond. When the game loaded, my firewall pop'ed a window to allow me to set a permission for the game, but it forced me to alt-tab to the desktop to access the firewall popup, and that would sometimes cause the game to freeze or lose sound.
So I turned off my firewall, and restarted the game, only to have the game freeze and crash, returning me to the desktop with a nice Windows messenger pop-up ad.
Bloody annoying.