Comments:

To: The Federal Trade Commission Re: CAN-SPAM Act Rulemaking, Project No. R411008d Commissioners, The CAN-SPAM Act is an excellent start on legislation to get the problem of unsolicited bulk email under control. There are, however, some concerns about how certain parts of the Act will be implemented. The one that's most disturbing is the possibility of applying the practice of using merchant-specific suppression lists to the sending of solicited email. (In this document, the term "solicited email" means that the recipient gave prior consent to the sending of the email, with conspicuous notice given concerning the nature of the content that would be delivered.) In the simplest implementation of suppression lists, any time someone unsubscribes from a list upon receiving an email to that list which contains one or more mentions of products or services that are determined to be commercial in nature, the address of that person must be sent to the merchan(s) involved and added to their suppression list. Anyone referencing commercial products in a way that might be construed as advertising must ensure that people on the merchants' lists do not receive the emails containing those references. There are a number of very serious problems with any such approach. They arise from the ways in which people use email very differently from other communications media, and the nature of email itself. In no particular order: 1. It is, in most cases, impossible to know the intent of an individual when they send an unsubscribe request, beyond that they don't wish to receive further email from that list at that address at that moment. People unsubscribe from lists for a number of reasons. In rough order of likelihood: The content no longer interests them. They get too much mail from that specific list. They get too much mail in general. Something in that specific email rubbed them the wrong way. They mistook the email for something it wasn't. (Spam or another publication are the most common.) They want to get that publication at a different address. They're unsubscribing temporarily because of an extended vacation or other absence, and wish to lower their email load while away. There are other reasons, but these are the most common. Very few people expect that everything they receive with any publication will be of interest to them. They read and use what is of interest, and ignore the rest. It is VERY uncommon for someone to unsubscribe from a list because of the mention of a specific product or service. If each of those unsubscribe requests, regardless of reason, leads to the sender being put on the suppression list of one or more merchants, you end up with a lot of people who might be interested in the product being unable to hear about it from the publishers whose mail they still wish to receive. With products promoted by affiliate programs (the ones most likely to be affected by inaccurate application of suppression lists), this leads to an odd problem. Let's borrow a term from the engineering fields and call it "Cascade Failure." Consider: All other things being equal, the best products are likely to also be the most widely promoted. The more widely promoted a product is, the greater the merchant's exposure to inaccurate additions to their suppression list. Every time their product is mentioned, every person who unsubscribes, regardless of their real reason, gets added to the suppression list. This could have devastating impact on their ability to advertise in or be promoted by the owners of publications or lists specific to their market. If there are more than a few publications in that market, this could wipe out some of the merchant's most valuable distribution channels, all while achieving little or no benefit to the consumer, who probably has no objection to hearing about the product in the first place. Add in the fact that unsubscribes tend not to be traceable to one specific email, and the inevitable "Suppress 'em all and let God sort 'em out" approach (the only safe one, given this scenario), will result in wholesale destruction of affiliate marketing via solicited email. This benefits no-one, and does nothing to advance the purposes of the Act. 2. It is often impossible to know which email in a series motivated the subscriber to leave the list. Most email lists publish at least bi-weekly, if not weekly or more often. People don't read all of their list mail as it comes in, sometimes saving up many issues and reading them in batches. Because of this, and because of the systems of technical operation of most lists, the publisher has no idea which ads might have appeared in the email they were reading when they decided to unsubscribe. 3. Many unsubscribe requests do not actually come from the person whose email address is in the request. Viruses grab addresses from various places on infected systems and insert them randomly in the From: and To: fields of outgoing emails. Most publishers simply assume that any address in the From: field of an email sent to their unsubscribe address wishes to be removed from their list. It's better than mistakenly leaving an address on the list belonging to someone who doesn't want to receive their mail. If the system automatically sends these addresses to the suppression list of the merchant mentioned in that message, even assuming that's trackable, a great many people will be added to the suppression list who never actually asked to be. If it's not trackable by message, one such virus-created email can result in the owner of the misused address being added to multiple suppression lists. This problem is compounded by the fact that people in specific markets tend to read the same or similar publications. They also tend to communicate with each other about related topics, so the addresses in any given addressbook or email program will tend to concentrate around one topic. Remember: Viruses don't just send one email per infected computer. It only takes a tiny percentage of the population of any market to place large percentages of that market on a lot of suppression lists without their knowledge or approval. This adds substantially to the problem of "Cascade Failure" mentioned above. Again, bringing no benefit to anyone, and not advancing the purposes of the Act in any way. An additional problem relating to the misuse of addresses in unsubscribe requests, or direct emails to the merchant requesting addition to a suppression list, is malicious forgery. It is a simple matter to use automated systems to harvest email addresses from topic-specific forums and web sites and send such requests without the knowledge or permission of the person who owns the address. People who participate actively in forums on a topic, or whose web sites discuss that topic, are also the most active buyers of products related to it. One person, armed with software that can be easily found online or created in a matter of a few hours, could devastate large sections of the market for a specific company's products or services. Again, no benefit to consumers and no furtherance of the goals of the Act. 4. There are huge problems of potential collateral damage with the way the various possible interpretations of suppression list usage intersect with the definitions of "commercial email" under the Act. Many publishers, in order to avoid having their solicited mail trapped by inaccurate content filters, will send a note to their subscribers letting them know that the current issue is online at their web site. Some will send the content via email, and later send a separate email letting people know it's been posted, in case it was blocked by such filters. With huge percentages of solicited bulk email being blocked, this practice is growing more common all the time. If they also promote affiliate products on their sites, they could seem (or actually be) required to use the suppression lists of every merchant whose products they link to. Failure to do so could well run them afoul of the suppression requirements. If this becomes the case, it will kill large segments of the email publishing industry. Specifically including those publishers who provide content that is valuable and useful even without the purchase of any of the products they advertise. When discussing this issue as it relates to mailers who send only to those who've given affirmative consent, this seems an undue price to pay, with little if any benefit to the consumer. 5. There are significant technical challenges involved in the use of suppression lists by mailers. They weigh much more heavily on the small publisher than the large commercial mailer. Many, if not most, list hosting services used by small- and mid-sized mailers do not use software that supports this function. Software that does also increases the cost of mailing. If the use of suppression lists becomes a legal necessity, it's likely that mailing houses that support them will also charge extra for their use. Add in the problem of large numbers of inaccurate and/or unintended requests for suppression described above, and you have a squeeze play that will put a lot of these mailers out of business. It will simultaneously mean the loss of much of the most valuable and desired content in many niche markets. Large mailers will face the same problems, to a somewhat lesser, but still important, degree. Mailers who use software that sends from their desktop computers and supports suppression (also called "exclude") lists will often find that their computers are unable to deal with the massive suppression files of popular merchants. Another group driven out of the industry, and more useful information lost to those who've requested it. The larger the merchant, the larger the suppression file. The larger the suppression file, the greater the processing requirements for the sending system. Thus, we have the same problem from a different angle: The more popular a merchant is, the more people will be unable or unwilling to promote their products or services, due to technical constraints. A separate technical issue is the problem of legitimate requests for suppression being lost before reaching the merchant. Lost email is becoming more and more common these days. The biggest cause of this problem is the congestion of the mail system caused by spam and the filters designed to stop it. It is not difficult at all to envision a scenario in which someone actually requests to be added to a suppression list, their mail is truly lost before reaching the merchant, and a merchant who is making every possible effort to comply is hit with the expense of a suit. This problem isn't entirely confined to people whose requests were lost. Many people use multiple email addresses that forward to one central mailbox. If they forget which address they used to subscribe to a specific publication and send their request from a different address, they can continue to receive the suppressed content even if the merchant has received and properly handled their request. If they assume it's simply a matter of refusal on the merchant's part, the same situation can occur: Suit without actual cause. For small- to medium-sized merchants, one such suit can be enough to severely damage them or put them out of business. The fear of such potential suits has already led some to stop publishing, even prior to issuance of guidelines on the matter by the Commission. 6. The administration of such lists imposes a number of significant expenses and problems for the merchant aside from that of unnecessarily lost market share, the potential for suits brought on erroneous bases, and technical challenges. The largest is the problem of avoiding misuse of the suppression file. All it would take to swamp a merchant would be for a competitor, someone with a personal grudge, or just some teenaged prankster who thinks the net should be entirely uncommercial to sign up, get their suppression file, and spam those people with ads for that merchant's wares. A public relations and customer service issue of Biblical proportions. Then there's the lure that all those addresses will present to spammers with no desire to harm the merchant. They sign up for the merchant's affiliate program, download the suppression file under guise of using it as it's intended, and slam the people who're on it with as much mail as they can send. Many people use what are called "tagged addresses." These are addresses which are given to only one sender. If they get mail to those addresses from another sender, they assume the first sender gave it out knowingly. In a case where a spammer gets hold of a suppression list with tagged addresses on it, the original sender to whom they were given can count on significant undeserved backlash. Contractual enforcement against such use could be problematic: Person A signs up as the affiliate and gives the list to Person B who spams it. There are potential technical solutions to this, but they just add another layer of expense and complexity without actually solving the problem. A smaller problem is the matter of the information about one's business that is relayed to merchants in the transmission of unsubscribe requests. Someone who understands the business can learn (or misinterpret) a lot about someone's business model from this information, and could conceivably misuse that in ways harmful to the publisher. 7. There are legal and privacy issues facing publishers who are required to give out the addresses of people who unsubscribe. When discussing a properly run list, meaning one that requires affirmative consent and has a working unsubscribe system, the subscriber is in complete control. They can stop any or all mail from any or all such lists at any time. The problems that the Act is intended to ameliorate do not stem from such publishers. Many of the best publishers have for years had a simple statement of their policy regarding sharing of subscriber addresses: "We won't. Under any circumstances." Is it within the intent of the Act that people who have assigned a right to another (use of their email address for delivery of specific content, with the promise that such use would be reserved to the holder(s) of that permission) should be required to be subjected to the potential harm described above despite the conditions of that assignment? In layman's terms, does the Act make it right for consumers to be potentially abused by forcing publishers to violate their agreements with their subscribers? Conversely, should consumers be refused the right to receive content from someone they want to get it from because they unsubscribed from someone else's list? Summary: There are other factors that suggest that the mandatory use of suppression lists is bad for consumers, publishers and merchants. The ones listed above are the most serious. They should serve to demonstrate to the Commission that suppression lists are not an effective way to solve any of the problems the Act is intended to address. In fact, there is significant potential for their use to make those problems worse. Because of these concerns, we urge the Commission to exempt lists which operate using the principle of affirmative consent from any possible regulations requiring the use of suppression lists.