Audience Bridge [Insights]

The ESP That Deleted Its Own Honest Open Rates w/ Travis Hazlewood

• Chris Miquel • Season 1 • Episode 19

Share episode

Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.

0:00 | 50:28

Every email platform can identify machine-generated opens. Almost none of them show you. 

Travis Hazlewood is Head of Email Deliverability at Ortto, now a Canva company. He's in his eighth year in deliverability, running global platform delivery audits, blocklist remediation, and sender reputation repair for high-volume senders. He sees the data your dashboard doesn't show you. 

He told me a story I hadn't heard before: Ortto built filtering to strip machine opens out of reporting, shipped it, and had to pull it when customers complained their numbers looked worse than on competing platforms. 

Honest reporting lost the bake-off. We get into why Microsoft is deliberately camouflaging its scanner clicks, what Proofpoint admits it does to previously delivered mail, why Travis calls send time optimization a crutch, and the term he coined for the damage marketers cause without meaning to: Secondhand Spam.

We also disagreed on air about whether the machine-open metric belongs in a dashboard at all. He'd delete it. I gate on it. Both positions are in the episode. 

Plus: APRF, the proposed standard that could finally tell you where your mail actually landed. 

Subscribe for weekly conversations on newsletter growth, deliverability, and monetization: https://www.youtube.com/@audiencebridge 

Get the newsletter breakdown of every episode: https://www.audiencebridge.io/subscribe 

CONNECT WITH TRAVIS: 

CONNECT WITH CHRIS: 

Who is Travis Hazlewood?

SPEAKER_00

Everyone is trying to do send time optimization based off of engagement.

SPEAKER_01

Users aren't engaging with your emails. You're not getting enough real opens, but I think it could also be opens. They know real opens and you're just not getting enough engagement. You're compliant, but you're just not engaging enough.

SPEAKER_00

I just try to work with marketers to be more confident that their stuff is actually worth.

SPEAKER_01

Schema, setup that you can do, that we're supposed to read, so we know exactly what you want, but we're just gonna throw that out the window. It's like it makes no sense.

SPEAKER_00

Every time we talk about gating these things, we're we're risking organic activity being lost because we just don't know what should be considered a real open versus a machine open. Microsoft, who is one of the biggest sort of offenders, they have found better ways to hide it because these are security systems. Links are the riskiest piece of content. It's going great. Thanks for asking me to be here. We've had so much fun conversations on LinkedIn, so it's it's cool to really get to just chat things out now. Yeah, there's lots of banter going on in LinkedIn.

SPEAKER_01

I love it. Yeah, yeah. Um so for those of you, I mean, Travis has great insights as far as on the ESP side, the platform side, he gets to see all the stuff, all the data coming in that's kind of a lot of time masked from what you see. You'll see like open rate and click rate and all this stuff. And you know, a lot of ESPs kind of hide the real numbers from you, and some of the data that helps them fix the liberability issues knows what's happening to then help you. But some people like to have more of that uh information themselves so they can do stuff, especially when you grow bigger and bigger

Opener bloat and the subscribers who never click

SPEAKER_01

and bigger. But we have them here, so we're gonna discuss some some nerdy data stuff and and things about opens and clicks that that'll be eye-opening for you guys. So that being said, I I've been talking about this opener bloat thing lately a lot to people telling them, you know, a lot of people I have a base sending segment where, you know, you're just a normal sending segment where you have your you know your clicker window, your opener window, and then your newly subscribed window, uh, just to, you know, optimize your your list automatically. Yeah. Um, but what a lot of people don't realize is that opener segment, there could be a lot of what I call bloat, which are just basically your habitual openers that have never clicked in their life, and they probably will never click because they're probably most likely bots or security scanners or whatever it may be. And it's kind of bloating what you're sending to, so it could increase your sending costs. And on top of that, I think that it can also affect uh your deliverability on the ISP side because they know what a real open is, they know what a real click is. So if you keep on sending these people, they know they're not open. So they're not giving you credit for those, and you think you are, which actually I've seen lately. Uh, you'll see, uh, I saw the other day on uh Postmaster Tools, you know, fully compliant, everything good, but on the bottom it's like users aren't engaging with your emails. You may want to do whatever, which is like either you're not getting enough real opens, or you're not, I mean, I'm I'm a click guy, so like if you're not getting enough clicks and stuff, enough engagement that that's there. But I think it could also be opens, they know real opens, and you're just not getting enough engagement. You're compliant, but you're just not engaging enough, which I thought was very gendered. Have you seen that yet?

SPEAKER_00

Yes, I've I've seen reports in the industry of that. It's pretty cool. I'm very excited. There's a couple things. There's that, and then there is a sort of beta going on, um, sort of similar to DMARC.

APRF: a new standard for real inbox placement data

SPEAKER_00

Um, so you if for those aren't familiar, DMARC is what you is a fairly newer sort of thing you can set up on your domain that provides uh reporting capabilities and also a little bit of a gating for uh it's more gating guidance for mailbox providers to know what to do with traffic that comes from your domain but isn't actually authenticated with your domain. Um, DMARC was obviously created with a very specific purpose in mind, and so it's reporting and stuff is around that. The like mailbox providers literally can know who to report via that DMARC thing. There's a new sort of reporting tool that's that's being beta right now. I I believe I saw that uh some people from Comcast and um Google even are beginning to sort of troubleshoot and see if this looks like something they can support, where it will provide a report instead of saying this is the traffic that is unauthenticated from your using your domain, which is what DMARC does. This one would actually say you got X percentage inboxing rate. It would be like the first time you get a genuine inboxing rate directly from mailbox providers. Um, and so that's in the middle of being sort of figured out if it if it's functional enough, if it's safe enough for them to do. So, like, yeah, there's some really cool stuff kind of come down the pipelines. Um I'm glad that Google Postmaster is already elevating something in their in their app because they can do that, they have the direct visibility of real engagements that you know everyone externally doesn't. But um, but even more so, outside of that tool, it would be nice to have a sort of industry-wide sort of standard that could be implemented by everyone, and then you would just set up like one little DNS record and you could get all that data in, and you it wouldn't become as much of a black box as it used to be, which is really cool. I'm very hopeful that it works out. I mean, a lot of people are hoping Gmail's actually better from it.

SPEAKER_01

It'd be a it'd be a major game changer. I mean, they they've done a lot, you know, uh especially when they're like, uh, we're getting rid of uh uh our version one, you're not gonna have to go main rep, IP rep, none of that stuff. Yeah. And it was like, okay, obviously they haven't killed it yet because uh there's too much pushback. Like, I mean, you can't blindfold everybody, but if you get that inbox stuff, I I mean, I think I find it hard hard to think that Gmail's gonna totally go through. I hope they do. I just it seems like uh I don't know. I'm uh maybe they'll just do it to certain for certain people.

SPEAKER_00

Um

The spam folder is a security tool, not a marketing one

SPEAKER_00

I do think it'll probably be like so a lot of people don't think about this, and this is really good context for um marketers across the board. Um, but you know, the spam folder is a security tool. Like that's the primary, it was built to help Gmail deal with the crazy well, not just Gmail, all of them, and Yahoo really is kind of the first spearhead sort of the the sort of I don't know if that's the real story, but that's the the story that goes around the industry that Yahoo kind of spearheaded the spam folder. But it was originally designed to help get a little more sophisticated and help create another gateway of security because that right blocking causes really problems, obviously. Um but because of that, a lot of people just see the spam folder as a nuisance from a marketing perspective. But from the mailbox provider perspective, you know, they are by by giving more information around what land what causes you to land in the spam folder versus the inbox, they're actually kind of ruining their security tool because the bad actors, who it's really mostly for, the more they can understand it, the more they can trick systems. And so they don't want to give too much info. And so I do believe if this were to come out, more than likely it would be right relegated to uh quality senders, kind of the way that uh Google works with like BIMI and uh with their their blue check mark and stuff like that when they were doing that. You know, it it you had to meet a certain standard of quality to even get that, and I think that'll probably happen here

Nobody agrees what an open or a click even is

SPEAKER_00

because they're not gonna do that for everyone. That's a lot of data for them to parse.

SPEAKER_01

Um what I think would is is what people really want to know here is uh you know, the difference everyone sees their open rates, um, obviously the click rates, and you know, every platform is different. And when you're talking about like a standard, like there'd be great, it'd be great if there's a standard that like this is what an open is, whatever, whatever the window of the open is, you know, like this is a click, like five second click doesn't count. Any other stuff you can identify, yeah. Um, through whatever ASINs or or stuff you're looking at, or or historical data and and all the stuff you're gonna do.

SPEAKER_00

You can't even uh agree on what a click rate should be named. Should it be click through? Should it be just called click rate? Is it called unique click rate?

SPEAKER_01

Yeah, like well that was like that was like uh uh for Beehive, for example, uh for a long time, like their click rate that they were showing everyone in the front was really a click to open rate. Oh my gosh. So you would like, you know, everyone's like, ah Mike, what's your click rate? They're like, ah, it's like uh, you know, two point something percent. And I'm like, well no, it's that's like 0.75 or something because your open rate's like a 35% or whatever. So like it's not it's not truly, you know, these people care about all the email you're sending, not just what's opening. So, but that's another thing. They're never gonna agree, probably, to come up with a standard. Uh too many different people. That being said,

Ortto built honest open rates, then deleted them

SPEAKER_01

I know you've worked a lot with identifying and helping your your the companies you're with identify what's considered a human open or or or a machine open, same with clicks. Yeah. Um, can you kind of dig in like what's behind? Obviously, everyone knows MPP, right? That's like the easy one to identify, like MPP opens. Everyone knows how to do that. They can get rid of it if they want to. Some platforms get rid of it. Beehive supposedly gets rid of it, for instance. Like at default, some don't. Some give you the option to remove it. Uh, but if you can just kind of go through like what you guys do or what your thought is on like what should be considered a a real open versus a machine open.

SPEAKER_00

Yeah, no, that's you know, when I when I first came on at Ordo, and that was one of the early conversations was whether we should uh filter out opens that were that we considered machine opens. Um and you know, we did utilize um some of the more technical aspects that uh are known industry wide. And so, you know, like even Apple sort of gave some of those details around the MPP, for instance. And so um we we we we played with it because you know other platforms have done that as well. But you know, immediately hit everyone started screaming that their open rates were worse at Ordo, and it's like, well, we'll we're filtering out, and it's like okay, so that looks bad. Um, so we pulled that. But I also agreed on another level with with pulling that. So, you know, I've worked with the teams to identify opens and clicks that are machine related. I built out the dashboards and templates based off of the um, you know, field that we use to identify which or which, you know, we utilize technical elements around like what data center, you know, data centers that have been identified as real, you know, relevant to say like a Microsoft uh machine click or whatever. Like some basically we we look at patterns and trends that are like, okay, well, you know, a thousand people clicking from the same source is not a real person clicking, right? The problem is that we've even gone even further with some other testings to find like to really push the boundaries because we have some B2B senders on our platform. We have a lot being a CDP, you know, and the problem is that even with more sophisticated changes, it looks like you know, Microsoft, who is one of the biggest sort of

Microsoft is disguising its scanner clicks on purpose

SPEAKER_00

offenders, Microsoft 365, especially, in doing these sort of click uh machine clicks, they have found better ways to hide it because once again, these are security systems. Um, you know, the the open, the MPP, Apple MPP is sort of a security, it's more of a privacy, but the the click oftentimes is related to some sort of uh security check because links are the riskiest piece of content, unless there's an attachment. But the problem is that they are intentionally making them more organic looking because they need the bad actors to not be able to figure out how to get around it. And we even tried some tests where it forced like maybe a secondary click, uh almost like a CAPTCHA page that it would go to like a sort of a landing. And I could tell it still didn't solve the problem. Market Microsoft 365 has the ability to go ahead and and preload, you know, Java essentially and fetch the next page and blah, blah, blah. And so, you know, these things are getting harder and harder, almost impossible to identify, even from an ESP standpoint. And so at the end of the day, you know, me and my team still work with the same data that marketers work with to still identify things. Um, you know, open rates ideally, you even with them inflated, is is still usable as far as deliverability trends to discover if you're in the inbox or not. Because, like with Apple MPP, which is the biggest one that's inflating opens, it only triggers if you're landing in the inbox with that email. If you land in the spam folder, it won't trigger unless they go and they open it and move it to inbox. Because if sometimes uh you know, different platforms can do different things with that in the spam folder, but usually they don't uh you know, load images in the spam folder. Um, Apple might, but I can't remember. With that said, you know, the the inflation just still helps because it's saying, okay, you're getting enough engagements to get into the inbox. So we can at least use that signal to understand the trend of whether you're going into the inbox or not. It really is becoming more important to find further engagements, uh,

The psychological wall between an open and a click

SPEAKER_00

like we've talked about with clicks and conversions and um even replies. I I speak to people because we have reply tracking that is way harder to obfuscate now, you know, or or you know, inflate. You know, obviously you can get some uh auto responses, but for the most part, you're not gonna get that unless you're sending more B2B. And so it is becoming more and more necessary in the industry to to move beyond opens to actually tail per subscriber engagement level. But I also know that that's a there's just no way to truly replace opens versus clicks because you know, in this uh easily, because one of the things we just don't think about is, and I was talking with another friend of mine who who does an email-related newsletter, how there is just such a hard uh psychological wall you have to get over beyond the open. Um, organically speaking, an open is more common and easy for the subscriber to do. Like they're that's what they're expecting to do. They're going into the email just to to browse. And so they don't mind. They don't mind just a casual open. They know they can scroll and they can get right back out. Any action outside of that, though, is a steep drop-off. Like, just think about, for instance, that the the uh sort of benchmark open rates have now become essentially 30, you know, uh 35%, ideally, at as your minimal target. Click rates are still around like two and a half percent, three and like having I know tons of marketing programs that have strong conversion and they still have a one percent click rate or lower. And so, you know, click rates are just so hard. And when you think about that, you know, getting someone to go to move into an open, it's like you have like in that sense, a three out of ten chance that you'll get them to at least open that email. But then it goes down like crazy to like two and a half out of a hundred chance to get them to click. Like that is a that just I think we don't think about that enough to really understand that that's kind of why clicks can't fully replace what an open is, because it's just such a mental hurdle. And so, what it requires is a content strategy that is probably different than what is common in marketing today, that it has to do two things. It has to not, it has to force further engagement, it has it has to maybe three things. It has to force natural um further engagement, it has to make that sort of enforcement feel natural to the subscriber, like almost unconscious, like, oh, I need to do this, or you know, kind of reactionary. And third, it needs to not do it in a way that ticks people off, you know, right? Because like the whole like clickbait thing that ticks people off. So it's kind of like that's the

Timing windows and where they break down

SPEAKER_00

future, I think, of email content is if we're going to have to rely on these further engagement signals, you're gonna have to earn it. And the and the content practices now that are normal, they don't do it. So uh a couple of things to unpack.

SPEAKER_01

Uh, first on the the opens themselves. So I've worked with uh a few people with their own custom uh email platforms that they use to manage your business, yeah. And and helping them identify what a what a real open is and what a real real click is, and I and I really focus more on the time window when things happen, kind of like your five-second rule um to identify those. Because I've I've seen the past where they've really hurt themselves, say doing open triggers, like an email opens, and they just send stuff out, but a lot of time, but those aren't real opens, and they just keep on sending tons of email to to people that haven't really opened, and yeah, that can actually purchase delirial if you you know end up sending 10 emails and knowing that they didn't even get one real open in like a two-day period, right? Oh yeah. So uh doing that on on your platform or in general, like what's your thought on like the trigger that's aspect of like automation?

SPEAKER_00

Yeah, that's definitely hard. Um, like for instance, we do um I we don't we we do filter out, for instance, clicks that happen within I think three seconds-ish of delivery, three to five seconds. It's really hard too, because we're also in the space where everyone is trying to do send time optimization based off of engagements. So they're trying to send when people are there. So, you know, um, but I do think it is uh uh you know unrealistic to assume that they are sitting in their inbox waiting and that you are the most important thing there and that they're clicking within the first 10 seconds of delivery or even opening. Yeah, and so a lot of times when I'm troubleshooting things to see, obviously, like I said, Apple MPP works in such a way that it it will not load another open. It will not trigger another open later as long as they're using a device that that was a part of the prefetch that had that that loaded. Now, if they go to a different device and they open it, or if they wait like a month and they open it or something, I can't remember what the timeline is for when they stop caching an image. Um, you know, it will reload then. But essentially, you don't have for the Apple MPP opens, you don't have an option to say, oh, well, let's only trigger on opens that happen after two days or or or or two hours, I mean, or an hour or something. You know, it's like you're just not gonna get that data source. You know, clicks you can you can do that to a degree, but I also have seen Microsoft 365, for instance, they do patterns that when I zoom out and look at different examples, they look organic, but it only if you zoom out and then start looking at different people do you start to see, oh no, that that had to be a machine, you know,

Proofpoint's retroactive server clicks

SPEAKER_00

and it wasn't in the first 10 seconds, it was an hour later, and it was uh a day later sometimes. And even proof point, uh, which is more of B2B, so probably most of your audience doesn't know, but there is a mailbox provider that is fairly bigger in the at least B2B space that has openly admitted that they will trigger uh server clicks on previous emails if you suddenly reach a certain complaint threshold with a more recent email. I'm like, and and they understand that all of us are like, why would you like that's so dumb? And they were just like, Well, that's how our system works. They don't care, and that's the thing, all these different services that do not care that they're obfuscating our understanding of what our uh uh audiences are engaging with, at least for now. But if they see more trends of crappier and crapper sending practices because of how they're they're gating everything, like probably this thing with Google and Comcast all baiting is probably a part of that. Um, because they're gonna deal with the consequences of that. But yeah, so you know, this is a fine line because I don't do a lot of work with people trying to get them to figure out is this a real because I don't even trust it. Like I have I have seen too many crazy patterns that are just like they're just not able to be truly captured and understood because they are trying to emulate human activity. And so, like I I just lean more on, you know, instead of trying to like, is this an open, is this a click or whatever? You know, you could build something where you have multiple opens or multiple clicks uh that are not right next to each other, you know, because sometimes those server clicks do happen like 10 clicks across the 10 links on your email in uh three seconds. Well, obviously that's not a real click, but you know, you could have all of you could have a server click happen and then literally have the human click within a few minutes or even seconds if someone's in their inbox. So, you know, ideally maybe you you could say we will at least wait for you know the second click to to be a real click, you know. But every time we talk about gating these things, we're we're risking organic activity being lost because we just don't know. And I'm not a fan of that. And so I just try to work with marketers to be more confident that their stuff is actually worth a real click and worth a real open. And um, and at the end of the day, I communicate that let's look at trends and let's look at improving trends. So let's say that 50% to 75% of your clicks are fake. Okay, well, you've got a 1% click rate, let's get you to two. Like, whether they're real or fake, like test it out. You're still gonna see those real numbers coming in later.

Send time optimization is a crutch

SPEAKER_00

Just do different things. Improvement is the cure, regardless. Just improve the training, improve the opens, improve the clicks.

SPEAKER_01

Um, you brought up uh send time optimization. Yeah, I get asked about that a lot, you know, on the ESP level where they have the setting like, oh, send at the optimal time for the person or whatever. I've always told people that I've I haven't seen any real noticeable lift by doing that. I want to get your thoughts on that. I mean, my my my position always like it's hard to use that tool when you're setting the same time every day. Right. If I'm sending a 8 a.m. every morning. Bulk of your stuff's gonna happen that period. Even though if the the person usually only at 4 o'clock p.m., maybe another part not even seeing it, they only open if they see it here. So like it just obfuscates like the true send time optimization you would think of. You have to like send like eight times a day and see, like, oh, where's this person opening? And then okay, I'll just send that time.

SPEAKER_00

Yeah. No, that's I think I did a post a month or two ago where I I I believe I called some time optimization. Actually, it might have been in my my LinkedIn newsletter, but I called it like sort of like a um, oh my god, I forgot the real world word for crutch. I I called it essentially like a crutch for sort of bad marketing because sometime optimization is using the unreliable engagements as accuracy. Like I said, the better thing you can do is trend, use it as an overall trend and try to build better trending and stuff. When you're trying to do accurate per second, per minute, whatever, you know, with those engagements, you don't know. Um, and those aren't real. And so you're right, because the Apple MPP is gonna just open that email whenever you send it. So yeah, it does look like it. And so, yeah, we have that in our in our platform. Um, I'm not a big fan of it. I I like basically never encourage the use of it unless you're trying to sort of time zone thing, you know, because obviously if you know the user's time zone and stuff, that yeah, I could see that.

SPEAKER_01

Um, but that's still that's still a you know, uh just try to hit the five. Right? Right.

SPEAKER_00

The three-hour window that you're yeah, you're not trying to hit yeah, and the uh the idea of being top of inbox is the thing that I fight against. Okay, because number one, you can't you can't do it, period. Number two, the new um tools that these major mailbox providers are creating are ruining the even concept of a chronological uh inbox. You've got Google creating the sort of the Gemini uh backed uh AI inbox where they're gonna be summarizing and trying to identify your core things you're important that are important to you, so that if people start leaning on that, then they're not even looking in their inbox uh at the list. And so there's gonna be that. Uh, Apple Mail already has a feature where you can actually go by the send the sender themselves. You can go in and you can look at uh a sender focused view and it'll show all of their emails, but it doesn't even show it chronological all the time. Sometimes it shows it by like one you already engaged with, and then it does it in like a weird backwards order. It's very confusing. And so they're they're doing all these things that really show that they want to change our uh subscribers' relationships with their with their mailboxes because the bulk marketing practices that have become spammy, they'd be fine if only one person was doing it, but the entire world's doing it. They're forcing these mailbox providers to have to figure out a way to elevate important things to their users, or they'll lose those users because someone else will do it. Um, because we're just saturating with sort of our crap when we're doing just generalized stuff. And so, you know, that I did a whole write-up where my perspective is that sentime optimization is a crutch that doesn't really do what you want it to do. And if you want the effect you're looking for, you need to create a relationship with your subscribers with your content where they go looking for you. Like you create consistency to when they know you're usually in the inbox. Like that's just by day and by a certain point in time in the day. And then you also create um, you know, an email experience that is not annoying, that is exciting, that creates expectation and gets them to open your email and and look forward to it when they're perusing all the different names that they're seeing in their inbox. And so that's a lot of trust, that's a lot of well strong marketing practices. That's really what send time optimization people think is doing, but it's not. And if you can do that, you can send whenever. As long as you do it consistently, they'll

AI summaries are replacing your subject line

SPEAKER_00

go find you. And that is strong marketing. That was well said.

SPEAKER_01

Um, the uh I I saw another uh one of your uh articles or or posts, I can't remember which one it was, but on you know, AI and you were talking about summaries in the inbox, how that's how that can affect the opens. If you want to kind of get into that.

SPEAKER_00

Yeah, no, that's a great thing to call out. Um, so the different the the crazy thing too is that the different mailbox providers are doing different things. You know, um, I believe it was like Apple was repla uh was playing with replacing the pre-header text uh with an AI summary. The Yahoo, though, was going so far as replacing the entire subject line and pre-header text with the AI summary, which is very scary. That's that's yeah, that's that's out of control. And then, you know, most of them are are doing the sort of AI summary card at the top, um, where yeah, they they try to summarize it at the very top. Sometimes that's helpful, especially like delivery notifications or whatever. But obviously there's a lot of inaccuracy with those um too. We've seen with them misreading the discount and the promotion and saying like get one 100% off and it was a BOGO deal, you know. And so it's like it didn't fully understand the BOGO. Um, so yeah. So there's obviously mistrust that that can cause. There's a lack of engagement due to a misrepresentation. But the the bigger thing from a deliverability perspective is that while we can't know for sure what an open is happening, Google and the rest can. And when their AI summaries are stopping people from doing an open, then they are going to consider that bad engagement and start punishing. And whether uh, you know, at least it's for now because they're still bidding, it's very likely to happen. Hopefully, they will build algorithms that understand sort of visibility enough when people are at least reading, which you hopefully I don't know how they would do it, you know, um, without uh, you know, kind of compromising privacy concern, you know, creating that. But yeah, it's gonna be really hard for them to have that signal if their summaries are are replacing opens by giving everything. And that's another thing that leans towards you just need to do better marketing that gets people excited to look for more. And it also really pushes, in my perspective, the idea that your your emails need to be focused. There is very obvious uh, you know, evidence and reporting around, you know, having that first sort of uh main portion above the header, as they call it, have a single point of focus. That is very important to do. But that the rest of your content, there should be more. There should be extras, there should be Easter eggs, there should be exciting content. Um, that when the summary summarizes the main part, people still open to find out those other things that they know that you surprise them with uh each time.

SPEAKER_01

And so, yeah, that's gonna be more important than ever. And that what's crazy is like you could even include schema for like your offers and stuff, and they could still do it. And they uh read that just overlook like we have this this exact you know, schema setup that you can do that we're supposed to read, so we know exactly what you want, but we're just gonna throw that out the window. It's like it makes no sense, yeah. Let's talk about your your secondhand spam uh that you like to talk about. Um kind of obviously everyone knows what a spam is, what a complaint

Secondhand Spam: when your signup form becomes a weapon

SPEAKER_01

is. Um, obviously it's the the most negative signal you could ever get. Um, and you you don't want those, that's what really tank you to Little Bodie Fast. Why don't you get into like what is secondhand spam? How can that kind of infiltrate uh your list without you kind of even knowing? Um so they they kind of know what that is.

SPEAKER_00

Yeah, I I feel kind of proud of myself for coming out with that term, a little alliteration there. Um, but it it is something that is a lot easier for um people like myself on the ESP side to to understand than maybe uh marketers would see on average, because it doesn't happen like you know, like the normal bad marketing practices. This is we're talking about things that are sort of incidental, things that you wouldn't expect that I see in the industry because I'm seeing thousands and thousands of marketers. Um, but you're not gonna see, probably, because it's not even 10% of marketers at any given time are experiencing this or something like that. So, what I'm talking about here are situations where some sort of uh bad practice or ignorance in your marketing stack or setup can cause incidental spamming. And so that would be things the first top of mind one is uh what we call spam bots that that attack signup forms that are unprotected. So uh we all know about CAPTCHA or re-CAPTCA's Google's uh version. Um, you know, we've all have the jokes and the annoyances with that, but they they actually do serve a serious need in the industry because bad actors are creating these these bots that search the internet and find signup forms. And any that don't have those CAPTCHA will be abused. And sometimes their intentional abuse uh where they are trying to hurt a particular sender or brand, um, oftentimes they're just they they can also be used specifically by bad actors that want to fill up uh certain recipients' mailboxes uh to hide their scamming and their malicious activity. And so, like someone's, for instance, if you know they they steal a bunch of people's credit card information, get their email addresses, and then do a spam bot to sign them up to a million different newsletters, and then they start getting all the welcome emails, they're not gonna see the email from their bank saying, is this a f a real charge? Is this a real charge? It's gonna be hard for them to even find anything in their inbox anymore. Um, so there's a lot of reasons, sometimes intentional for them for the marketers brand, so oftentimes not, where just not having catcha is being abused

Name-field injection and hijacked welcome emails

SPEAKER_00

um and your system is being. I've even seen where they get so smarmy, they figure out that you're using a first name even in the welcome email, and they'll take people's real email addresses that they steal online, they'll put them in the form, and then they'll put some sort of scan link, or even if, and this is another thing, you didn't even have like a text limit on the name field. Like you should have, you wouldn't think you'd need to put like a 30, you know, um character limit. 30 character limit, but I've seen where they have it, and I've seen where it's where they put a whole email worth, a full paragraph of content, and they are essentially hacking this person's this brand's welcome email, and so that instead of it saying hi, first name, and then giving the welcome pitch, it goes hi, does a full scam pitch, and then all of the real welcome email information is lower. So this captcha thing is really important because you have no clue how it can be abused in that way. That's the the bigger one that I see most often. Other than that, I do also see things like um, you know, obviously you've got the the unhealthy or or sort of forgetful list practices where you accidentally um sent to the wrong segment that was a bunch of people you were trying to exclude, but you didn't put exclusion properly in the segment name, and then you accidentally sent to it. I would call that secondhand spam because you're just unintentionally sending to these people and they're gonna work for spam.

SPEAKER_01

The time that you, everyone who's been doing email long enough has sent to the entire list on accident. Yeah. Oh good out to two million people. Ouch.

SPEAKER_00

Yeah, that's definitely uh uh what I would call it's like a secondhand spam. Like you didn't mean to, it's just kind of like happened because of the environment. Uh, some some you know, yeah, problem there. Uh another thing would be um uh what a lot of marketers struggle with is this idea of content-specific consent. Um, everyone thinks that they can just send to whoever gives an email address and it's fine. And maybe legally, yes, you can do that in places like the US and stuff that you all you have to do is get one consent and you can send whatever you want to send. Uh the problem is that that's not how the mailbox providers' algorithms are really responding, they're responding to engagement. And so if you're sending stuff that is undesired, it is not gonna get the engagement and it's gonna be considered spam. And so you've got marketers, uh, whether they're doing it with their own content or they're just opting people into this newsletter and that newsletter and this product nurture and that product, you know, and not realizing that someone just wanted this one little thing, and so then they're gonna like stop engaging and then they're gonna global unsubscribe because they're just sick of it, um, to the point, or to the other side where you've got these big conglomerates that own multiple brands, and then they try to share lists. And it's like, I don't know that you know, like Clorox is related to Fabriz, that they're owned by the same brand, and then you're but you're sending from the Clorox uh domain, and then suddenly I'm getting an email from Fabriz. Well, I didn't sign up, it's spam, it's unsolicited. And it's like, well, they have the legal right, but the there is a psychological factor on the subscriber side that there is a content-specific consent that needs to be work with because subscribers are the ones that are ultimately deciding what spam is, not the legal system.

Reply tracking, and whether it scales

SPEAKER_01

Um so yeah. You um you've mentioned that replies are the the most positive signal you can have. Right. You did mention too, like the that you can track replies. That's kind of interesting. How how can you guys track reply?

SPEAKER_00

Yeah, I mean, sort of like it's not too sophisticated. You know, you just do sort of like an uh a sort of alias rerou uh forwarding system kind of thing. Um, and then we just log it. We just log a few uh field factors, say like, you know, you you will have like maybe the subject line that who you know who replied so it can, you know, obviously ping back to the actual subscriber address in the data platform.

SPEAKER_01

If you're just updating it, you're updating the reply to to this, that's the right.

SPEAKER_00

Yeah, we just basically wrap. It's kind of like click tracking, we just re-wrap it with our own that we forward. Simple. Yeah, because I mean that's the best stuff that works in email because it is an older system, so you you don't want to be too sophisticated.

SPEAKER_01

So replies are obviously definitely the number one signal, but when you are a large sender, uh B2C, and you're sending like newsletters out, like you know, if you're if if you like think of the 1440, right? They're sending the three and a half million, four million subscribers on their list active, you're not probably gonna get enough replies to really move the needle. Yeah. So where do you see that and that in that size game? Or is it more for B2B style stuff? Like, where do you see like the maybe there's too much volume that it really probably is not gonna move the needle, or or where and another thing is what's your thought on all these like inbox warm-up tools, right? All these these seeds and they're replying and they're engaging to kind of warm up your your your your domain or your IP when you're moving over or migrating somewhere. Yeah. How does that kind of go? Like I've I've seen some people use a warm-up tool. I I won't I won't name the the company. Um, and it and it helped initially, right? Initially got good reputation, was allowed, you know, warmed it up, was fine. And then they kept it in what do you want to call maintenance mode or something? And then over time, it kind of started taking the lower ability. I don't know, it's because it's the same emos they see doing the same thing. Like, all right, these guys have been too too engaged for way too long. This is just garbage and then a ding of now. So where's your where's your thought on like replies on on its on its impacts to to lists and to like at what levels? And then, you know, what's your thought on these warm-up tools?

SPEAKER_00

Wow, these are two great questions. Okay, I'll try to be succinct. Um, so number one, with replies, you know, the the easiest reply you can get is is with the welcome email, I would say, especially in newsletter space. Like that's the point where you can really create a sort of one-to-one human experience and conversation for those that um are providing the type of content that that would deserve that. Like if you're really trying to create that human newsletter experience, yeah. Um, and it's also good to to to put that in your welcome email because whatever you do in that welcome email is really setting the bar um of relationship. So the if you want people to be more engaged, you know, doing deeper engagements and opens, then you want that welcome email to do that too. And oftentimes I try to do like do a reply confirmation. If you can do reply tracking, don't just do like a click to confirm, do a reply track uh confirmation. And that will give you a very strong signal with the mailbox provider, possibly even get you on an allow list for that recipient, and so on. Beyond that, you know, that's gonna help set up uh make it a little bit psychologically more uh easy, I think, for your subscribers to do more than they are naturally inclined to do. Most subscribers are not naturally inclined to do more than open, possibly click if they feel like actually getting out of their scroll mode, uh, basically is what I would call it. Um, and so it really does become about that same thing I talked about earlier that you've got to do content strategy that almost unconsciously provokes a reaction. Um, it's the same with everything. It's like this on LinkedIn, it's like this on Twitter or or you know, any sort of um, you know, major channel. You the best way to get people to respond uh more consistently is to evoke a response. Now, I'm not trying to say like, you know, make them angry. I'm just saying like we all have natural impulses, whether that's um anger is obviously the easiest one, the other one is like, oh yeah, I felt that thing too. That's and it like excites you into like agreeing. Um, and then as well as like, you know, the the saying things wrong confidently often gets people to respond, to correct you. You know, uh, I'm not saying try to like trick people into responding because you you do need to have genuine relationship, uh and genuine reactions, not just little uh hacks, but that's the kind of direction your mind has to think. Like you, you know, a lot of times as marketers, we just assume that uh the presence of a link warrants its click by default, um, or even the presence of a link that we want clicked is enough to deserve it. And in reality, there's a whole we should be focusing first on the psychology of the recipient and what you're dealing with there. And so it really depends on the type of relationship you have, the type of like, are you trying to make it feel like human-to-human? Are you still keeping it sort of brand-to-human, you know, and then going from there and you want to work within that relationship and it's just gonna take testing. I'm doing that with my own LinkedIn newsletter, and I'm telling you that that you would think it'd be easy. I've done things like comments and emoji if you liked this specific content versus this one in my newsletter, because I I usually have multiple pieces, and I still don't get those comments. The only comments I get are the things that are natural reactions um that came from the subscriber. And so nothing that I asked for. And so and so we have to remember that and start focusing on what would earn the reply, not going, please reply. That's not gonna get it. So that's kind of the thing that I would suggest is just like you're gonna need test and try to work with that psychology and and start early and then and then go from there.

SPEAKER_01

Since, since uh just let me uh step in here real quick, since you know the the best entry point for reply is the welcome, then it makes sense. We we all we all agree with that. Um on the technical side, you know, everyone talks about like, hey, you gotta have your you know, your marketing here, your your transactional here, your whatever on these different IPs, maybe different subdomains for just a welcome email. Since it's such a powerful email, would you include that in your regular

Warm-up tools and the deliverability scam economy

SPEAKER_01

newsletter domain or or IP, or would you have them are they separated?

SPEAKER_00

Yeah, I mean it is technically transactional, but um that one would be an ideal one, especially if you're gonna do a reply, uh, to to just go ahead and be introducing everything, sort of where it's gonna be coming from and who it's from, um, to try to help, you know, just immediately build, okay. So this is the sender from this is the sender address I should expect each week. Even what, you know, I guess maybe not the time, because you usually want to trigger that, but even outlining the time in the email, like that welcome email should be your perfect intro into that, both both literally in the like content explaining things and also like sort of like representatively of like this is what the experiences look like, even to the point of you're gonna see the logo here, you're gonna see the content in this format. Like, you want to make the welcome flow straight into the the regular content.

SPEAKER_01

Yeah, perfect. All right, now let's hear your your thoughts on on the warm-up tools.

SPEAKER_00

Yeah, I'm not a big fan of the warm-up tools, uh, because number one, Google has even like blocked them because they really do not agree and and believe that it is an attempt to circumvent, which it is. Uh their security systems. And on the other hand, I regularly regularly help people warm up every day with no problem without using such tools. It's the email space is just filled with sort of what I'd call scammers. Uh, I mean, well, I guess they're just they are scammers, but I mean, I I should probably be nicer. I don't know. These are real brands and stuff, but they they they feed off of people's uh insecurity around inboxes and this sort of black box experience. And instead of bringing clarity, they are benefiting by with your ignorance and not actually resolving it. Um, and so I highly discourage anyone paying for tools they do not need. And most of the deliverability tools that are paid are not needed. Right. I agree there.

SPEAKER_01

I tell people I'm not a big fan of the of the warm-up tools. Um, you can warm up stuff with your own data, your own historical data and stuff, right? Right. Just move your most engaged people over first.

SPEAKER_00

You just do if you follow basic good practices, you you can warm it up yourself, no problem. And so it really is meant for either people who don't know that or the bad actors that are trying to get around Google's filters.

Lightning round: tools, myths, and the metric he'd delete

SPEAKER_01

Well, I think we're coming toward the back end of this. Is there any other topics or or things you think would be cool? Um, if not, have a few little wrap up questions here we can go through. Yeah. Lightning round. Lightning round if you want. Yeah. All right. Since we just talked about there's really no cool useful deliverability tools. What's one free deliverability deliverability tool most operate operators aren't using?

SPEAKER_00

I mean, I feel like I'd be surprised Postmaster tool. Google Postmaster tool is crazy, and there's a lot of people who don't even know it exists. So highly recommend that one. And I would, I would, I would, I would second that with Yahoo Sender Insights, because that's newer and no one's been using that one. I honestly mostly just use the organic data, even when I'm troubleshooting deliverability issues. Yes, because you have it. Most people don't have access to that. Well, that's a bad ESP. No, I'm talking about I use the in-account data that the customer can see. I oh okay, okay. When we do screenshots, when we we do audits for our enterprise accounts, and and we use the tools they have. We screenshots to show them this is how you can use our tool. So all right. If you don't have access to your ESP, you got a bad ESP.

SPEAKER_01

Well, yeah, I mean everyone's got their little thing, but they're also, I mean, the thing that you know, a lot of what's crazy is there's still a lot of ESPs that don't give you domain level breakdown. Right. We have that.

SPEAKER_00

So I guess we are so.

SPEAKER_01

To me, it's like I mean, you you can't like if you're if you're if if you're if you have a business, you're running in any size, mediumable size of a list, you don't have that breakdown, you're you're a trouble.

SPEAKER_00

You know, there's not really much free tools otherwise, you know.

SPEAKER_01

No spark post Well, there's like the there's like the uh melt mail mail tester, you know, send your email whatever.

SPEAKER_00

Um I don't love C tests either though. So what about what about any paid tools? I have friends at Socket Labs, they can integrate with a lot of different um like if you have a SendGrid or a Mailgun or something like that, and their dashboards are built to work on top of that and and give you what the others don't. And so I love my friends at Socket Labs and I think their tool looks cool. So yeah. What's one metric in every ESP dashboard that should be deleted? That's funny. I mean, my my instinct is actually I don't know that it's all of them, but my instinct is actually the the machine metric. The machine identification. Which is like the I think the opposite here, but that's truly how I feel about it. I I think identification is so unreliable that it's becomes it's become something that marketers are distracting themselves with when they can just still do the work without knowing.

SPEAKER_01

What about um this one where I kind of go back, spam complaints? Like to me, it's like you don't get it for Google, right? So if if it within the dashboard, unless you're pulling in from unless your ESP is just pulled in from Google Postmaster Tools, they might be able to help you kind of like see overall what's happening. Right. But and then a lot of these ESPs, especially the ones for starters that don't have domain level breakdown, yeah. It's like, oh, your spam complaint's great, but if you're 90% Gmail and you're 10% say Yahoo, your Yahoo spam complaints could be through the roof and you would never know. And then on top of that, you then you do got, you know, you get a Yahoo feedback group that's giving you the complaints, but they only care about the complaints that happen when whatever gets in the inbox. So if you're you're judging that on your full Yahoo send of 100,000, but only 40,000 inbox, now your complaints are based on that, not the full number. So to me, like the complaints are are our ideal uses for people. They I tell everybody, you gotta go to Google Postgre tools, you gotta go to Yahoo, you have to look at that every day just to make sure you're you're you're within your complaint.

SPEAKER_00

No, 100%. And you're right. It's the complaint rate is just so so broken because it can't properly represent the real world complaint rate. But I'm so greedy, I'm almost like a like a man dying for thirst in the desert. I'll take every spam complaint I can get. So I don't want it to go away. Um because that's as little as you can get. Now I wish we could do like like a per provider average, but then that would still get skewed because Google, you're gonna get zero percent every time. So what's the most persistent myth and deliverability? Obviously, spam words is the the biggest one. Um, right next to it, though, I would say, you know, blacklists being the biggest cause of people's problems, most marketers are not okay.

SPEAKER_01

Number one, you are have you ever heard of spot spam house? Oh yeah, right.

SPEAKER_00

Yeah, yeah. There's literally only a handful. So most of the time when people are like, hey, is my IP, or even we have shared IPs. Your shared IP is on a blacklist. Well, yeah, we send email. There's probably thousands of blacklists that technically exist. You could build one yourself if you have an at-home server, you know, and and put it on the web online. But uh the ones that actually are being referenced, you know, Google and Yahoo, and a lot of them are not referencing any other external blacklists anymore. Maybe, you know, Hotmail used to do really heavily with Spam House, um, and they're doing even a little less than they used to. Uh for most of Spam House has diversified their different blacklists. But at the end of the day, it's still there is like I'd say three to five that I would uh rely on. And most of the time, people are not really there. It's usually that they don't want to realize they've been sending bad.

SPEAKER_01

Yeah.

Where to find Travis

SPEAKER_01

Exactly. Well, cool, man. I appreciate uh the conversation. Thought it was a lot of good stuff there, a lot of great insights uh for the listeners. Um, you want to share anything about where people can find you? They want to sign up for your newsletter or something, or your LinkedIn newsletter, and you know, anything about the company?

SPEAKER_00

Thanks. Yeah, that'd be great. So I'm on LinkedIn, you can find me. Um uh I can get that the link to the LinkedIn to Chris. Um, but also um, you know, it's Travis Hazelwood. You can look me up. I I work at uh Ordo, but technically we just got acquired by Canva. So it you should see the Canva logo and it says uh says that on my profile. I have a uh newsletter on LinkedIn um because I wanted it to be searchable and referenceable very easily. I didn't want to just gate it behind an email inbox thing. And it's called Inboxicated. Um, and I think that's how we all feel about our inboxes and or even you know the the experience related there. And so it's a lot of fun. And I try to really utilize real-world scenarios, real conversations I'm having, even doing research from the ESP side of like trends that I've noticed around things, and I think it could be really fun. Um, so please feel free to join in and take a read at some of those. Yeah. Travis, thanks again for coming on. Uh, I really enjoyed the conversation. Thanks, Chris, for having me. This was so much fun. Yeah.

SPEAKER_01

Yeah. And and to meet in person, well, yeah, virtually. Yeah. Uh maybe we'll see each other at the next next show. Yeah, that'd be great.