Audience Bridge Insights #18 - The AI Standing Between You and Your Subscribers — with Chad S. White

The point of the engagement is to create influence.

The biggest impact that's going to happen is the actual AI learning for what's actually going to inbox. Just what's going to happen with that learning, to me, the scary part.

This moment is actually more disruptive than all of that.

You got to just write them very simple so they can't mess up what the heck you're trying to say in it. Can't be too complicated, too witty or whatever, because it could confuse them.

I think the inbox writers are still figuring things out. I think there's absolutely going to be a strategic advantage for somebody to like not have any AI.

If you're a multi-day sender, high frequency, you should never send an email to the same people more than once.

Promo emails are very well written. They need to monetize the inbox. I get that. But now you're being transparent about what that content is.

Welcome back to another edition of AudienceBridge Insights. Today I have Chad White, who is the Group Vice President of CRM Strategy at Zeta Global and the author of Email Marketing Rules, which I think is on its 4th edition or 5th, something like that. Chad, glad to have you on here. How you doing?

I'm doing very well. Thanks for having me on, Chris.

Appreciate it. So we're going to dive in quite a bit on the deliverability side, uh, what Chad's seeing. He's got obviously tons of of years of experience, more than I do, I would say. So we're going to dig into a lot of things that have been happening, especially in the recent years. So with that being said, you've written nearly 4,000 articles on email. You have the book. This latest wave from like 2024 to 2026 with all the new changes, the Gmail, Yahoo rules, Microsoft kind of adopting those, their new requirements, Apple tabs, the AI world and all the summaries that are happening now. I guess everything you've covered in the past, is this the biggest structural shift you're seeing in email?

I think it is. And that's, I think, a double-edged sword, right? It's, I think, pretty amazing to see InboxFighters investing so much in inboxes. I think that is a very positive sign. That is not to say that all the things they're doing, I think, are positive for marketers. I think, I think we all know that that's that's not true. But there certainly are some things that I think are pretty positive. I think the, the Yahoo-Googles, uh, requirements on net are very positive. Microsoft's, you know, sort of half signing on is positive. The stuff around AI summaries, which I know we'll talk more about later, I think considerably less positive. But certainly, like, the amount of change in these past years is definitely the most change we've ever had to grapple with. And I think, you know, and that's saying something because I think that exceeds the level of change that happened in the wake of the iPhone being introduced in 2007. That set off massive waves of redesigning emails, tons of talk about, hey, what's that, you know, what's that interaction flow look like? Where does it go? And everyone debating about how, hey, it looks like people don't want to convert on smartphones. And it turned out it was just because the experience was just so lousy.

They were wrong on that one.

They were wrong on that because the experience was just garbage, right? Email experience was garbage, the, the mobile web experience was garbage. And like, so that's why people, you know, went to desktops. But there was— people were completely misunderstanding the signs. And we were talking about mobile-friendly design for, oh my gosh, like from 2007, like the fury kicked off pretty hard and fast. We were talking about it pretty strong until 2013. And I think it wasn't formally like really put to bed. And I think until like, I think 2015 is when Google said, hey, if your website is not mobile-friendly, we're going to be like penalizing you. And I think that was the final thing that got everybody, all the laggards, to get up to speed. So I mean, that was a long, long time, and that was a huge sea change. But I think this moment is actually more disruptive than all of that.

The one, the one thing I, I will say on that is, is, I mean, it's been like, okay, design mobile first for everything, you know, for the last, you know, whatever, 5, 6 years, 7 years, however it's been. But I keep looking on the data And I'm still seeing a big percentage of desktop users, 30% or so on, on many places. So I don't want people to forget about that experience. There was always like mobile first, but don't, don't totally like push your desktop aside because you're definitely getting a lot of traffic from there as well. That's— you need to be user-friendly. It's a totally different experience than mobile.

So absolutely, it's both. You need both of them. And I think it's also really critical. We sort of— I think we lose this. Sometimes in the math, in the data, is that we'll look at what's happening among our subscribers and we'll say like, oh, you know, on this campaign we saw this engagement in these channels and we're like, okay, like that's our split. And that's, that's not real, right? Because people, people float between channels. We see the same thing with like dark mode optimization. Like we'll look at this particular campaign and we can tell like, oh, roughly this amount open in dark mode, um, and it'll be like, huh, okay, so it's, it's only this big, but no, it's, it's bigger than that because there are some people that switch between light and dark depending on what's going on in their environment. And so the numbers are actually wider, like it's bigger than that. And it's the same thing. Uh, only other thing I would add about mobile friendliness is that I'm still a little bit sad that, you know, 2007 was when the iPhone came along. There's still brands that are not getting mobile-friendly correct. Uh, and part of that I think is because they're adopting responsive design and they're washing their hands and they're walking away. And that's a huge mistake because I see brands routinely using responsive design, but their font size is way too small. They're like a 10-point font in there.

But it's stacking. It's stacking.

But it's stacking. It's completely unreadable. Like so small. And links way too close together. So let's not like tick the responsive design box and think we're done. Now readability is really important, you know, the experience around links and not having them too close together. So you're fat fingering the wrong link, like there's no patience for that, right? You think you're clicking one thing and something else pops up, that interaction is probably done. People are just gonna move on. So Yeah, that's, that's the thing that I think is really kind of amazing about like the mobile-friendly journey is that you still see brands, big brands that are whiffing it because they're not paying attention to font size and link density.

And then touching on the AI aspect of it, you know, everyone's worried. I know we'll talk about it later, but everyone's worried about the summaries and the rewriting of subject lines, preheaders. Like, why am I even writing stuff? They're just going to change it. But where, I mean, the biggest impact I'm going to— I think it's going to happen is the actual AI learning for what's actually going to inbox. They already do it now, you know, with all their algorithms. It's like you can get it in your spam, I'll get it in my inbox, he'll get it in the promo folder. It's so dynamic now based on domain reputation, even down to the subscriber level. Just what's going to happen with that learning and stuff is, to me, the scary part more than summarizing and things like that.

I think it's the mixture of all those things.

Oh, the mixture is—

Yeah.

The double whammy is not great, but—

Yeah. No, the summary stuff is what actually worries me the most at the moment. But you're right, it's a mix of all these things interacting with each other that I don't think the inbox writers have really figured out, honestly, completely.

Cool. Let's touch on sender requirements stuff. So It's been 2 years since the big bulk sender requirement change Gmail, Yahoo adopted, and now Microsoft is jumping on board. Who actually got hurt from this shakeup? Was it the spammers or is actually legitimate high-volume publishers who just, you know, didn't have their ducks in a row, a little bit sloppy, especially on the authentication part that really hurt them?

Yeah, I think in general, this is actually a really good thing. You know, historically, deliverability was a black box, right? And, you know, I know quite a few deliverability consultants and it's like it's the dark arts, right? Like they're like reading the tea leaves and for legitimate reasons, the inbox fighters don't want to tell us too much because when they tell us too much, then the spammers game them. And so it's really been just incredibly black box. Like my entire 20-plus years in this space. And so it was really good to see them say, hey, look, you need to fully authenticate and have a DMARC record published. You need to keep your spam complaints below this threshold. I think it was like, you know, below 0.01, definitely below 0.03. That's really great. They're just being explicit about it. Didn't have that before. So I like that for a lot of the brands. Like, I historically have worked with a lot of like very large enterprises, you know, think like very large retailers and airlines and things like that. And these requirements were almost a non-event for a lot of those folks. They were already at— already had DMARC records, already at P=REJECT, which is not even where we're at yet with those requirements. They already had list unsubscribe headers. Maybe some of them weren't, weren't the web links. They were like the email links. So it's just like a small change. And really, for the most part, inbox for ESPs, email service providers, were handling the list unsubscribe stuff. Generally it was at the ESP level. And for the most part, like those settings for spam complaint rates were very reasonable. And a 2-day unsubscribe process. In this day and age, incredibly reasonable. I think people look back at CAN-SPAM and they were like, wow, like you're going for like CAN-SPAM says 10 business days and now you're saying 2 days. But CAN-SPAM was passed in 2003. That was like 100,000 years ago in like tech years. And so like nobody, even then the 10 days was really to accommodate for brands that had like highly distributed— think like franchisees and branches and like insurance branches and things like that where they're like sharing unsubscribes like in batches overnight, you know, and it was like kind of cascading through the system. Like that's what that 10 days was for. Like 2 days in this age, totally fine. That's not going to break anybody's back. So again, I thought those were all really positive. And so, and really I feel like they're trying to get ahead of what AI is going to do to spam. I feel like that was like a big driver of these requirements. So that's how I kind of read it in that context, is that they were actually trying to be proactive. They kind of see where things are headed. I think already we're seeing a lot of AI-driven spam and they were trying to get ahead of that. So I Nothing but applause for that particular move. Now, that said, did it hurt some senders? Absolutely. Tended to really hurt, uh, senders who are on shared IP addresses. A lot of, uh, those ISPs specialized in shared IP addresses, like smaller senders. They definitely struggled to figure out exactly what they were supposed to do. And like, there was a lot of uncertainty, and I think all that's kind of settled out. But for those folks, for smaller spenders, smaller senders, and for ESPs catering to them, I think it was like a good rough year for those folks. And that's unfortunate because it's obviously not like their fault at all, but it was definitely rough for them.

You've said inbox providers are now standing between the brand and the subscriber before the open. Can you spell out what that means mechanically? And what exactly is Apple Intelligence or Gemini rewriting, compressing, or reprioritizing?

Yeah, so there are like sort of two different things happening in that sort of AI summary space. So they're, and they're recognize that they're very different animals. So on the one side, you have the AI summaries that are happening that are driven by inbox providers. So Gmail and Yahoo Mail. Replacing preview text with summaries. In the case of Yahoo Mail, sometimes replacing subject lines and preview text with AI summaries pre-open. And then post-open, Gmail doing deal cards that they're automatically applying to emails. Yahoo Mail, in the primary tab, they have little— I forget what they call their AI summaries that are very much like Apple Intelligence summaries. So that's like, that's largely happening with no consumer choice, right? And certainly no marketer choice, right? I think by and large we would love for all that to go away because it's deeply problematic for marketers. Because as you point out, like, we're being disintermediated. Like, they're getting in between us or reinterpreting. We're playing this weird, weird game of telephone where we're saying something very clearly and the inbox fighters are repeating it sometimes accurately, most of the time accurately, but sometimes very inaccurately. And it's like, why, why are we playing a game telephone here? It's like you're not adding value, right? You know, so why are we doing this? It was a really interesting story that I was at. I was at Litmus Live. I think it was just last year and they had a panel of a lot of these inbox providers, and they were talking about AI summaries. And there was this really interesting moment when like 2 of them on the panel were like, oh, PTA emails, indecipherable, like crazy email, like really hard to understand what's going on. That's why we have AI summaries, is to like fix that problem, like pointing at PTA emails.

PTA, bro. PTA meeting.

And like, okay, I get it. Like, I have, I have 3 sons. I've gotten a lot of PTA emails and it's true, they tend to be like sort of long and rambling and kind of all over the place. And I was like, but I'm like hard pressed to see how that justifies all of this activity, especially when most of these AI summaries are not being applied to PTA emails. They're being applied to the promo tab, right? To commercial emails. Which, you know, I'm gonna go out on a limb here and say that PTA emails, not written by a team of highly trained people in the email marketing space, unlike commercial emails written by a team of highly trained people who are great at communicating. I think that promo emails nowadays are very well written and subject lines are really clear. The preview text is really clear. Like, I think we've got it all buttoned up, and that wasn't always the case. But I think by and large, like, you know, A-minus at least. Like, we are doing so well, and to go in and like kind of tear all that up, like, that's where, that's where AI summaries are really having the biggest impact, is on the emails that arguably have like the most considerate, like, most amount of time invested in them. So it's It's a bit odd for sure. Now, that said, there has been some encouraging motion there. So for instance, Yahoo Mail, they used to have deal summaries after you open the email at the top of the email, pushed your email content down. You have these deal summaries and as of like a couple of months ago, they've largely transitioned those, those deal summary spaces into just ad space and now there's just ads. And I think that's great because before it was like this very odd mix of like some stuff that was pulled from the email, some stuff pulled from other emails in your inbox, and then ads. And like you never knew what the heck you were clicking on. And so I feel like let's just, they just turn it into ad space. Like they need to monetize the inbox. I get that. That's fine. But now you're being transparent about what that content is. Whereas before it felt like a little bit. Yeah, maybe deceitful is a strong word, but like it wasn't clear for sure. All right. So, so that's the one kind of AI summary that's all driven by inbox writers. The other one is, like you mentioned, like Gemini. There's definitely a growing number of people who are using Gemini and other LLMs to summarize their inbox. This is a consumer making the choice. They're like, hey, tell me what emails arrived overnight, you know, or the last time I checked my inbox, and just give me a summary, and then I'll decide, you know, what I need to go and like look at in more detail. In my mind, that's very different, uh, because that's a consumer making that choice. And we tend to see much more moderated behaviors when we see consumers like controlling those choices versus just like by default Everybody has to use something. So that's, that's one of the problems I have with these, the AI summaries that the inbox writers are kind of forcing on people is that you can't turn those off. The consumer can't turn it off, brands can't turn it off. And in both cases, when it's wrong, consumers neither market or marketers, they don't have any recourse. Right.

I tried, I tried integrating one of those and I had it on for like 2 days and I turned it off. It was just driving me insane. Like it was like trying to like bucket everything in certain folders and this and that. And I was just like, I can't live like this anymore.

Yeah. But it's good for people to have choices. So I tend to have the view that consumers make very good choices and when you take choices away from them, it's bad. You give them choices, then it's good.

Right. But that's the— that statement right there is actually the good thing is that, and I had this conversation, I think, my last podcast with Dan Oshinsky, we were talking about AI from the inbox and stuff, and he was saying like, you know, the email inbox providers, they don't even know what the email is going to look like a year from now, 2 years from now. They're introducing stuff all the time. They're testing it. How are people using it? What's the user experience? And they chop more things than they keep on the platform. So you're not, like you said, they had the summaries up there for the and now it's just the ads. Like, they'll continue to do things like that.

Yeah, I just don't— yeah, I do think we're in a period of flux. I don't think people should get too excited to think like, oh, it's always going to be this way, or, or worse, right? Right? Like, I don't think that's how people should think. I think that— I think the inbox writers are still figuring things out, and I think over time it's gonna change. I think there's absolutely going to be a strategic advantage for somebody to like not have any AI, right? Like there's absolutely a growing backlash among consumers, and some consumers are going to say, hey, you keep forcing this on me, I don't want it, therefore I'm going to switch, you know, or I'm going to do something else. Um, but you also asked like, all right, how, you know, how is AI, you know, like interpreting content? Like, you know, so I read in reading into that question is like essentially how can we influence AI summaries? And there are some things we can do there as well. Obviously having a good clear subject line is great, but in terms like the biggest thing you can do is HTML text. HTML text in the email is really powerful. You got to have that. Alt text is not routinely ingested. Sometimes it is if you starve it of HTML text. Sometimes it'll, it'll kind of pluck out some of the alt text, but in general HTML text is the way to do it. I've seen research indicating that hierarchy is, is also important, so stuff that is higher up in the email gets greater priority. With some of our clients, we will have, you know, invisible, like white on white, uh, like pre-header text that provides a bit of a summary. We find that that tends to be pretty successful in sort of guiding the summary in the right direction that we would like rather than, you know, rolling the dice because we certainly see AI summaries pulling stuff from the legalese at the bottom of emails like that. Legalese, all those disclaimers have gotten really—

become really dangerous copy, like, because you'll see out of control, especially in the finance space, they have like freaking 1,000-word essay at the bottom with the disclaimers. It's like, yeah, the—

and I understand why that's there and why you got to have it, but that is introducing quite a bit of risk in terms of what an AI summary is going to say or what a deal card is going to say, because we've seen, you know, deal cards in Gmail like pull out like percentages out of like disclaimers and stuff. And how like, oh, employees, like their discount is capped at X, like, and then X shows up in the deal card, which of course is not what you want to show up there.

Right. This next topic is what keeps me up at night, what keeps a lot of my clients and colleagues up at night. We'll just call it the unfair engagement loop. So AI is summarizing your email and you were saying like sometimes it could be bad, sometimes it's okay. But in general, like if it keeps on going and you're getting all these AI summaries and they're rewriting things and it affects your open rates, it affects your, your click rates, it affects, you know, what's happening within your emails, then theoretically it's going to affect your, your reputation because you're not seeing enough engagement with the subscribers that you have. And then you're going to get penalties for that. And then all of a sudden you're not delivering as well. But then the summaries are like there's this whole loop of like, I mean, how are you going to be able to know how to deliver if engagement, which is always the primary, you know, both positive and negative signals was always a primary thing that we looked at to, to judge our deliverability. What's going to happen there? How can we even figure this out? Or is there not going to be an inbox anymore?

Yeah, no, that scenario is something I worry about as well. So traditionally there were a number there have been a number of factors that influence sender reputation, and most of them were, were attached to the sender itself. So, so just to quickly list them. So the first one was infrastructure, and that's really the main one that's controlled by your ESP. So like your choice of ESP definitely can affect your sender reputation. You know, how emails are sent off, how throttled, you know, to the volumes that inbox providers want to see, right? They don't want to see like a gazillion emails go out and like hit their servers all at one time. They want to see a particular rate. They want to see, you know, your IP and your domains warmed up appropriately. So the infrastructure there is important, uh, volume is important, the content of your email is important, although today it's mostly like code-based, uh, issues. And link reputation, that's big, uh, complaints, IP reputation, engagement, uh, domain reputation. Those have been like the traditional 7 for a long time, and all of those are your choice, right? Your behaviors, what you are doing, how your subscribers are responding, your choice of ESP. Those are all within your control. And now Inbox AI, as you pointed out for the very first time, inbox providers are measuring their own performance. If their AI summaries of your content are good, are accurate, that's— that helps you. And if they're bad, that hurts you. And And that's really like a— I can't overemphasize, this is the first time inbox writers have ever measured themselves. They've measured their own performance, the performance of their AI systems when measuring sender reputation. And I think that's really, really huge because you're absolutely right that I could send an email, it could be a great email, and the AI summary could mess things up, get the details wrong, make it not appealing. I've seen all kinds of instances of this, including instances in Gmail where like retail brands send an email and Gmail says that the deal is expired when the deal has not expired. Like if I'm, if I'm, if I'm a subscriber and I see that it has an expired tag, you know, on the deal, like, am I opening that email? I'm probably not. Like you just, Gmail's like just torpedoed that entire send and things just roll along, right? Like there's no, there's no recourse. There's, there's nothing to be done. You just like your, your campaign just got sabotaged, not intentionally, but, you know, the net-net of it is that very few people are going to open an email that has an incorrect expired tag on it. But like, that's what we're talking about. And that's, again, 100% brand new. That is a very, new challenge that we did not have to reckon with before. And that's tough.

Do you think there, there could be some, I mean, there's always regulation stuff they're trying to pass, you know, especially when it comes to email privacy. Do you think there's a case of a policy kind of enforcing them to not be changing this stuff because it's not what the sender wanted to give the subscriber? Like, is that so-called legal to just—

Yeah, I don't think that there— it's not. I've talked to some lawyers and it's not clear that what's happening now is necessarily illegal in and of itself. But I do wonder, like, the, the area of the law that I am concerned about is about, like, It's like misleading, like, you know, ad copy, like misleading copy. So like, I'm kind of surprised this hasn't already happened, to be honest, because we are so litigious in America. But like, for instance, in deal cards, one of the things that Gmail really struggles with is, is BOGOs. So buy one, get one. And so I will frequently see brands that— and I think maybe, maybe they're— maybe Gmail is on to this. I think I've been seeing them less lately, but like it's very common for a brand to have like a buy one, get one 50% off and the deal card just says 50% off. That is a huge difference between buy one, get one 50% off and just 50% off. And I'm surprised that there hasn't been a consumer that has tried to take the retailer to task to say, hey, you know, your email says 50% off, give me that 50% off and like sue them for that. And then in court, you know, of course the retailer would say like, nope, we didn't say 50%, we said BOGO 50% off. It's actually Gmail that said it was 50% off and that was inaccurate. But like Gmail isn't a party to that lawsuit. So then that spawns a separate lawsuit. But the honest truth that everyone has told me is that nobody wants to sue Google. Like nobody wants to like flush 5 years of their life down the toilet for very uncertain goals at the end of it. And that's sort of the problem with all these is that all these inbox providers are also just massive tech companies with, you know, fleets of lawyers. Nobody wants to take them on, but I do, like, I do wonder about the legality, like, of that space about how, you know, hey, that deal, you know, is not being accurately portrayed.

That— Yeah, I think that they have to change. Like, they shouldn't be trying to extrapolate deal cards out of content. Like if they want to do it, just have the metadata forced to be included in the email and they'll pull that. They don't have it, don't show a damn card.

Yeah, no, I mean, that was how this all started. It started as an add-on, right? So brands were encouraged to add deal annotations to their emails and the feedback loops like weren't good. So brands could not see if like these were like extra coding effort and they couldn't see if it was actually making a difference. So they stopped doing it. And, you know, like at one point I think it peaked at like 5, maybe 5% adoption, and then it went to essentially to zero. And that's not what Google wanted. So then Google started using automatic extraction to apply it to emails that it seemed to think was appropriate. And so, and it has like a mixed track record of getting that right. Oftentimes it gets it right, but sometimes it gets it very wrong.

Well, maybe hopefully now with AI and everyone create using it to help be more efficient with their emails and stuff, everyone could just use it for creating their own cards on the fly, putting in the right schema or whatever is needed.

Yeah, we certainly have encouraged brands to do that more often. Like if you have a deal that can be easily represented in a deal card because deal cards don't handle all kinds of deals well. But if you are sending like a, you know, a straight up 25% off sale kind of thing, like code it, like, you know, part of what we're struggling with is, is accuracy. And so if that's a step you can take that will make that deal card more likely to be accurate, you should do that. Now, the frustrating thing is that even if you do that, it's a little bit of a dice roll whether or not Gmail is going to honor that. And that's super frustrating, right? When you as a brand take the time to write that additional code into your email, to have Google just ignore it and then like automatically extract its own content, that's super frustrating. But that happens.

I mean, I would say that there's got to be a liability on that one.

I mean, come on. I think if that ever did go to court, that would be harder to defend. Like, right. Yeah, we even included the correct content, but now it just got ignored. Yeah. Again, I think it kind of goes back to You know, nobody wants to sue Google. No one wants to sue Microsoft. No one wants to sue Yahoo. Like, these are really big companies. And certainly in the case of Google, I don't think anybody wants to sue that.

Well, Amazon is just as big.

Yeah. I don't think people want to sue Amazon either.

Anyway. Amazon could sue Google.

Oh, yeah, well, yeah.

They're messing with their deal card or anything that they have going on.

Yeah, I don't even know that the— I don't think the big guys generally want to fight each other. Either. Again, I'm hopeful that this— that we won't get to that, that point. I know, again, I'm hopeful, like, I'm really encouraged by what Yahoo Mail has done, right, where they were doing these deal summaries. The experience was, I think, not good for consumers, and they evolved that space into simply an ad space, which is totally cool. I don't have any problems with that. You know, the content is clear, it's clear that it's ads, it's clear what happens when you click on them. You know, and then the consumer can decide, you know, am I interested in that ad or am I not? Or am I just going to scroll down and engage with the email? That's fine. So I'm hoping that some of this stuff will self-select away because, I mean, I think everyone is, is realizing that, like, you know, AI is not cheap, it's not free. And so, like, they're also like increasing their overhead to supply this. And I think at a certain point they're going to wrestle with, okay, is this additive? Of, is this better? Is it creating a better experience for consumers? At what cost is it creating a better experience? Is this giving us a competitive advantage, or is it just ticking people off? I don't think we've— like, I think Yahoo clearly, like, reached that, that threshold. And I think that, you know, the Gmail may not be far behind them and having, you know, kind of rationalizing their use.

Yeah, because of that actual, uh, fact there is why I think that now, especially with the full adoption of the list unsubscribe in the header and everything, that they're going to just start auto unsubscribing people from people's lists that aren't engaging, aren't activating or whatever, just so they stop getting all these emails that they don't even need to process. They don't need to do stuff in to save on infrastructure costs. They have the technical setups all there. They'd be like, oh, this person hasn't opened an email in freaking a year. And a lot of the senders don't manage your list very well. So let's just say this guy unsubscribe them for people that have like full inboxes. This guy's not using his email. He hasn't logged into his email in freaking 3 months. Unsubscribe him.

Well, I mean, I think we already have essentially a process for that. You know, it's called the spam folder. So like to have messages go in there and eventually just get blocked. So I would be surprised if inbox writers were to take that step to like automatically, like to actually unsubscribe someone versus just, you know, routing those emails to spam and then eventually blocking them. Right. Like, I don't think they need to, but they already have a process for handling that stuff. That would be alarming to see because unsubscribes come from people.

Sort of the emails that we send people, the right copy and the right text and the right link.

Yeah, no, that would— well, that would— I would consider that to be quite an escalation of where we are now. That would be Yeah, definitely worrisome.

I think you touched on this a little bit ago, but, you know, writing the preheader text was always a craft, like how we're going to support whatever our subject lines are doing. Subject lines are obviously always a thing as well. If the machines are rewriting both, what is there for us to optimize? Or how should we optimize? Is it like you got to just write them very simple so they can't mess up what the heck you're trying to say in it? Can't be too complicated, too witty or whatever, because it could confuse them.

Yes. No, I do think that that's where we're at. So before I got into marketing, I was a journalist. And, you know, I remember when SEO really came on the scene, it had the effect of taking all of our very clever and engaging, you know, titles for stories. And all of a sudden we were rewriting those, those really creative titles into very boring, straightforward titles because we were now optimizing for search engines. And I think that's kind of where we're at now with email is the creative subject lines, like the really clever ones. I think this has hastened their demise. Clarity, clarity, clarity. That's the name of the game. And it's both good for helping inform AI summaries. Definitely subject lines are ingested. You know, preheader text that shows up as preview text that gets ingested. So these are very important bits of copy that have very high hierarchy, very high priority in summaries. Those are really important to get right. Also, if an AI summary whiffs on understanding your email, you want your subscriber to at a glance be able to like, maybe I didn't understand that, and then look at the subject line and be able to know Hey, what's the truth? Right? So like, I've gotten, I've gotten emails that say my package is now going to be delivered on Tuesday and the AI summary says it's going to be delivered on Monday. And so it gets off by a day, but the subject line says right there that it's going to be delivered on Tuesday. Like that is the voice of the brand, of the sender. Like that becomes the truth. And I'm not saying that's not confusing, But definitely, you know, the subject line is— and in the preview text is your opportunity to say what is most critical, like, and to guide the AI summary. So like, it's those two points.

Got it. I've always said, I mean, everyone, not just me, but, you know, opens have basically been a dead single for a long time with MPP prefetching, security scans, bot opens. Is this AI layer now going to just kill it for good and are clicks next?

Yeah, so I, I don't really share that view that, that emails are— that, that opens are dead. You know, I feel like all of these changes have made it so we've kind of pushed email opens into the category that they always should have been in, which is that they're, they're health metrics. They don't really tell you the performance of things. For most brands. I know in publishing it's a little bit different, it's sort of a KPI, but for most brands what we've seen is opens being used much more appropriately as a health metric, and, and that's good. I don't know if, if everybody here in the U.S. is aware of what's going on in France, but the privacy, the privacy organization there has just, you know, 8 years after GDPR decided that GDPR actually applies to open tracking as well. So as of, as of July, mid-July, a rule went into effect where you have to ask for consent for different kinds of open pixel tracking data usage. And so there's not just like, hey, can I collect, you know, opens to use them for whatever I want? You actually have to ask for specific use cases, which nobody's going to do. This is really It is a really, really high bar. But even there, the privacy organization there, CNIL, or I think it's just called CNIL, actually has recognized that open tracking is important for inactivity management, so for managing deliverability. And so there's a carve-out where brands can save the last open. So they can't keep a history without an opt-in, but they can keep a record of the last time an email was opened from you as a way of being able to manage inactivity, to understand, hey, is, is the subscriber engaging with these emails, you know, so that I can either stop sending them emails or send them at a lower cadence or do something different. And so there is this sort of fundamental understanding that that is one of the core uses of opens. So it was great to see Neil recognize that because I was not actually in their initial draft of this, this new rule. So they kind of, after consulting with, uh, with marketers, they added that in, which is fantastic. But I think for most brands, like, that's one of the most valuable uses of opens. And certainly the other thing with opens is like auto-opens. A lot of brands now get auto-opens. And that's not engagement, but it's an amazing inbox placement signal. I don't think that's what Apple wanted to give us, but that's what they did. Uh, if you're getting an auto-open, it means that email is landing in the inbox. It's not landing in spam. And so when that signal goes away, it means you're landing in spam for that particular user. That's an incredibly valuable signal. So there are absolutely uses for opens now, but I do think that in general everything seems to be pushing engagement measurement. You know, it's like it's moving from the campaign level up to the customer level, and at the same time it's spreading out from the email channel across all channels. So you need like a much more holistic view of your subscribers, like as customers. And that's kind of always been the case. Like 10 years ago, if you had a subscriber who you could tell was a very happy customer, right, you saw engagement, you saw conversions, you saw all this activity around them, but you didn't really see much email activity or any email activity. Like, I would have told you 10 years ago to keep that subscriber. Keep emailing them because our visibility into email engagement is, is not great. It wasn't great 10 years ago. It's certainly more questionable now. Um, and so that advice I think has just become more vital. Like, if you were, if you were seeing at the customer level a very healthy amount of engagement, you know, if you're a publisher and you're seeing, you know, uh, engagement in your app If you're seeing web visits, but you're not really seeing a lot of email visits, it could be because they're getting that email and they're seeing, you know, the subject line, or maybe they're seeing, you know, a summary. And they're like, ah, all right, I'm going to go check out like their website today. And so like maybe they're not reading the email, but it's driving them to your website or to your app for content. And that's the name of the game, right, is engagement. And influence.

Got it. Saw that the Atlantic CEO warned this spring that a Gmail newsletter tab could be coming out, separating out for different publishers the way promotions did for retailers. Obviously, it doesn't exist right now, but Gmail does have the new AI inbox, which you can have on or off, which filters out whatever it decides is unimportant. And I think that came out in January or something. Is that basically the newsletters tab by another name? And what does the promotions tab history tell us about what could happen with this newsletter tab if it does come to fruition for all those publishers, media brands that are out there?

Yeah, I think that marketers love to demonize tabs, and I don't know that the evidence really backs up these accusations. So tabs, obviously not new. 2013, Gmail introduced tabs, and so they've been around a long time. Consumers are really used to them. They know how they work. And it's been interesting because in recent years we've seen a lot of mistabbing by Gmail. So we've seen, for instance, retailers all of a sudden not in the promotions tab anymore. They're in the primary tab. And what we tend to see happen is that opens go up quite a bit. But clicks don't budge. And really what that tells us is that people are like, what the heck is this doing here? And like, they'll look at it and they'll be like, oh, all right. They're like, there's this idea that if I could only get more opens, I would get more clicks. And that is just— that is not borne out. Like, that is just not true. If you can trick people into opening your email, then they're magically going to be more interested in your email content. That, that just magical thinking is not how the inbox works. And that's why I think on some level, the fact that we're being forced into having more descriptive subject lines is a very good thing. That way you're not, you know, you're attracting the right people to open that email because like the funnel, you can't optimize the funnel in its individual layers because it's all the same person, right? So you want your subject line to help subscribers self-select into the funnel, right? Because you want them to not just open it, but to click, right? And ultimately convert. And to do that, you need to have alignment all the way down. So if someone opens, like almost the last thing you want to happen is to have someone open and go, oh, that's not what I thought this email was about. And then go away because that's going to hurt your chances the next time around, right? You really want them to read that subject line, open it and say, yep, that's what I thought this email was about. Great. And then click through and like hit a landing page was like, yep, yep. That's also what I thought was going to happen next. Like you want that all to be connected. And so this, this magical thinking of like, oh, if I could just be extra clever, extra entertaining or extra vague, that will just make people really curious and they'll open it and they'll go like, oh, man, you really got my attention and all right, and I'll buy your thing or I'll engage or whatever. Like, that's, that's not how it works.

So I think it was like a couple of months ago or maybe a few months ago, the accidental experiment happened where like Gmail was just pushing everything in the inbox for like a 24-hour period or 48-hour period. Forgot what it was. And like, it's exactly what you said. Like, no more clicks, no more conversions, open rates spiked up a little bit, but only thing that, that, that really was affected is unsubscribe rates went up.

Yes. Yeah. That tends to be like, you get more opens and you get more unsubscribes. Probably not what you want to happen. Yeah. So yeah, again, there's, there's, it's not like a ton of data, but we have like these instances, especially with Gmail in recent years where they've, you know, they've had some issues with their tabbing algorithm. And so we've been able to kind of peek and see like What happens if we were suddenly more in the primary tab? And it's not good. I think what it shows is that, you know, first of all, in Gmail, like, anybody can move anything out of the promo tab into the primary tab and have it show up there all the time. Like, they can, they can do that. Um, so if they want that, they could do that. Um, but I think what it shows is that people are very used to tabs. It's not some mystical black hole. Sometimes we talk about the promo tab as being like another spam folder, which is like the farthest thing from the truth, right? Like, people are very comfortable with that. And I think in general, it's, it's very positive. Anything that makes the inbox a more positive, controlled experience for our subscribers is a good thing. So I don't get terribly alarmed with the introduction of tabs. I want to say that Yahoo Mail actually had a newsletter tab at one point. I'm not sure if they have it anymore. I'm pretty sure that didn't, like, you know, eradicate, newsletters or do something horrible to them. So I don't get too excited about that. Two other things I would point out is that yes, uh, you know, Apple's just introduced tabs. It's worth pointing out that half of people roughly have turned off Gmail tabs. Like they just decided, I don't like—

I don't use them.

Yeah. They're like, I don't like this. I don't like tabs. I want everything. I want the monolithic inbox. And again, roughly half of Gmail users have just turned it off and they go back to the monolith. I think the exact same thing is going to happen in, in Apple Mail over time. And I think the exact same thing will happen with sort by relevance versus, you know, sort by recency. The same thing will happen there is that some people will like that and some people won't. Do I think we need to bring the social media algorithm to the inbox? I don't think we need that. Um, I think people can make good decisions and for themselves, but some people will like it and some people won't. And so I think all these things that we think have like a really dramatic impact, I, I think some of them don't. But again, the AI summaries where people don't have a choice, that's, that's the part that gets me the most alarmed. But I feel like when people have the choice, a lot of them opt out, and then, you know, the people who have it, like, it works for them, and c'est la vie.

Cool. I have a few more questions. It's getting close to the end of the hour. I don't know if you have a hard stop.

Nope.

Okay, cool. So most deliverability advice out there assumes people are mailing once or twice a week. A lot of it goes to B2B, e-com, things like that. My audience mails daily, sometimes twice daily, sometimes more than twice daily. An actual send as a publisher, a media brand is the product. Where do the standard guidance actively hurt a daily sender?

Yeah, so I'm not sure I love the language. Like, the guidance is certainly not designed to hurt anybody, right? The guidance is a direct reflection of what inbox providers allow senders to do. So like, that's, that's the heart of the guidance. Uh, over a lot of my career, I've worked with a lot of retailers and again, a lot of like other like B2C brands that are sending once a day or twice a day or 3 times a day in some cases. So definitely I've worked with some, some high-frequency, right, uh, senders. And it's absolutely true that they have sort of a higher bar, um, because someone— if you're a brand who's sending once a week and someone goes inactive and they stop opening your emails over the course of a month, if you're sending once a week, they've just— they have not opened the last 4 emails, that's not nearly so dramatic as, oh, someone stopped opening your emails for a month and now that's 30 emails that have gone unopened. Like that is going to much more dramatically affect your engagement rates. And, you know, as I mentioned earlier, engagement is definitely one of those factors that impacts sender reputations. It's a pretty sizable one nowadays. Brands don't want to— I mean, it's like inbox fighters don't just want to see that their users are tolerating your emails. They want to see that they're engaging. But like, I don't think that those engagement bars are, are really unreasonable. What's in most cases what happens with these high volume, like retailers that I've worked with is that when you start to see a long period of disengagement, you're not seeing any activity. The wise thing to do is to put them into like a re-engagement segment where you're then sending them emails like at a much lower cadence. So if you were sending every day, now you're sending like twice a week or, or maybe once a week until you see engagement again and then you throttle back up.

Right.

That's the kind of thing that I would also recommend for media companies. And it's obviously a little bit different because there are expectations, but You know, what are you going to do? You have to protect your sender reputation for all of those subscribers that you have who are actively engaged. And that's always been the trade-off is, all right, yes, I absolutely want to protect or try to, you know, keep people on the list, you know, that are disengaging because, you know, some of them are going to reengage over time and, you know, those are going to be good customers, right? But you can't allow pursuit of that to threaten all the subscribers who are actively engaged with your emails and expect to get your emails. You can't allow one to jeopardize, you know, the fringe, essentially, like the, you know, the hopeful, you know, 20% that you have to jeopardize the 80% that's driving your business. And that's, that's always been the calculus. And so I would just remind folks that that's what the math is, unfortunately. You never ever want your most marginal customers to jeopardize your best customer and the types of engagement and experiences you're building for them.

Right.

That's, that's just the math. And so, you know, you should be constantly monitoring your deliverability and monitoring your engagement to be able to understand for your brand, for your subscribers in the inboxes that are most important to you, where are those lines? Where do you start to get in trouble? That, all right, if I am having, you know, if I'm looking at inactivity at the 3-month mark, you know, and taking folks who are inactive for more than 3 months and putting them in a special segment where I'm emailing them a lot less, is that enough? Is that enough to keep my deliverability, you know, really high? Or, you know, does it need to be, you know, does it need to be 10 weeks? You know, or am I doing fine? And maybe it could be at the 4-month or the 6-month mark. You know, everybody should be constantly kind of, you know, doing that math and watching that. And that's just the— that's the reality of things. But I don't think that those thresholds at most inbox providers are unreasonable. It would not be my take.

So I've always had the policy and I tell everyone that we work with that, like, if you're a multi-day sender, high frequency, you should never send an email to the same, same people more than once. It should always either get tighter, like you might have one that's like your looser segment with your bigger send and then you're tightening up for the next one. So it kind of controls the frequency automatically. You're adding triggered sends into the game where you're really getting more engaged people or people who have engaged in the last week versus 3 days. And that's how I have them segment who they're sending to, to send more frequently. Because I'm going to feel like if you send 3 emails in one day to say you have a list of 200,000 people, you send 200,000 people all 3 emails and 150,000 of them don't open all 3 emails. I mean, you're just telling the inbox provider that you don't really care about these people. You're just sending them all these emails they're not caring about, and it'll, it'll take your reputation faster than tightening up across the way.

Yeah, certainly. Yeah. Now, one of the, you know, in the age of AI where it is becoming easier for brands to produce more emails, one of the things I would love to see more of is more segmented sends. I would love to see more brands replacing a broadcast send with a segmented send or, or spinning up more triggered campaigns that really leverage engagement in the retail space. There are somewhere in the neighborhood of like 15% of retail brands get the majority of their email engagement and revenue from triggered campaigns, which generally represent like 5% or less of their email volume. Like, that's a great opportunity for brands, you know? So like, that shows you what's possible. When you are responding to interest in a very specific and targeted way, the right content. And I think publishers, you know, have some similar opportunities that they could, they could leverage. Certainly when you're talking about segmenting by like interest and stuff that like that follow-on story or that related story and then going omnichannel with your engagement.

So if you see them on the website, you know, on a certain page or, you know, whatever, then might be a time to trigger them something. Or if they're in, in their app or they're in receiving a text, an SMS message, and they reacted with that one. Okay, well, maybe you can follow it up there. So don't just think that the only engagement you're looking at is an email. You can use it to other channels to trigger things within email and the other channels themselves as well.

100%. Absolutely.

So I work with a lot of financial publishers and they monetize with sponsored sends and ads in the email. How do you think or see AI summarizers treating commercial content, editorial emails, and should publishers be restructuring how ads sit in the template?

Yeah, to my knowledge, AI summaries do not differentiate between editorial content and commercial content. I don't— I could be wrong, but as far as I know, there's, there's no difference one way or the other given. Um, I mentioned before that like hierarchy is definitely a consideration. It's not like everything, but, um, but certainly content that's higher up in an email gets more priority. I think you do need to be a little bit careful here. Like if the goal is to try to get like advertising content into the summary, like that may not be actually the thing that you want. Like, they— that may not drive engagement with the email the way that maybe you think it would. Like, right, maybe that would be good for the advertiser but the bad for the publisher more broadly. Um, so I think you have to be a little bit careful there. But as, again, as far as I know, there's no differentiation between different kinds of content in an email.

Right. I'll see, you know, I'll see publishers will kind of have a— put their obviously native or commercial ad at the top of the page and then add, you know, content and articles underneath it. But then sometimes the subject line will be tied to the— would be tied to the article. But the native ad is that, like you said, hierarchy is all the way at the top. So kind of really pushing that one first. But then that's the whole almost tricking them into the open. To see the ad first and doesn't match whatever the subject line was versus just putting maybe the, you know, the subject line or headline of the offer up there so that you're not really confusing.

Yeah, you need to be a little bit careful. But again, the hierarchy thing is important, but it's not like absolute. So it's not like, oh, if I put this ad at the top, totally going to get pulled into the summary, especially if it's just a a graphical ad, then it's not at all, right? Because it's only HTML text that's really informing things. Again, even if you were to have alt text on that ad, probably not going to get pulled in except for in some fringe cases. So again, if all the ads are purely graphical, then they're kind of invisible to AI. All right.

We're going to put our prediction hats on here. 5 years out, is email still a direct channel? Are we all just negotiating with AI for attention?

Yeah, so, uh, I think that the future will have us sitting in between these two audiences for sure. Uh, I think that we'll be absolutely still communicating directly to some people, and in other cases we will be communicating directly with machines that are working on behalf of some other people. And I don't know that that's necessarily bad. One of the things that, that we haven't talked about that I think is really interesting is Google's Personal Intelligence. It's this new program that Google has launched that allows you to connect Gemini to your accounts for Google Search, Google Photos, YouTube, and Gmail., and this information gets pulled into essentially like your Gemini profile. And when you ask it questions or ask it for recommendations, it then knows all that stuff about you. And I really think that like, okay, Google Search, that's definitely good, good history information. Google Photos, sort of indirectly, like it can sort of tell like where you've been. Um, and the same thing on, on YouTube, it can tell what kinds of things you're interested in. But I think that The Gmail is like, that's the hidden goldmine in all of this. Just think of all the things that are in your Gmail account. Everything you've ever purchased online is a receipt in your Gmail inbox. I know there is in mine. Everything I've ever bought online and probably a good chunk of things I've purchased offline as well, if I have an e-receipt, all that stuff is in there. So that's like all the brands that you care about the most are in there. Anything that you are getting promotional emails about, right? Which are probably, you know, the 5 to 10 brands you care about most, right? That you're open to hearing about. That's all in there. You know, emails to friends and family, so you can build like a social graph, that's in there. Loyalty program emails are in there. So probably loyalty account balances are in there. Any rewards you might have, like that kind of information is really critical if like I'm trying to decide between like 3 products and I have like loyalty points in a program for one of those brands that helps me get that product for way less, man, if I'm doing e-commerce, I have to know that. Or if I have a promotional email that is just for me, that's like a unique discount because of something I did or what have you. Oh my gosh, if I'm an e-commerce brand, I have to know that in order to be able to truly recommend the best value for the price, you know, in a particular product area. So that is a— that's an example, I think, of a very powerful way in which brands would absolutely or should be absolutely very happy to have their emails ingested by personal intelligence so that Gemini can give their customers good recommendations. Right? Like, why wouldn't you want that? That seems to me like a plus. That's a win. But it's not going to the person, right? The person is probably not reading all those emails, but it's constantly educating Gemini on your latest offers. And again, like loyalty point balances and things like that. That seems really good. So, you know, I think I mentioned it before, like we have to make sure that we focus on what we're really trying to achieve with marketing and advertising. And I think it's, it's influence, right? Like, yes, we'd like to have engagement, but the point of the engagement is to create influence. And so I think we need to kind of rethink a little bit like, what are we trying to accomplish? And it's mostly influence in the end. I think the huge challenge is how do we measure that? And that I think will be a major, major challenge for brands over the next 5 to 10 years is to get a hold of that. You mentioned omnichannel. We don't do a very good job of understanding omnichannel engagement today. We absolutely have to get better at that and we absolutely have to get better at understanding, you know, how AI is in the mix, you know, especially when it comes to things like e-commerce, which is definitely going to grow in the years ahead.

So don't forget about the travel emails, all your flights, hotels, Absolutely.

Yeah. I mean, treasure trove. Treasure trove.

Treasure trove. Well, Chad, a ton of insights there. I think the audience is going to love it. Appreciate it. The conversation is there. Would you like to share where people can find you on LinkedIn or whatever? Anything you want to pitch or book?

Yeah. So certainly everyone can find me on LinkedIn. X and Mastodon @chadswhite. You can also find me at email@marketingrules.com. And of course you can find everything about ZetaGlobal at zetaglobal.com.

Awesome. I missed you at the last conference. Maybe I'll catch you at one coming up here soon.

That'd be great.

But appreciate your time. Thanks for coming on and we'll talk soon.

Thanks for having me.

Thanks, Chad.