Archive for the ‘Deliverability’ Category

Taking Action on Inactive Subscribers

Tuesday, January 3rd, 2012

I don’t normally talk about industry averages because they rarely apply to any individual company within an industry, but for illustrative purposes here I’ll make an exception. The number most commonly offered up as an industry average open rate is 20%, and industry average click-through rates are typically reported between 2% and 4%. Your results may vary of course. But given these numbers, 80% of recipients on average do not open a message, and 96% – 98% do not click on any given message. Now some of these unresponsive people were responsive last week or last month, but it turns out that for many email marketers, a sizable percentage of the non-responders are chronic, and have not opened or clicked on a message for months or even years.

These are your inactive subscribers. You know them well enough to have them on your list, and they know you well enough to not unsubscribe. But for some reason or another, they’re just not that into you anymore. Some maybe never were, filtering your messages from the very outset into a folder that receives little attention, or even signing up with an email address they never check. Since the dawn of email until very recently, we had no real incentive to do anything special with these subscribers other than continue mailing to them hoping one day to get through. The cost of each subsequent message is very low (particularly compared to the upside from a click-through) and email success is a function of absolute activity, not relative. 200 click-throughs has always been worth more to your business than 175, even if you have to mail to 10,000 more people to get there.

The landscape has changed, however, and engagement metrics are now beginning to impact deliverability. What that means is that the unresponsive people on your list are making it harder to reach the responsive people. So it’s time to take a look at who these people are, and decide what you’re going to do about them, or else risking some downward pressure on your deliverability. There are three options to consider:

1. Remove inactives from your house file
2. Move inactives to a separate group for different treatment
3. Leave inactives in your house file and treat no differently

I’ll take a look at each and provide some context that might help you make a decision.

1. Remove inactives from your house file.
Taking the unresponsive people out of your house file altogether will result in an immediate lift on your engagement metrics. For example, if 5,000 people from your 20,000 name list has not opened a message in 13 months or more (the time frame I typically start with, since some responsiveness is geared around annual events such as membership renewals, annual conferences and holidays), simply omitting them from your next mailing can boost your open and click-through rates by 33%. If you’re at industry averages, your 20% open rate that drove 4,000 opens previously will still see the same 4,000 readers, but from the 15,000 people remaining on your list. You’ve climbed to 26.7%, sacrificing no absolute results and increasing the percentage of your recipients who interact with your message. This lift in engagement metrics signals to ISPs and email administrators that you’re a legitimate sender, decreasing the likelihood that your messages will be shuffled off to junk folders or bounced – a bona fide upside across your entire email program. The downside is that there is always the risk that you’ve assumed incorrectly – that someone who appears inactive is still seeing your messages in the inbox and waiting for the one that grabs her attention. The longer someone is inactive the less likely this is the case, but it is still something to factor into your decision.

2. Move inactives to a separate group for different treatment.
This approach mitigates the downside from option 1, in that you’re not cutting inactives loose entirely. Instead, you’re targeting them like you would any other segment of your list. Only instead of sending them messages based on what they have done or bought in the past, you’re sending them messages based on what they haven’t done. One approach is a dedicated win-back campaign, where you expressly point out that you haven’t heard from them and try to re-engage in some way. Another is pare down the frequency of the messages, only sending them the ones of the greatest strategic importance. These may be the announcement that registration for the big conference is now open, or changes in content or products or features that may appeal to them and shake them back into activity. Like any kind of targeting, this approach requires a little extra effort. But you will recognize the same lift in engagement metrics on the messages they don’t receive, which will still signal to ISPs that you’re on the level and your messages deserve to pass the velvet rope.

3. Leave inactives in the house file and treat no differently.
Technically, this is the “do nothing” approach and is a viable option based on your circumstances. If your deliverability and engagement metrics are already strong, despite a significant number of inactives in your list, maintaining your status quo could be justifiable. Remember that even unread messages carry a branding impact. If you do elect to leave your inactives in place, make sure that your subject lines still speak to them and work towards re-engaging them. Telegraph your message content so that even subscribers who only glance at their unread messages still feel like they’re in the loop with your brand.

An important consideration for any of these options is where your inactive subscribers came from in the first place. People who actively subscribed to your list or made a purchase showed a significant level of engagement at one point, and are more likely to be re-engaged in the future. But if your inactives are names acquired through a channel with low engagement (such as a list of trade show attendees you exhibited at, or a purchased list), the chance of them suddenly awakening to your brand after a year of ignoring it are slim. Better to cut them loose and lift your deliverability and engagement metrics.

Posted in Best Practices, Deliverability, Engagement | Comments Off

Whose Address Is It Anyway? The Pitfalls of Subscriber Sharing

Tuesday, May 3rd, 2011

It is not uncommon for a company or association to operate under several brands. For example, a publishing company may have a dozen or more different titles, and an association might have several different divisions. One of the potential advantages for this type of organizational structure is the opportunity to cross-market: the publishing company can sell its subscribers on some of its related titles, and the association can program its conferences so that they appeal to members of several of its divisions instead of just one. Email plays a big role in these cross-marketing efforts, but organizations should be aware of some of the potential pitfalls of email cross-marketing and subscriber sharing.

Is it legal?
If Publication A has enlisted an email subscriber, who actually owns the permission to email that person – Publication A, or the parent organization? It depends on what the CAN-SPAM law defines as “Affirmative Consent.” According to the provision, an email address can be shared by multiple parties as long as there is “clear and conspicuous notice at the time the consent was communicated.” You have seen this in practice, I’m sure. When you sign up for an email list you might see two options, such as “( ) Please subscribe me to email communications from ABC Company” and “( ) Please send me offers from partners and affiliates of ABC Company”. In our publishing example above, if the subscription option says “( ) Please subscribe me to email communications from Publication A” then only Publication A can mail to that subscriber. If cross-marketing is an objective, the publisher would need to change the language to clearly communicate that Publication A and other publications owned by Publication A’s parent company will email the subscriber. Alternatively, the organization might employ a second subscription option reading “( ) Please also send me communications from other Parent Organization titles that are relevant to me.”

Does being legal make it acceptable?
Your organization may follow all the CAN-SPAM rules to the letter, providing “clear and conspicuous notice at the time the consent was communicated” without fail, giving you every legal right to cross-market to your heart’s and division budgets’ content. But your objective with Affirmative Consent is not simply to abide by the law; it is to ensure that your subscribers anticipate all of the communications that you send them. If someone subscribing to Publication A signs up for email from sister publications at the same time, it does not necessarily mean that the subscriber will realize she has a relationship with your other titles, or recognize them when they show up in her inbox. If this happens, even though you’re on solid legal footing, your subscriber could ignore the message, delete the message, unsubscribe, or even mark it as spam. Being legal may make you right, but being right won’t guarantee you are successful.

Subscriber sharing pitfalls
So what happens if a cross-marketed subscriber doesn’t respond as intended to a message from another part of the organization?  No matter what her response, it has ramifications on the rest of your email program. Here’s how:

- If she ignores or deletes the message, that means she has had no engagement with it whatsoever. Increasingly, deliverability is impacted by engagement metrics. As greater percentages of subscribers fail to interact with messages, a sender’s reputation will begin to drop, which makes it harder for any messages to any subscribers to get through. This is a change over the past year. Previously, there was no danger in emailing to unresponsive list segments; today and in the future, there is a deliverability penalty to be aware of.

- Unsubscribing means that she would no longer receive messages from the cross-marketed publication, which is ideal for the reason above – it means you are no longer mailing to this unresponsive address. But she may also unsubscribe from all the organization’s lists, either wittingly or not. This means that not only is the opportunity to cross-market gone, but the original publication that earned her permission in the first place can no longer mail to her.

- Marking a message as spam is particularly problematic for business-to-business emailers and trade associations. When a subscriber marks a message a spam, an internal email administrator or spam monitoring service is notified. When it happens often enough, the sender could be blacklisted from the subscriber’s entire domain. In the case of b-to-b, this would mean that the publication can no longer reach this subscriber or any of her colleagues. And depending on how the sender’s IP addresses are structured, it could mean that none of the publications can reach any of the people at that particular organization. The sender could become invisible to the entire company.

How to avoid the pitfalls
The best way to avoid these pitfalls is to over-communicate. When your subscribers sign up, tell them clearly what they’re in for. When you send a cross-marketing message from an organization they do not have a direct relationship with, use the opening copy of the message to tell them why they are getting this message, and make it very easy for them to unsubscribe just to this type of message, keeping their original relationship with your organization intact. Remember, your objective is not solely to remain above the law; it’s to nurture and grow your relationships with your subscribers.

Posted in Best Practices, Deliverability | Comments Off

An Engagement Proposal

Tuesday, February 8th, 2011

Andrew Barrett is Sr. Director, ISP Relations & Deliverability for Real Magnet

For all of the vast resources the big four free inbox providers (Yahoo!, AOL, Gmail and MSN/Live/Hotmail) have expended over the decades to tweak, test and improve their filtering and delivery processes, they’ve been remarkably quiet about it. But their reticence is for good reason; if they were to publish the juicy details, they’d be undermining their own efforts, as spammers would surely make use of the information to evade filtering changes almost as soon as they are implemented.

So when a major inbox provider actually publishes specific details about how they make delivery decisions – as Gmail recently did in a paper explaining how Priority Inbox works – it creates a splash in the deliverability pond.

In January, Google posted on one of its research blogs a four page document containing the mathematical algorithms Gmail employs to decide how important any particular message is to its recipient by gauging the probability that the recipient will act quickly on the message.

It’s nothing short of a mathematical model of recipient engagement.

The real utility of the paper, in my opinion, is that it gives senders a clearer picture of what engaging e-mail looks like to the ISPs. I propose that Priority Inbox is a reasonable proxy for any recipient domain where it comes to measuring engagement. Senders who test and optimize their mail for preferred placement in the Gmail Priority Inbox should see gains in deliverability most anywhere else.

So what does engaging e-mail look like to Gmail? I’m not mathematically savvy enough to evaluate the algorithms in the paper, but the author provides some useful narrative:

Labelling: First, mail that the recipient marks as important is definitively important. The real trick, as the author notes, is how to infer importance without explicit recipient labelling. To do that, the algorithms measure hundreds of “features” of the message that fall roughly into one of three categories: Social, Content, and Thread features.

Social features: This has nothing to do with content-sharing activity on social networks, as some might assume. Instead, it’s a measure of the degree of interaction between the sender and the recipient, like the percentage of a sender’s mail that is opened by the recipient.

Content features: Gmail looks for keywords in the content, header and subject lines of the message that correlate with higher opens (or lack of opens).

Thread features: Gmail looks at aspects of the thread (or “conversation”) to which the message belongs. Has the recipient taken action on other messages in the same thread? If so, what percentage of the overall number of messages in the the thread were acted upon? Did the recipient actually start the conversation? If so, then Gmail assumes the message is important.

Beyond a treatment of the features of a given message, the author addresses a pair of interesting aspects:

Important or merely interesting: Opening a mail is a strong signal of importance to Gmail, but the author notes that many users also open a lot of mail that is “interesting” rather than “important”. So Gmail’s math tries to calculate where to draw the line between the two.

Time to action: The real goal of all the math is to predict the probability that a recipient will act on the mail within a given number of seconds following delivery to the inbox. The less time it takes, the more likely the mail is to be scored as “important”.

Part of the problem, of course, is that where Gmail places the message in the inbox directly affects how long it takes the recipient to notice and act on the message. Gmail tackles this, too, but at this point in my reading of the paper, my slide rule burst into flames.

This mathematical treatment of engagement metrics offers a rare glimpse beneath the kimono of a major inbox provider, and it confirms in a general sense much of what the deliverability community has been able to infer largely through observation: engagement metrics are playing an increasingly large role in automated delivery decisions at recipient domains. Senders who are serious about improving delivery (and, ultimately, ROI) will test and optimize their mail for engagement.


Posted in Deliverability | Comments Off

Half-True Predictions

Tuesday, January 18th, 2011

The thought of blogging predictions for the New Year in Deliverability makes me cringe. I’m not exactly sure why; it may have something to do with the fact that everyone else has already done it – we’ve had a bumper crop so far this year. Mostly, I think, I’d hate to be proven wrong later. So, instead of offering my own predictions (which would doubtlessly sound much like anyone else’s), I’d like to take a look at two different trends in deliverability that came only half-true in 2010, but that are still worth your time to continue to watch in 2011. The two trends are Domain Reputation and Engagement Metrics.

If you can recall what it was like to switch mobile carriers before the winter of 2003, you’ll understand why domain reputation is a big deal. If you wanted to switch carriers then, you had to get a new mobile number, because the local number portability rules for cell phones hadn’t yet been enacted. Coworkers, friends, relatives, and vendors all had to be notified of the change if you wanted them to stay in touch with you. It was a royal pain — painful enough in many instances to keep folks from switching carriers in the first place.

Today, sender reputation is largely associated with the IP address from which e-mail originates, and not with a business or brand. If a sender wants to migrate their e-mail program off of the desktop onto a hosted platform (like Real Magnet), or change ESPs, they’ll have to leave their sender reputation behind and start from scratch because the move means they’d be sending from an entirely different IP address.

Domain reputation is like local number portability: a reputation score is associated with sending domain (e.g., mail.yourbrand.com) instead of a dotted quad (209.108.90.17). On an Internet in which reputation is assigned to a domain rather than to an IP address, sender reputation can follow the company or brand, for better or worse, across changes in originating IP address.

Domain reputation was one of the hot new things that was supposed to gain traction in 2010, but that mostly didn’t. To be sure, ISPs are filtering on domains and URLs in the content of e-mail messages (for sites that are known to host malware or to be associated with spam, for example) but that are not necessarily domains that belong to the sender. Some of the big free inbox providers use their own, internal sender domain reputation systems for filtering, but these are still miles away from the kind of granular reputation scoring data that is available for IP addresses.

One key reason why domain reputation hasn’t enjoyed widespread adoption is because it relies on the use of sender authentication protocols to tie reputation to the sender, and lots of senders aren’t bothering to implement those protocols. But just last week, the smart folks at Gmail announced that businesses who host their mail on Google Apps can use DKIM, a particularly flexible authentication protocol, just by ticking a box in their application settings. If other large mail hosts follow suit, domain reputation has a chance to pick up steam in 2011.

The other prediction that became only half-true is the broad adoption of detailed engagement metrics by ISPs for automated deliverability decisions. Hotmail and Gmail made a huge splash last year by incorporating some engagement metrics in their filtering processes. Priority Inbox at Gmail, and Sweep at Hotmail/Live give preferred delivery – and even preferred positioning within the inbox – to mail that the recipient has opened or clicked through from before.

But beyond those two standouts, and only in a fairly limited way, none of the other large ISPs or reputation scoring firms seems to have done much more than experiment with collecting and using the data. The incremental improvement to their existing filters does not appear to have justified the huge overhead associated with the collection and management of all the additional data. For a time, we’d heard that ISPs were measuring all kinds of engagement metrics – like social network sharing activity, and even how long a recipient appears to spend reading messages from a given sender. But beyond the opens, clicks and spam complaints, other engagement metrics just don’t seem to be all that much more useful for deliverability.

That’s not to say that senders don’t need to worry about engagement – engagement remains incredibly important. Those clicks, opens and complaints remain significant components of sender reputation. Tracking engagement is critical to improving ROI. Senders who perform even a modicum of content or subject line testing gain incredible insight on how recipients are interacting with their mail, and can learn much about what will generate the best conversion rates. But when it comes to getting into the inbox, detailed engagement metrics do not appear to be on track to becoming as important to recipient domains as blocklists, content filtering, and IP reputation scores.

Detailed engagement metrics and domain reputation scoring are two big trends of 2010 that happened only half-way. But don’t get me wrong – they remain important concepts for deliverability, and successful senders will keep their eyes open for broader adoption in 2011.

Tags: ,
Posted in Deliverability | Comments Off

Street Legal E-mail

Tuesday, January 4th, 2011

In this third set of questions following our recent deliverability webinar, we’ll try to clarify some confusion about the current legal state of affairs where bulk commercial e-mail is concerned. We’re also about to see some big changes go into effect in Canada that may have some impact on your e-mail strategy. We received this question from a webinar participant after the live session (if you haven’t caught it yet, you can still see the recorded version):

My e-mail is CAN SPAM compliant, but it still gets bounced or filtered. It’s not spam if it complies with the law, right?

First, let’s be clear: CAN SPAM does not actually make spam illegal, a common misconception among businesses that are new to e-mail marketing. Here’s a quick, simplified checklist of what the law actually requires of bulk commercial e-mail soliciations:

Don’t lie about the content or the source of the mail: If you’re sending an advertisement for a product or service, it has to be obvious that you’re mail is a solicitation. For example, senders can’t send mail purporting to contain photos from an uncle’s birthday party, when it really contains a sales flyer.

Provide clear instructions for opting out: Online opt-outs must use a single web page to accomplish the unsubscribe request. Forcing recipients to log into an account before they can opt-out is a no-no. Any opt-out mechanism (like an unsubscribe link) must remain functioning for at least 30 days, and opt-out requests must be honored within 10 business days.

Tell recipients where you are: Senders have to include a valid physical postal address in the body of the e-mail. Your business location or headquarters should appear here. A registered post office box is fine, too, as are any of the mailbox rental firms that are established under Postal Service regulations.

Perhaps what’s most notable about this short list of requirements is what’s missing: a prohibition from sending spam (howsoever one chooses to define the term). So, even if your mail is fully CAN SPAM compliant, that doesn’t necessarily mean to the ISPs or to recipients that your mail must not be spam. In fact, ISPs see millions of unsolicited bulk e-mail messages (a common definition of spam) every day that fulfills each requirement imposed by CAN SPAM, and they devote enormous resources to filter it.

So, CAN SPAM requirements actually represent the bare minimum for e-mail marketing standards, not the guarantee of delivery to the inbox that most newcomers assume it should be. To answer the question directly, then: mail that is CAN SPAM compliant can still be filtered or bounced by ISPs. In fact, CAN SPAM includes separate language that holds ISPs harmless when they filter mail.

What about the new Canadian spam law? Do senders in the U.S. have to abide by the law if they send to recipients in Canada?

Canada recently passed the world’s most stringent anti-spam law late last year, covering a broad range of electronic messaging, and it is expected to take effect in September of 2011. The Canadian law does what CAN SPAM never did: it requires senders of e-mail within or into Canada to have or to obtain explicit permission from their intended recipients. For most ISPs and recipient domains, it is a lack of permission that turns ordinary commercial e-mail into spam.

In theory, the Canadian law is enforceable in the U.S., though it wouldn’t be cheap or easy. Canadian plaintiffs would have to obtain a judgement in Canada, then find a court with jurisdiction in the U.S. that’s willing to enforce it. This requires a great deal of time and expense, so enforcement is likely to be rare. But if you’re already CAN SPAM compliant, and have implemented other best common sender practices, you’re likely already in compliance with the Canadian law (once it takes effect). Check my earlier blog post for a more complete analysis of the Canadian law.

That wraps up our brief look at spam laws in the U.S. and Canada. In our next installment of the deliverability webinar questions series, we’ll look at various types of content filtering, and what senders can do test their content for optimal deliverability.

Andrew Barrett is Sr. Director, ISP Relations and Deliverability for Real Magnet.

Tags: ,
Posted in Deliverability | Comments Off

Getting a Handle on Soft Bounces

Tuesday, December 21st, 2010

Welcome to the second in a series of posts focusing on questions we received from participants during and after our recent webinar on the topic. (If you missed it, we’ve posted a recorded version online, and we’ll announce when we’ve scheduled our next live deliverability webinar.) Last time, we talked about open rates and metrics. Today’s questions focus on soft bounces. After the webinar, an attendee e-mailed me the following question:

How many times should a recipient soft bounce before I suppress them?

Before we delve too deeply for an answer, it’s worth taking a moment to explain what a generic soft bounce is, and what the recipient mail server may be trying to tell you when they send it.

A generic soft bounce often indicates a temporary deferral, or de-prioritization of your e-mail. Unlike specific bounce types (e.g., user not found, mailbox full), a generic soft bounce does not indicate a recipient-specific condition. Rather, generic soft bounces are a sign that the receiving ISP has a problem with the mail or with the sender.

It often happens that, in the middle of the server-to-server transaction in which mail is presented to the recipient domain for delivery, the receiving server decides that it has accepted all of the mail it is prepared to accept during the current transaction. The balance of the mail in that transaction is then soft bounced, without any additional resources consumed by processing or filtering the mail, and with no regard to the recipient address, deliverable or otherwise.

There’s a range of possible reasons why an ISP will generic soft-bounce mail. Some of the major free inbox providers – like Microsoft Live Hotmail and Yahoo! – limit the volume of mail it will deliver from a sender during a given period of time based on reputation, how many messages in the send are addressed to non-existent accounts, and other considerations. Once that limit is exceeded, they’ll bounce any additional mail until the rate falls below the limit.

ISPs are very reluctant to share how they calculate these types of limits, because they don’t want bad actors to use that information to game their systems and evade their filtering processes. Instead, senders should try to implement those practices that improve sender reputation, so these limits can be raised or lifted entirely.

So, back to the question at hand: the generic soft bounce rate should not be a consideration in a decision of whether to suppress a particular recipient address. If you’re seeing a large number of generic soft bounces even after resend attempts, take it as a sign that it’s time to focus on best sender practices.

A follow-up question: what about “mailbox full” bounces?

Unlike the generic soft bounce, the “mailbox full” soft bounce is specific to an individual recipient e-mail address, and it means pretty much what it says: the mailbox is full, and any additional messages addressed to it will continue to bounce until the owner makes room by deleting some mail or adding capacity.

Senders can continue to send to recipients that bounce with a “mailbox full” message, but these bear very close scrutiny. A full mailbox can be a sign of an abandoned account (or one that is about to be abandoned), especially if the address is hosted by one of the big, free inbox providers, like Gmail or AOL.

Persistent attempts to send to full mailboxes can damage your sender reputation. The thinking goes something like this: if the sender continues to send to the same full mailboxes month after month, they may not be particularly careful with other aspects of their lists. ISPs know that even confirmed opt-in lists run into full mailboxes – it happens all the time. But they also know that senders with weaker permission run into the problem more often. Either way, the longer an address continues to bounce “mailbox full”, the less likely it is to become deliverable again, and it should be suppressed.

Senders can be somewhat less aggressive in suppressing full mailboxes at smaller receiving domains, like corporate e-mail accounts. Lots of corporate IT administrators aren’t as diligent about deactivating the mailboxes of former employees as senders would like, and there’s usually no reputation damage associated with them. From a ROI perspective, though, if you think a particular “full mailbox” should really be a “user not found” because the employee has moved on, it’s a good idea to suppress it.

That wraps it up for today’s questions. During the webinar, we touched briefly on what the law requires of senders. There have been some significant developments on the e-mail legal front since then, and in our next installment, we’ll take a deeper look.

Andrew Barrett is Senior Director, ISP Relations & Deliverability for Real Magnet.

Tags: , ,
Posted in Deliverability | Comments Off

Canada Passes The Ten Million Dollar Spam Law

Thursday, December 16th, 2010

Earlier this fall, we alerted our readers to the imminent passage of Canada’s strict new electronic messaging bill. On Tuesday, the Canadian Senate voted to adopt the legislation, and it was enacted by Royal Assent yesterday.

Our neighbors to the north may be the last of the G8 countries to adopt an anti-spam law, but it is the very strictest, creating penalties of up to 10-million Canadian dollars (or just under 9.87-million U.S. dollars) for businesses who send spam into or within Canada.

The new law, dubbed the “Fighting Internet and Wireless Spam Act” (or “FISA”, for short) imposes new requirements on senders of just about every type of electronic messaging, including mandates that stretch well-past the minimal requirements for e-mail under the U.S. CAN SPAM Act of 2003.

Under CAN SPAM, senders are required to abide by a series of labelling requirements, provide a working unsubscribe mechanism, and honor unsubscribe requests within ten business days. CAN SPAM, however, has never required that senders obtain prior consent from recipients. FISA requires either explicit permission, or implicit permission in the form of an existing business relationship or a conspicuous publication of the recipient e-mail address. If the publication of the address is accompanied by an instruction not to send unsolicited e-mail, it doesn’t count as implicit permission.

FISA creates a two-year window from the date an address was collected with implicit permission to try and convert it to explicit permission. If after two years explicit permission is not obtained, the sender must suppress the address. Both CAN SPAM and FISA explicitly preclude sending to addresses that have been automatically “harvested” from web sites.

CAN SPAM grants enforcement powers to the FTC, and gives ISPs the right to bring action against infringing senders themselves. FISA, in contrast, provides no criminal penalties, but allows both ISPs and individual recipients of spam to pursue civil action against senders.

The requirements seem to create significant new hurdles for senders, but authors of the Canadian law insist that the legislation is aimed squarely at only the worst of the worst offenders. FISA includes a “due diligence defense”, in which senders should not be held liable for violations if they can show they were making reasonable efforts to abide by the law when the offense was committed. Honest mistakes won’t count against senders.

Should U.S. senders be worried about the new Canadian law? Obviously, the law doesn’t apply if you’re not sending to recipients in Canada, but senders may not always know where (geopolitically speaking) the owner of a particular address receives their mail. However, if you’re already abiding by CAN SPAM and best common practices, you’re likely already in compliance.

The short answer is that (in theory at least) FISA is enforceable in the US, though the process is neither simple nor cheap. It takes about as much time and and money to obtain a judgement in Canada as it does in the U.S., so enforcement action is likely to be as rare, and therefore reserved only for the most egregious of offenders.

Canadian plaintiffs would also have to find a U.S. court willing to enforce the judgement, which is by no means a given. However, there is an open pledge between the U.S. and Canadian governments to support law enforcement efforts across borders. Earlier this year, for example, a Canadian court was willing to enforce a judgement obtained by Facebook in a California court against a Canadian spammer who racked up $873-million in fines for CAN SPAM violations. It will be instructive to see whether U.S. courts will be willing to reciprocate once FISA is enacted.

Within the e-mail community, the new law is regarded as further evidence of a trend in which legal requirements and best practices appear to be converging, albeit at a glacial pace. The take-away for senders, then, should sound familiar: adhere to CAN SPAM and best sender practices. Send to those who have granted permission, and try to engage with and obtain permission from any segments for whom you do not have it.

Andrew Barrett is Senior Director of ISP Relations & Deliverability for Real Magnet.

Tags: ,
Posted in Deliverability | Comments Off

All About the Opens

Tuesday, December 14th, 2010
Last week, Real Magnet hosted its first Deliverability webinar – if you missed it, you can catch the recorded version any time. We spent about 40 minutes looking at a few key concepts that senders should understand to begin to actively manage and improve deliverability. We received a number of compliments from attendees after the session, but we also received many more questions – far more than could be answered in the time allotted for Q&A after the session.

Today’s post is the first in a series that will attempt to answer some questions we received during (and after) the webinar that we just didn’t have time to address, or answer more completely. We’ll focus on a pair of questions about open rates.

What’s a good open rate for e-mail marketing? What kind of open rates can I reasonably expect?

This is a question we hear often from senders who are just getting started on their e-mail strategy. A quick check of the search engines will show that some of that data has been available from time to time, but it’s not a good idea to use these as a gauge for the success of your own program.

There are so many variables in the way senders approach e-mail – even within the same industry (non-profits, insurance, accounting services, etc.) – that the data is not terribly useful as a benchmark. This type of data rarely addresses some of the qualities of the mail it’s reporting on – information that’s far more important than the category of industry:

  • Are they sending newsletters or more promotional mail, or some mix of these (and in what proportion)?
  • What is the quality of their lists, and what “flavor” of permission have the recipients given to the sender?
  • How much effort does the sender devote to list hygiene, and to subject line and content testing?
  • And how does all that compare with what you are actually doing or planning to do?

Without a way to qualify these and many other variables, it’s very hard to get an apples-to-apples comparison. But if I had to produce an average open rate number upon pain of death (or unemployment), I’d say that a sender in any industry who strictly adheres to best practices, including confirmed opt-in acquisition methods, should reasonably expect to see between 15% and 25% of their mail opened.


What kind of technologies do ESPs use to track when e-mail is opened? How reliable are they?

At Real Magnet, we insert a transparent image file into each message as it is built immediately prior to the send. The image is only a single pixel in size, and each image file is uniquely named so that it is associated with a specific message and individual recipient. The images are stored on our server, so that when the recipient opens the mail and their e-mail software displays it, we record the call to our server to download the image file. We make a note of the time and date of the download, and add it to your tally of opens. Just about every ESP of any size uses a nearly identical process to record and track opens.

When a recipient reads mail with images turned off, the image file isn’t downloaded and the open is not recorded. However, when the recipient clicks on a link in your mail, we report both the click and an open.

It’s difficult to know when a recipient opens or previews a message with images turned off. Some e-mail software or recipient domains always turn images off by default. Many of the older models of Blackberry smart phones don’t render images at all. Most ISPs and e-mail clients give the recipient the option of changing the display settings to turn them on by default or on a sender-by-sender basis.

Many of the most popular e-mail packages and free inbox providers will display images in mail from senders who appear in the recipient’s Contacts folder or address book. That’s why it’s always a good idea to ask your recipients to add your From: address of your e-mail to their address book.

In the next installment, we’ll answer a question from a sender about what to do about recipients for whom mail always seems to bounce, and what to do with recipients who seem to have stopped opening or clicking through on your e-mail. Until then, keep engaging!

Posted in Deliverability | Comments Off

When Blacklists Die

Wednesday, November 17th, 2010

Blacklists have been popular targets for complaints and criticism for years. Senders complain they are too stringent and lack transparency. The anti-spam community howls with outrage when they’re not as aggressive as they think they should be.

One blacklist in particular, called the five-ten-sg.com block list, has been a thorn in the side of ESPs since 2001 – but not because lots of ISPs use the list to block mail. In fact, they don’t; the list generates too many false positives, and as my colleague Al Iverson demonstrated a few years ago, you’d get significantly better results by randomly blocking any mail from an IP address in which the number 7 appears.

The list operator is a guy named Carl Byington, and I’ve been reading what he has to say about spam and e-mail for years. He’s a smart, reasonable guy who’s always been honest about the nature of his list. He lists sources of bulk e-mail for a broad range of reasons, and he’s quick to agree with anyone who points out that his listing criteria are not useful for filtering decisions in a high inbound e-mail volume production environment. But it’s his list, and he can do with it what he pleases – and ISPs and other network operators are similarly free to ignore it.

ESPs, on the other hand, have been getting an earful from their customers about Fiveten for a long time.

When a sender runs into deliverability problems, they’ll often turn to web sites that offer to look up an IP address on a bazillion block lists all at once. In altogether too many instances, they discover they’re listed by Carl. They’ll fire off a few angry e-mail messages or phone calls to their poor, harried deliverability guy. It always seems to take a few days to explain why the listing is almost certainly not the root cause of their deliverability issue, and to redirect time and energy back to the real issues.

Not all lists are created equal. Some are more important than others, because ISPs find them more useful, and so they become deployed more widely. In the scheme of things, Fiveten is not that important, and a Fiveten listing has no measurable impact on deliverability.

This weekend, Fiveten went dark. On Friday, any lookup at the site yielded a response reading “blackholes.five-ten-sg.com has been retired.” As of this writing, the domain doesn’t answer at all. Carl hasn’t provided any public explanation for his decision to decommission his list, and he really doesn’t have to. No one has to pay money to use his list, and maintaining a list takes more time, energy and resources than most folks realize. I suspect Carl simply ran out of one or more.

Senders have a love-hate relationship with blacklists; they do a good job of keeping the deluge of pill spam, virus and malware messages at bay, and are an important reason why e-mail remains a viable channel for marketing and commerce. But when senders find themselves at the pointy end of a listing, they often feel as though the listing must be capricious, or even malicious.

The demise of Fiveten demonstrates that, contrary to all the complaints over the years, block lists as a category generally are not capricious. It turns out that market forces are as immutable for block lists as for any business, and block lists operators are just as answerable. Over-aggressive listings are not useful to ISPs, because they tend to generate false positives by blocking wanted mail. When a list isn’t useful anymore, ISPs stop using it, and it goes away. Just like Fiveten did.

Blacklists will continue to exist and operate much as they always have, and I predict that both senders and anti-spammers will continue to complain about them just as loudly. If either side were to stop – well, that’s when I’d start to worry whether blacklists are still doing a good job.

Andrew Barrett is Sr. Director, ISP Relations & Deliverability for Real Magnet.

Tags: ,
Posted in Deliverability | Comments Off

Yahoo! Refreshes the Inbox with “Minty”

Tuesday, November 2nd, 2010

With Yahoo!’s major release of interface upgrades last week, three of the big four free e-mail inbox providers have now implemented important changes in the ways recipients can interact with mail. Only AOL has yet to go live with the full complement of new features currently planned. Earlier this year, Hotmail rolled out Wave 4 with Sweep and Time-Traveling Filters that penalize “gray mail” by sweeping it out of the inbox. Gmail’s new Priority Inbox now rewards messages that it deems most relevant to the recipient with preferred positioning in the inbox.

Yahoo!’s mail release, dubbed “Minty”, was rolled out to their customer base last week, with a slew of front- and back-end changes. The most noticeable changes are to the user interface, with a leaner, faster-loading (and more purple!) site that more closely resembles the mobile apps (read: iPad) versions of its offerings. The interface changes are designed to increase market share in non-US markets, where consumer broadband is not as common, and load times can still be an issue for larger percentages of the user base.

In addition, Yahoo! is giving stronger positioning to an existing disposable e-mail address feature. AddressGuard allows users to create disposable e-mail addresses to protect their main e-mail address. The disposable addresses can be handed out in lieu of their primary Yahoo! address, and mail received to the disposable addresses are delivered directly to the users’ inboxes or other folders (where they can be subjected to existing filtering). However, the user can disable the address without impacting mail to their real inbox any time they want, if they feel the disposable address is receiving too much spam.

AOL will take the wraps off of their own set of upgrades (named “Project Pheonix”) later this year. Few details about specific changes to the user interface have been confirmed, but insiders say we should expect tighter integration with AOL Instant Messenger, SMS messaging and MapQuest. They’re planning a Gmail-like archive feature and enhanced search across users’ e-mail folders. AOL say they will present fewer ads at entry points, but will serve up more targeted advertising when users drill down into AOL’s content sites.

A recent interview with AOL Mail Ops President Brad Garlinghouse also hints at a larger effort to deliver an integrated messaging platform for all of the users’ various inboxes – be they traditional e-mail inboxes at Yahoo!, Gmail or MSN, as well as IM, SMS or social media inboxes. He notes that Internet users now manage an average of 2.4 e-mail addresses, up from 1.9 five years ago. Garlinghouse intimates that Phoenix will position the AOL mail site as a hub for users to manage them all.

It doesn’t appear that either Minty or Pheonix include any major changes to the way they filter or deliver inbound mail. The message for senders is that social media and e-mail are converging quickly: Google’s abortive attempt to socialize e-mail with Buzz did not spell the end of other efforts, and it’s possible that more than one will succeed. If that happens, senders may have a better chance to turn their recipients’ entire social networks into prospects.

Andrew Barrett is Sr. Director, ISP Relations & Deliverability at Real Magnet

Tags: ,
Posted in Deliverability | Comments Off

« Older Entries
Newer Entries »