More Evidence Can-Spam Isn't Canning Spam
from the but-why-the-discrepancy dept
Last week, we had a story about an anti-spam company that checked out 1,000 spam emails and determined that only three complied with the new CAN-SPAM law. This week, a different anti-spam software company did the exact same study and, while they came up with the same overall conclusion (CAN SPAM isn’t working), their survey revealed 102 out of 1,000 spam messages complied with the law. Either a lot more spammers started complying over the course of the past four days, or one (or both) of these studies is likely to be flawed in how they picked their “representative” spam.
Comments on “More Evidence Can-Spam Isn't Canning Spam”
No Subject Given
Either a lot more spammers started complying over the course of the past four days, or one (or both) of these studies is likely to be flawed in how they picked their “representative” spam.
Or, a third option, that there are lies, damn lies, and statistics…
1,000 emails is not a scientific representation of all emails out there, since in most cases the spam is that which the reporter receives to base their statistics on. Even if someone was able to manage to grab all the possible SPAM messages out there for one particular day, and remove all the duplicates, they would still not get a good representation of the SPAM as some spammers may only send messages on a certain day or set of days which may fall outside of the scope the author of the statistics was testing within.
Unless the statistics are based over a larger time period, with more of a representative amount, any numbers they come up with will make little sense to analyze for commonalities. The MX Logic article stated “MX Logic looked at a random sample of over 1,000 unsolicited commercial emails during the course of a seven day period beginning New Year?s Day and found only three of the messages complied with the CAN-SPAM Act.” Considering the fact that many spammers probably took a vacation after the Christmas madness, the numbers may be bogus. The Audiotrieve announcement doesn’t even give a time period, meaning that they could have started collection the 8th, or later, when spammers started spamming again.
However, if these folks continue their survey over time, and average the results, we will probably get a much better picture of the reality.
Time period
Apparently, the first study covered 1-7 January, while the second was reported on 12 January and said the emails were gathered “over the weekend” – possibly 10-11 January. A simple assumption would be that little to no effort was made to comply before 1 January (after all, it wasn’t illegal then). 1 January was a holiday – automated mailing routines ran, but not programmers. 2 January was a “hammock” day – between a holiday and a weekend, so the programmers took it off as well. 3-4 January were the weekend. Finally, on 5 January, some effort could be made to comply (after catching up on cow-orkers holiday stories). In short, fixes would start to be in place by the end of the week, so it’s perfectly reasonable that a study from 1-7 January would find only 3 of 1000 compliant, while a study from 10-11 January could find 100+ compliant.