When the script doesn't work

DJ asks in the comments of Friday’s post:

As Seth said, great reminder. For those that have great processes/channels in place, I’ve found incredible success. However, sometimes I’ve found my answer on Twitter (i.e., @godaddyguy). Also, there have been times where I’ve gone through the script (i.e., shaw.ca) and have never heard back. What then?

This is the reason ISP relations folks get paid the big bucks, DJ! To sort these kinds of problems out.
Generally, if I’m having problems getting a response from an ISP, and a client is having problems delivering mail I sit down with the client to have a discussion about how important that domain is. We look at data including percentage of addresses on the list affected by the block and the monetary value of those addresses on the list. If the client decides that this is a block worth getting lifted then we try to determine the type of block. Content based blocks are different than URL based blocks are different than IP based blocks.
If we decide the block is probably content or URL based then we start making changes to the email to see if that resolves the issue. Most of these kinds of blocks are “bucket” blocks and tweaking things can often keep the bucket from overflowing. Senders have quite a bit of control over what is happening for this type of block, and usually can restart email delivery without contacting the ISPs just by changing how they send email.
If the block is IP based, that’s a little harder to resolve without ISP involvement, but if the block is IP based and the ISP does not respond to requests for help, then that itself is a very telling piece of data. It usually means that something the sender has done has convinced the ISP that the sender is a spammer. The ISP isn’t interested in resolving the issue. If the ISP is large, this can be a problem for the sender and may take enlisting the help of a 3rd party to act as an independent negotiation. If the ISP (or domain) is small, the sender may find it easier to write off those recipients and just ignore the block.
My client base is a mix of issues. Some of them get advice from me and blocks get resolved without having to contact the ISP at all. For those who have hard blocks and ISPs won’t talk to them, most of my time is spent collecting data from the ISPs and providing the client with a roadmap to externally visible changes. Once the changes are implemented, then I approach the ISP on their behalf. Usually, if there are real changes in sending, the ISPs are much more likely to listen, whether I’m using their scripts or using a backchannel.

Related Posts

Unsubscribes made difficult

Dennis blogs about his experience trying to unsubscribe from classmates.com list over on deliverability.com. His experience touches on a number of points I have discussed recently.
Dennis initially signed up for a free account at classmates.com around 10 years ago, but has asked to be unsubscribed multiple times. Recently classmates reactivated his subscription again, sending him marketing mail he did not want. Reactivating subscriptions is an extremely bad idea. Not only is it a CAN-SPAM violation to send mail after an unsubscribe has been received, but senders really end up annoying recipients by doing this. Think about it, these are people who have actively told the sender that they do not want mail, and the sender goes out and decides to override the recipients wishes.
I can only imagine how horrible the delivery for this mailing was. ISPs measure how many non-existent addresses senders attempt and mailing a list that has addresses accreted over 10 years is going to have a massive number of dead addresses. Not that many people have the same address now that they did 10 years ago. Some of those dead addresses are probably now being used as spamtraps by the ISPs, another hit to delivery rates. Finally, there are the complaint rates to consider.
For those people who received the mail and want to unsubscribe, Classmates.com does everything possible to discourage that. Dennis describes the process he went through.

Read More

Email and the Obama Campaign

Late in the summer there were people talking about the spam coming from Senator Obama’s presidential campaign. At that time, most of the discussion was focused on the open subscription form on their website and that there were some individuals who had been fraudulently signed up and were now receiving email from the campaign.
Last week, the Senator’s campaign again became a topic of discussion among some anti-spam groups. The maintainer of one of the more respected public blocklists and members of his family received mail from Senator Obama’s presidential campaign at their personal addresses. Because the mail was unsolicited and met the qualifications for listing, the sending IP addresses were listed on the blocklist. In response, the campaign’s ESP started moving the Senator’s mail to other IP addresses, resulting in those IPs also being listed on the blocklist as well.
I talked with the blocklist maintainer and I believe that his address, and those of his family members, were added to the Senator’s mailing list as the result of an email append. All of them are registered Democrats and they all live in a battleground state.
This may have made for good campaign strategy, not being an expert I cannot comment on that. It is, however, very poor email marketing strategy.
First, the campaign decided to appropriate permission to send email. There is not ever permission associated with an email append. Just because you have a name and a street address does not mean that you have permission to send email. In very, very limited circumstances, an opt-in append (click here to continue receiving email) may be acceptable. However, that is not how appending is normally done.
There is no pretense of permission to send email. Just because someone is registered to a particular party does not mean they want to receive email from that party.
Second, when the campaign started seeing delivery problems they started sending off different IP addresses. Moving IPs around is out and out spammer behavior, no questions asked.
Now, I know this is a very hotly contested election and I know that some people believe that any method of getting the word out is good. I also expect that there may have been some positive reaction from recipients. The overall reaction, based on the IPs changing, may not have been so positive.
Do I really believe that Senator Obama is a evil and willful spammer? No, not really. But that does not change the fact that the Obama campaign seems to be sending email without the permission of the recipient and seem to be attempting to evade blocks by moving IP addresses.
From a marketing perspective, the campaign may be using email effectively and doing everything right. But from an email delivery perspective, they are getting many, many of the basics wrong and are looking like spammers in the process.
Other news and blogs that talk about spam from the Obama campaign:

Read More

Marketers missing out

Many delivery blogs have posted about the recent ReturnPath study showing that marketers are missing prime opportunities to use email to develop a strong relationship with recipients. I finally manged to get a few moments to read through the study and comment on it. Over a few days in February ReturnPath researchers signed up at more than 60 major retailer brands. They then monitored the subscriptions to see how often and what kind of mail the retailers sent.
Overall, it seems the researchers were disappointed in how the retailers were using mail. Even the title of the whitepaper captures this feeling: “Creating Great Subscriber Experiences: Are Marketers Relationship Worthy?” The answer seems to be more no than yes.
From my perspective the data is not all that surprising. In many cases it seems bigger companies rely on the recognition of their brand to get them through minor delivery problems (like complaints) rather than good practices. Whereas a smaller company will have to work harder to develop a relationship, larger companies with wide brand recognition can fall back on their brand.
There were a few areas ReturnPath measured.

Read More