Israel has a new anti-spam law that takes affect today, December 1st, 2008. Ultimately, it seems to boil down to: Opt-in is king. Companies who don’t already live up to the opt-in standard (with appropriate tracking of opt-in permission) seem to be rushing to reconfirm their email lists. ExactTarget’s contract requirements already require opt-in, so that’s a great start. But, if you send mail into Israel and/or have a presence there, I’d strongly recommend contacting an attorney for guidance. My (very limited) Hebrew skills make decoding the Israel Internet Association’s anti-spam website a bit difficult, but it does suggest to me that getting permission in writing is a requirement, as is labeling the message as an advertisement.

There’s definitely some confusion out there regarding what’s allowed and what’s not allowed under the law, and it’s unclear to me what ability there would be to take action against senders located outside of Israel, but it’s not the kind of thing that I recommend you find out for yourself. Our recommendation is always to comply with the legal requirements of every jurisdiction you’re knowingly serving mail to.


“We want to build our list FAST! We want to put a notice on our high-traffic website that says, the first 1000 people to give us the email addresses of ten of their friends will each receive a $15 coupon code to use on our online store. Is this legal?”

Sure, it’s legal. But that’s the wrong question. Marketing success takes quite a bit more than just making sure what you want to do is legal. The right question would be something like: Is it wise? What kind of problems are you going to have as a result?

From a deliverability perspective, viral marketing campaigns + incentives = disaster.

Why?

The perception of spam is a big problem here. The “friends” are going to get this mail, not recognize that they signed up for it, and they’ll report it as spam. In a normal “forward to a friend” scenario, the volume is low enough, and the friend has enough control of the message, to minimize the associated risk. When you open it up broader, by way of asking for X email addresses, or by offering an incentive, you’re going to end up with a lot more addresses – bad addresses. Even if you don’t assume that people will try to game the system by giving you garbage addresses, the volume increase is the kind of thing that’ll hurt you, as there’ll be a greater number of spam complaints as a result.

Also – as I’ve blogged about before – any viral marketing program that utilizes incentives must comply with CAN-SPAM. That means that you have to make sure you do not send email to any “friend” who may already have unsubscribed emails from you. If you fail to scrub the submitted addresses against your internal unsubscribe records means you are failing to comply with the law. And the affirmative consent standard applies here. Meaning, if you don’t have affirmative consent from the person you’re emailing (and you won’t), then you have to label those messages as an advertisement. That’s bad news. That means you are pretty much labeling your mail as spam, and inviting ISPs to block it.

Forward-to-a-friend isn’t inherently bad, but it is risky from both the perspectives of legal compliance and deliverability success. Our recommendation is that you never offer any sort of incentive for viral marketing or similar programs, and that you never, ever have a webform that allows people to submit multiple email addresses. 


“Our terms of use specifically require that all members abide by all local, state and federal laws, including CAN-SPAM and we make sure to educate our customers accordingly,” says an email I received today.

That’s a red flag. When you brag to me about CAN-SPAM compliance, it tells me that you need something to brag about other than permission, other than clear and direct consent.

Sure, there’s spam out there that is far from legally compliant. But there’s such a thing as legally compliant spam, and it’s just as unwanted. What governs your ability to get your mail delivered? Permission, not just legal compliance. And legal compliance isn’t a substitute for permission.

Bragging about legal compliance is a bit like bragging about how fabulously shiny your subject line is. It makes for great marketing filler, but it means absolutely nothing.

If you’re bragging about CAN-SPAM compliance, it tells me that you either don’t understand how email reputation works, or you’re attempting to deceive, because you know you don’t have a solid level of permission involved.

I’m not sure which is worse, but either way, it means something doesn’t smell right. And in this case, the rest of the email message did nothing to allay my concerns. There was absolutely nothing about permission. No mention of clear consent. Just a lot of talk about legal compliance, and how anybody can opt-out, and we suppress X million addresses, and how they’re leaders on the lead generation and prospecting front.

Really? Leaders? Without permission? I don’t think so.


On May 12, 2008, when the Federal Trade Commission (FTC) approved four new rule provisions under the US Federal anti-spam law, CAN-SPAM, they also included some very important clarifying information in the related Statement of Basis and Purpose (SBP), which was published shortly thereafter in the Federal Register. The most important bit of that clarifying information relates to forward-to-a-friend messaging and how that messaging is covered under CAN-SPAM.

In a scenario in which someone either receives a commercial e-mail message and forwards the e-mail to another person, or uses a Web-based mechanism to forward a link to or copy of a Web page to another person, the FTC explains that, generally speaking, if the sender offers something of value in exchange for forwarding a commercial message, then that company must comply with CAN-SPAM.

Meaning, if you offer an incentive to people to forward your message to a friend, you have to ensure that they do not forward your message to somebody on your unsubscribe list, else you are liable for a CAN-SPAM violation.

Let’s break it down even further. Imagine this scenario:

  • You have a Forward-to-a-friend, invite-a-friend, or any type of viral marketing mechanism where a recipient has the ability to submit an email address, and you then send that person an email.
  • You offer an incentive to people to submit their friends’ email addresses.
  • You then send an email message to those friends.
  • If any of those friends were on your unsubscribe list, if they had previously opted-out from your messages, then you’ve just broken the law.

Not good!

Our recommendation is that you don’t offer an incentive to subscribers to forward your email unless you have ability to honor existing and future opt-out requests. This isn’t typical, standard functionality. If it’s something you plan to do via ExactTarget, make sure you talk to your account manager and the deliverability services team, to better understand the legal liabilities and technical requirements relating to any sort of forward-to-a-friend process.

In closing, here’s one additional thing to keep in mind: If that message you send purports to be from the friend, the friend must have control of content. In the past, the FTC levied record fines against a company who sent messages as the friend, but didn’t give the friend control over the content. This is the kind of thing where they’re likely to actively prosecute bad actors. As always, make sure you don’t implement this in a way that makes you look like a bad actor.


Project Honey Pot is a group that collects and collates data on harvesters, spammers, dictionary attackers, and other bad actors, and they make this data available for spam filterers and those looking to sue spammers.

Email address harvesting, the process of obtaining email addresses by extracting them from public information sources (such as web pages) via automated means, has long been a common way for bad guys (spammers) to build up their spam lists.

According to the folks at Project Honey Pot, harvesting is illegal under CAN-SPAM. "The law defines every message sent to a harvested address as "spam" and imposes potential liability on the sender. This is regardless of whether the sender complies with the law's other requirements. In other words, including an opt-out link and following the Act's notice regulations is not enough to spare bulk mailers sending to harvested addresses from liability."

Not everybody agrees that this interpretation of how harvesting is handled under CAN-SPAM is correct. Regardless, mining public data sources for email address clearly is not the best path. Even if it were universally believed to be legal, ISPs are quick to block spammers. Mailing to harvested email addresses is one of the quickest ways to get yourself tagged with the "spammer" label. If you want to get your mail delivered, you don't harvest email addresses. It's that simple.

For email addresses retrieved from services such as Zoominfo and Jigsaw, do opt-in requirements and CAN-SPAM requirements still apply?

Absolutely, yes. You need affirmative consent to be able to add an email address to your list, unless you want to be labeled a spammer, get blacklisted, and/or find yourself suspended from ExactTarget.

This applies regardless of where an email list has been obtained from and regardless of how an email address was obtained.

Contacts found on sites like Jigsaw and Zoominfo (and similar services) haven't opted-in to receive emails from you. If you take email addresses obtained from a service like this, and you add them to your list, they're going to report your mail as spam in very high numbers. It's going to get you blocked at ISPs. It's going to cause blacklisting issues.

I've talked to various clients and prospects on this very topic. Invariably somebody will say, "Hey, wait a minute. Jigsaw is just B2B lists. This is business contact info. This isn't B2C. I'm not sending mail to Yahoo."

CAN-SPAM (and opt-in permission requirements) still apply in the B2B realm. Keep in mind that B2B filters like Postini, Barracuda, MessageLabs, and others, they all work in a manner very similar to how ISPs work. They receive spam reports from unhappy recipients, people forwarding spam, or people clicking on a "report spam" button in an Outlook plug-in. They look at reputation measures in much the same way that ISPs do. And they will block you for spamming, just like the ISPs will do.

In the B2B world, it can be harsher on you when this happens. If you're blocked at Yahoo, you know specifically that you're blocked at Yahoo. But, if you're blocked by Postini, you're blocked by the thousands of companies that use Postini as their spam filter. It becomes a much broader issue, one that can be a lot more difficult to investigate and resolve.

The moral of the story is, when you're getting a person's email address from somebody other than the user of that email address, then you shouldn't be emailing them. You don't have permission from the user of that email address, and if you add them to a list, and send them email, you're spamming.

Contact databases are a useful tool, but not for email list building.

(This blog post was re-purposed from a question I received during a Q&A panel I participated in back on September 17th for MarketingSherpa.)

Recently, somebody wrote me, very certain that if somebody is a prior customer, it's okay to send them commercial emails (i.e. advertisements), even if you don't have prior affirmative consent (the opt-in standard defined in CAN-SPAM). The person's point was that if you buy something from me, any message I sent you afterward must be "transactional or relationship." A lengthy debate followed, but what it boils down to is this: Wrong. The law has very specific definitions surrounding message types, and the FTC applies the primary purpose test to make the determination.

Mickey Chandler over at Spamtacular agrees and takes the time to explain what primary purpose means. "[It] means just that.  It is the purpose of the email, as determined by the content of the email, that determines whether it is a “transactional or relationship” message, not the underlying relationship."

What he means there -- is that a message is transactional based on what's in the message. It's not considered transactional just because somebody is a past purchaser. Mickey goes on to say, "Just because I once sold you something (and thus established a “prior business relationship”) that doesn’t mean that I can then begin to send you email about everything I have for sale because we have a prior business relationship."

You can find his whole post here.

ExactTarget has always ensured compliance with the FTC & FCC mandate regarding wireless domains. We don't allow these domains to be imported or mailed to, as required by the law. For more information on why, you can find quick links to the FCC's pages on the topic by visiting our Wireless Domain microsite.

We at ExactTarget are not alone in our interpretation of the wireless domain requirements. Just the other day I ran across the following on Mickey Chandler's Spamtacular blog. What should best practices be, according to Mickey?
  1. You should make certain that you regularly download the list of wireless domains maintained by the FCC and wash addresses in those domains from your general lists.
  2. You should segment your wireless domain addresses into their own list.
  3. You should make certain that you implement double opt-in for addresses on the wireless segment, even if you don’t use double opt-in in general.
ExactTarget does periodically download the list of wireless domains maintained by the FCC and we include it in our list detective filter. This ensures that addresses in those domains are never uploaded or mailed to.

If you're looking for the ability to be able to send to these domains, please feel free to contact us to discuss. We can help you feel out whether or not there is a legitimate business case to proceed, and if so, how to do so in a way that's compatible with both the legal requirements, and ExactTarget's opt-in policies. The process can be a bit tricky, but we're happy to help you navigate your way through.

Here's a quick post from Alex Rubin at ReturnPath, with some useful tidbits relating to international spam laws in Singapore, China and Hong Kong.

My esteemed colleague here at ExactTarget, R.J. Talyor, was quoted extensively in this recent article in the NY Times "Bits" blog. He talks about Obama's use of SMS to announce his choice for VP, and more generally about how organizations and marketers use (and don't use) SMS currently.

SMS is a "brave new world" for me. As I learn more and more about how deliverability works in the land of SMS, I'm learning that it's a lot harder to trick somebody (like an ISP or your email service provider) into allowing spam to be sent. Sure, people still get SMS spam, but the wireless providers are much quicker to clamp down on unexpected and unacceptable usage. There are a much smaller number of wireless providers, compared to the number of email postmasters out there. The number of filters is much smaller. The sets of guidelines are much fewer. And the guidelines, the requirements are much more explicit and specific. No spam. You want to do SMS? You have to tell the providers what you're going to do with that short code. Before sending for the first time, you have to explain how you're going to use SMS, how things are going to work. And if a provider doesn't agree, you can't send or receive to or from users on their wireless network.

Somebody asked me, if the entire world converts from email to SMS, does that mean that I'm out of a job? Will there be no more need for policy compliance enforcement and best practice guidance people in the SMS world? I don't think so. I think a lot of organizations don't yet understand what you can and can't do with SMS. I think there are many, many opportunities to guide and impart best practice knowledge, just like with email. I have a feeling that even in an SMS-only world, I'd never run out of things to focus on.