Thanks for having me, Derek. I’m excited to talk email marketing deliverability for awhile.

Well I’ve been in tech my whole life. I’ve always been fascinated by computers. I began at Ontraport about seven years ago as a customer support rep. We had a really small team of about five people, and everyone wore a lot of different hats. Because of all those different hats, I got to find out what I had a strong affinity for, and what I was good at. That happened to be email marketing deliverability, and test email marketing. I think it’s interesting because for many people it’s a really dry, boring subject, but it’s so important to our customers businesses that I just gravitated towards it. I could talk about it all day.

There’s one that stands out above the others, and that is treating email marketing deliverability as a reactionary problem. People don’t care about deliverability until it’s an issue; until they get blacklisted, or until they get an email from an angry customer that’s saying, “I just bought your course, where’s the email that I paid for?”

Whatever it is. It’s always reactionary. For most other portions of people’s online businesses, it’s not reactionary. You go into it with a plan, ready to go. Email deliverability should be the same way. So what I try to advocate for is a more well rounded approach where ideally in the beginning you would establish best practices–or even in the middle if you’re already into a promotion–to make sure that you don’t have those trigger problems that will freak you out down the road.

The easiest way to prepare is just to ensure you’re sending relevant information to people who want to hear from you. Some people would argue that, well, if someone’s on my list, they want to hear from me.

But that’s not always the case. Someone could sign up to get a free report from you, and then they forget that you’re emailing them every week. Or they change emails. Or they don’t like you, so they drop you into a folder and forget about it.

The ESPs–Gmail, Yahoo, Hotmail–they see that happening. They think, I see Brendan or Derek sending out emails to our customers every week, and they’re not reading them. These emails must not be very valuable. Then, your name as a marketer gets associated with non valuable email. You gravitate towards the promotions tab or the spam folder, or you won’t be delivered at all.

The easiest way to combat that is ensuring that you’re removing people from your list if they haven’t been opening, reading, or clicking–because they’re probably not going to purchase.

You’ll see a huge pickup in the number of leads that are getting and responding to your leads.

You’re totally right. If you think about the history of email, when it was first invented, it was incredible. You could send an email from your computer to another country. Then it evolved into this tool that people used to market and to spam people, and the ESPs had to respond. It’s been quite the evolution–it was first used in one to one communication, and then people used it to market, and then people started to break the rules and harvest email addresses to deliver crap. So now, the ESPs have realized they need to pull the crap out.

By adhering to the rules, you’re making it really easy to get into the inbox. You just need to make sure that you’re segmenting your list and only sending to your best people.

Exactly. The reason for that is that in the beginning, spammers could ruin an IPs reputation, and then just get another IP. There was nothing stopping them. So ESPs have had to find new ways to identify mailers and then decide if they want to deliver their mail or not.

There’s a bit of both, honestly. With Google, you have the promotions folder, which is a bit more complicated, but if you’re starting out, it’s not as important to worry about the promotions folder versus the inbox. It can make a difference, but really, it’s a minor detail. If you’re just getting started, you want to be focused on laying a solid groundwork. You want to be focused on having systems in place to deal with those inactive people.

If you start with a company like Ontraport, we have package codes that you can enter into your account which will create our active re-engagement for you. What that does is it uses automated rules or sequences to send an inactive person on your list and say, we really like having you on our list but we also don’t want to annoy you, so click here if you’d like to stay on our list, or do nothing, and we’ll remove you in a week.

That shows ESPs that you care about your list and you care about segmenting. It’s great if you make it to the six month mark and you only care about the primary inbox–but the ESPs will still be seeing that people aren’t engaging. I really like to focus on that foundational piece first, rather than the smaller promotional piece.

It didn’t used to be super accurate, but now it’s getting much better. ESPs have realized that they can display images automatically, meaning that the pixel is downloaded every time someone opens an email. So it’s completely accurate. Most of the other big ISPs are following suit.

So they may be inaccurate by maybe 1% at most, and if you’re really a stickler for 100% accurate stats, you can always add in an opt-in that lets people display the images on their own. But it’s much more accurate now than it ever has been.

I was just doing that this morning! We just launched a program called Ontrapages, which is a landing page generator. We launched it at our annual conference, Ontrapalooza, which was an awesome time. What I’ve noticed is that our opt-ins for Ontrapages and our sign-ins, it’s about 50% Gmail, whereas even a few years ago, our list was only 27% Gmail.

Gmail has really taken over in a big way. If they continue to grow the way they are, I think we’ll all wind up with a Gmail account.

Yes! That’s not even counting Google Apps.

Yes. So I’ve done a lot of research on this and I’ve found that it’s actually simpler than I thought it would be. One of the things they look for is the number of links. So for example, we were setting up a welcome campaign for Ontrapages. Before we went live, I did our email deliverability test. We had four links in it, and it kept hitting my promotions folder, no matter which words I changed or took out. It made no difference. That’s just where we were going to live.

The next thing I tested was taking the number of links out. I took one link out, and it hit promotions. I took one more out, and voila. We hit the primary tab. It appears to me that one of the main factors is the number of links, and your image to text ratio. You want to keep that around 80% text, 20% images. So it’ll be a little different for everybody, but as a blanket rule, if you keep it to two links, and 80% text with 20% images, and you’re not purposefully misspelling any words, you’ll get in there just fine.

I did not include that because I figured everyone would have one, so you’re technically allowed to have three links:your unsubscribe, and two others in the body of the email.

It can. The thing that many people who send mass mailings don’t realize is that reputation is iterative. There’s the worst of the worst, spam score ten. Then there’s the best of the best, spam score zero. Let’s say you have a spam score of six. By following best practices, you might get down to five, and they won’t suspect you of being a spammer.

My suggestion is to always follow best practices. None of this will be a problem. For those of you that are already stuck with a bad spam score–now is the time to start fixing that. If you do one single thing after this call, go to your list, segment out your inactive subscribers, and get rid of them. It’ll take a few weeks to see the impact, but I promise you will see a change in how people interact with your emails.

I know it’s hard to get rid of part of your list, but if they haven’t bought or clicked or opened in six months, they’re probably never going to. You’re not really losing that much, and plus, you’ll save money on sending fewer emails.

Yes. We have an integration with SpamAssassin. It’s not the be-all end-all, but each of the ESPs have their own private filters. Naturally, they won’t tell you what those filters are, because everyone would get into the inbox. So they can’t give out what they’re filtering for.

So we use SpamAssassin. It’s the largest open source spam filter available. People that are hosting on GoDaddy or BlueHost or they have their own server, they can install it for free on their email servers and it’ll scan your email. It’s a great place to start, but it won’t be 100% accurate with Gmail or Yahoo.

Yeah. If you google SpamAssassin Filtering List, it’ll return a bunch of the things that it looks for. Things like the HTML version not matching the plain text, or raw URLs that go to different places. We also have lots of reputation trackers in place for things like bit.ly, so that will be flagged, because so many spammers use URL shorteners like bit.ly. If you use it in your email, it makes you look like a spammer. It’s all mostly content stuff.

I do most of my testing on Gmail, simply because it’s so prevalent.

Yep. Once is nice but twice is better.

They used to be one of the most important factors, four or five years ago. But more recently, I have seen that subject lines aren’t really having a huge impact on your inbox rate. If there’s lots of symbols in the subject line, that’s caused some problems.

A lot of people like to write RE: or FW: in the subject line to make it seem like it’s a real email, and it might be a reply or a forward, but that’s been shown to make a difference in spam folders. If it looks like you’re trying to impersonate another email, you will go to the spam folder. So stay away from reply, forward, and symbols in your subject lines and you should be fine.

Outside of that, just split test email deliverability. I know there’s lots of gurus out there that can be really specific and say that if you include a certain number of tips to make a certain amount of money per month, you will get the best open rates. Just remember that everyone’s list is different. Listen to those gurus, but always verify with your list because what works for them may not work for you.

Honestly, I haven’t tested it too much so I can’t give you a concrete answer either way. If your HTML is too long or too bulky, emails will start to get clipped, but I haven’t done any extensive testing on a code, because it hasn’t yet been a problem.

There’s definitely a small impact on deliverability with regards to your domain. But things are starting to move away from domains more and more now, because spammers are able to hop on an IP, blow an IP out, and then get another and mail from there–that’s also doable with domains.

What ESPs are starting to do is to try to thumbprint mailers. They do this by way of list makeup. Let’s say you’re mailing, Derek, and it goes out to one million people. When you mail, it’ll always go out to the same 400,000 Gmail inboxes, with one or two difference. Your list is very unique, because it’s unlikely that anyone else has that specific 400,000 emails. So the ESPs are able to identify your segment that way, so if those same 400,000 people get an email, they might think, maybe his list was stolen. Or maybe he’s sharing. Or maybe, he’s hopping around from domain to domain trying to spam people.

They’re developing ways to target, judge, and assign reputation based on your list, more than your IP or domain. There’s some people that are learning to break up their list and be strategic about who they send to when to avoid being thumb printed by the ESPs. So, yes. It does play into deliverability. I think it has more to do with recognition for your customer.

In my opinion, your transactional addresses can go out over a different IP, so that your people will get the important stuff no matter what. Even if your reputation isn’t really strong, you know that you can get your most pristine communications out over this IP.

As far as whether it should be involved with other mail, people will argue that if that one IP is so pristine, we should let it reflect on our other IP to even things out. To help bring up the reputation of the others. With the way things are headed, I don’t think it’ll make much of a difference, but for the time being, use a different IP, use a different from name, just so you can establish a really healthy reputation for transactional communications. But eventually, your reputation won’t depend on your IP or from name, especially if we get to IPV6, IP reputation will go straight out the window.

It’ll have a huge effect, because we’re going to go from the amount of IPs currently available based on the limits of the numerical structure we have, to basically innumerable IPs. It’ll be way easier to hop around, easier to get, cheaper–there’ll be so many more options. IP reputation will be irrelevant.

Not really. I do know that lots of people have been saying that we’re going to run out of IPs and it’ll be a catastrophe, but that’s been happening for years. So there’s speculation, but I think we’ll be good for the next several years.

For the time being, yes, because if it’s only transactional email, it’ll only be going out with invoice or product information, so all the mail it’s lumped in with is only of the most importance. Right now, it’s perfect. In a few years, it will not make a difference, because we’re moving away from that type of tracking.

When you connect to Sendgrid for example, they will filter and screen your emails and separate them into which will go on the transactional IPs and which will go on their normal IPs. Don’t take this the wrong way, Sendgrid and Mandrill are great services, but when you’re using their normal IPs, it’ll be mostly the same as using your ESPs IPs. There’s only so much you can do on the technical end. it won’t make that much of a difference, especially because most ESPs out there also have reserved transactional IPs. So if you were an Ontraport customer, we do send all double opt-in mails and transactional mails in a separate IP pool.

We recently released Ontraport basic, which is a lower priced option for people that want a powerful tool like Ontraport. The goal was to release something that would allow people to start small and allow us to scale up with them as their business grow. That wasn’t previously available. Most ESPs like Ontraport are really expensive. We wanted to change that. So go check that out at Ontraport.com. If you’re into landing pages, check out Ontrapages.com. We have a free version if you just want to dip your toes in, and we also have a premium version with more templates that can connect to any ESP out there via FormConnector.

Of course. Thanks for having me, Derek!