[00:00:00] ​ [00:00:00] Phil: What's up [00:01:00] everyone. Today we have the pleasure of sitting down with Jeff Lee, Lifecycle Marketing Tech Lead at Calm. [00:01:06] About Jeff --- [00:01:06] Phil: Jeff started his career as an IT specialist at IBM. He then joined Merchant Cycle as a front end web dev and eventually ended up managing a team of web developers. [00:01:14] He later joined Flipboard, a popular social magazine app, where he spent five years embedded in email dev and marketing ops, he built and grew their email capability to sending. Over 400 million emails per month. And today Jeff is life cycle marketing tech lead at calm, where he architected their adoption of push notifications as a channel and now sends over 300 million push notifications and 2 billion emails per year at calm Jeff. [00:01:39] Thanks so much for your time today. I really pumped a Chapman. [00:01:42] Jeff: Thanks for having me. [00:01:42] ​ [00:01:42] Darrell: [00:02:00] [00:03:00] Awesome. And by the way, I'm a daily comm user. Like legitimately a daily comm user. I think I started, I hope this is not a bad word, but I started with like headspace in, in the past. And then I, I got turned on to calm just because of the ads. And, I found it so [00:04:00] great cause they have a, such a great mixture of content. [00:04:02] there's like sleep, soundscapes, meditations, and different teachers. So, just wanted to say it, I feel like comms changed my life. So it's nice to have you on. [00:04:12] Jeff: Who's your favorite,teacher right [00:04:14] Darrell: I like Jay, Shetty, Chibs. Oh, Karakay, I want to say his name is. Yeah. but one of the delights about calm is like coming to find a new sleep story. That's like kind of right along your alley and they have like sleep stories for almost everything. It's just like boring fairy tales or like tales about being on a train, but nothing happens. So you just fall asleep cause you're just like, nothing's happening. [00:04:37] Phil: As a new dad, Daryl, do you find that like you need. tools to help you fall asleep better, less now, because you have less sleep and now you just pass out easier. [00:04:46] Darrell: I don't know. I feel like you and I have talked about this, but there's like sleep anxiety where you've, you are worried the baby's going to wake up. so you have trouble sleeping. So, so I honestly have started to do this thing [00:05:00] where I watch like maybe some videos that are like interesting. And then I'll switch to a story that's semi interesting. [00:05:08] And then right before I go to sleep, I'll put on something about like history. the Ice Age. And then I'll just Hehehehe [00:05:13] Jeff: Hey try the I think it's the rules of tennis like calm has some random ones It's oh, here's somebody reading the rules and regulations of something that'll put you out like immediately [00:05:23] Darrell: Who would have known there would be a market for [00:05:25] Jeff: Yeah, I [00:05:25] Darrell: someone reading the rules of tennis. [00:05:27] Jeff: It was read by I want to say Jimmy Connors, but I know it's not like that. Oh, McEnroe, John McEnroe. [00:05:33] Darrell: Yeah. [00:05:33] Jeff: the most raging is to like tennis guy, whoever, and he's trying to put you to sleep. It's a complete like [00:05:39] Darrell: Yeah! There's like celebrities that, that, yeah, I think like some A lister celebrities too. That, that, that read the stories. [00:05:46] Phil: that no one is listening to The Humans of Martak as a way to put them to sleep, but,we'll try to make this one really fun. [00:05:53] Jeff: I'll make sure this is the first episode that we can feature on the sleep. [00:05:57] Darrell: yeah, we need a different, like an offshoot called [00:06:00] humans of Martech sleep version. Um, all right. Yeah, let's get into it. [00:06:04] Building Engineering-Marketing Partnerships With a Technical and Emotional Blueprint --- [00:06:04] Darrell: so first question, Martech has been getting seemingly more and more complicated. we have to orchestrate multi channel marketing now with different channels. So I wanted to hear from you, what's been your experience, like How is it working with engineering teams? [00:06:19] You've worked on both sides. Are there any like best practices to working with engineering teams? have you gone through a journey where it was hard to work with them before and now it's much better? Could you riff on that a little bit for us, Jeff? [00:06:30] Jeff: Yeah. so yes, it was in the beginning. after transitioning into the marketing side, not sitting inside of engineering, it was hard to get people's attention, right? Cause they were just like, Oh, we got plenty of other things to do,that don't include you. So I think part of it was just, okay, I'm not going to get support for now. [00:06:50] I'm going to fix this myself. Cause I can, right? but eventually when you start fixing your things yourself, then you can prove hey, now I'm fixing it at [00:07:00] least on a small scale. It's it's not the perfect system. It has its faults. It's going to fail. I need, actually need your help to like make this work every time, like consistently. [00:07:09] And then they get in board and they're like, oh yeah, okay. Now we know you're using the data that you said you needed. We know that it's proving some result, right? for example, we had a field like for like subscription end date, right? It would I need to know that this person's subscription has ended so that we can send them a promotional email. [00:07:26] Or that they've renewed so that we don't send them a promotional email. So we fix that on a small scale and be able to say okay, now we're saying the right thing to most people like 90 percent of the time, but there's still 10 percent of people who are getting the wrong thing. Let's get that down to 1%, right. [00:07:41] And then they get on board with it. And then now we're at a spot where I was like, I proved we, we proved okay, working with the lifecycle team actually has some result that's beneficial to the customer actually reduced effort from. Their team, because they were having to do like random things for me all the time. so now I have a little bit of goodwill so that I can ask for [00:08:00] more things in the future to like, so it's a snowball that accumulates so I can do more with them. [00:08:04] Darrell: Was it like, was it the actual, because you had evidence that something works, or do you think that they also gave you some credibility because you knew what you were talking about from like a technical standpoint? Little bit of both or [00:08:19] Jeff: yeah, probably so there's a technical like ability. So like I would go to them and Oh, okay. I fixed it. this is how I did it. This is the query that you could just drop it into something like whatever tool you're going to use. improve it for your guys's sake and then, productionalize it. [00:08:33] so there's that. being able to speak the same languages, definitely speeds things up. It's less of a burden to them. So, trying to,streamline the process for them and make it as easy as possible helps to get things moving. That doesn't work every time, but,it helps. [00:08:46] Phil: Yeah, definitely. [00:08:47] Why it Took 3 Years to Convince the Product Team at Calm to Implement Push Notifications --- [00:08:47] Phil: Walk us through that process for push notifications at, at calm. Like when we first chatted about getting you on the show, you mentioned that it took a whopping three years to get push notifications on the roadmap at calm, and you [00:09:00] mentioned four years for ML recommendations. [00:09:02] What were your initial approaches that didn't work and how did you like tweak that strategy? Any amusing stories that kind of come to mind about the resistance you encountered along the way? [00:09:13] Jeff: Yeah. So the push one was an interesting one because we needed a screen in the, on the subscription app where somebody could just go in and turn it on or off inside the app. I think that was like a requirement from Apple that was newer, right? That had rolled out that they just required it. I got to think about that. [00:09:28] Like what happened for a second here? [00:09:29] Darrell: I'm just wondering because I've seen this, I've experienced this before too. Like in what world does some, does, do people think that push doesn't work or it's not for a B2C app. like I'm going to try not to like name names here, but I started working for this company and they said Oh, we don't need like an app. [00:09:49] Like it was a consumer company. Like we don't need an app. And like the. because it's so basic, it, you often don't even know how to respond. You're just like, what? I don't know if that was the [00:10:00] experience for you, but like in my mind, it's so difficult for me to grasp how, how business people can say, no, we're not going to do push. [00:10:09] It's not important. [00:10:10] Jeff: I, it could be coming from just pre conceived prejudices, I think, against certain channels. Right. Because when I was, before I started getting into email and push before that, when I was working on the web stuff, I hated email. I thought it was like the scammiest thing, right? Everybody calls email spam, right? [00:10:28] And it's nobody wants to deal with it. But then once I got into it and we started sending a lot and then you can see the business incomes, outcomes, like of doing those things, right. It makes money, it makes engagement, it drives everything. so if somebody like. I think I don't want to put like thoughts into whatever the people who were running the product were before, where it wasn't just something that they just didn't want to do. [00:10:50] They probably had some something that were just, maybe they use too many like slot machine apps and you get the, like the push notification, like every five seconds. So they don't like that. Right. So they're like, Oh, we don't [00:11:00] want calm or, our app to be that [00:11:03] Phil: Yeah. [00:11:08] Jeff: Like at least give us an option to be able to do it right. and then eventually we got the, we had some new leadership come in because, people rolled out and then it was the easiest thing. Cause he, like the new chief product officer had come from a background where they use push successfully at his company and it was like, I sent, I presented them a little thing of Hey, I need these three buttons on this screen. And he said, yeah, okay. Yeah, obviously we need that. and then they built it like six weeks later. It was not like a large lift, it's really just,that was the thing where I was like, realize okay, now we have to actually understand like who you're talking to and the organization, like the product person or the sales person or the engineer person as a salesperson, I'm the lifecycle person selling an idea to the product manager or to the engineer. [00:11:49] That's solving something that's affecting them, right? Or something that they're like predisposed to want to buy, right? So push notifications, like they, they didn't want it. So and I was going up against it and trying to start small and say, [00:12:00] Hey, here's. with the assumption like, okay, if I send 100, 000 pushes, I'm gonna get a 10 percent rate, re engagement rate, which was definitely not the case, right? [00:12:07] It's I don't know. I forgot the numbers, but maybe 10 percent is wrong. maybe like a hundred percent, right? Something like insane, where you can't say no. it'd be so good that they can't ignore you, right? It was not good enough for them to not ignore, right? They were just like, because it was like, it was the, maybe the audience is too small. [00:12:23] The engagement lift was too minor at that level. Right. But then once you compounded over the year, over a course of a year, then, then you're seeing real results, but it's just it was hard to prove as a one off also. Right. I think that was all the problem. It's I think for email and push is something you just have to do it and do it. [00:12:39] And then eventually it's that's it. It turns into this large thing, which is like SEO too. [00:12:43] Phil: when I had to do that at a couple of past companies and they didn't have like push set up as a channel either, the approach that we took was like competitive benchmarking. look at these four competitors that we have, we signed up for All four of them are setting push notifications.[00:13:00] [00:13:00] We are the only one in this bucket of five tools doing the same thing. That doesn't send push notifications. Like that to me, even though we were able to show like industry benchmarks and these are projections about engagement rates, and this is how it's going to ladder up to other KPIs, like just being able to say these other three, four competitors were doing it. [00:13:20] Like the CPO was just like shit. We got to do it. what were some of the initial conversations that you have that you just, you got pushed back on and what did you learn from this? if you were to go to another company or if you were to like turn that off and you had to make the push for getting it to come back again, like, how would you go about this differently? [00:13:36] Jeff: interesting. Yeah. I mean, part of, maybe a part of the problem or part of the friction was that we were the number one for that during that time. So it was like, we're number one. We don't need to do these things that other people are trying to catch up. if we turned it off, I mean, I think if we turned them off, we would see the dip. [00:13:53] Darrell: It'd be like, you would see the engagement rates go down and it would say, hey, we turned this off on this date. [00:14:00] It's just, it's there. like a, it's, I don't know what that is, like a reverse A B test. It's essentially an A B test, right? In order to sell it back, or like maybe we're trying to sell it into another organization or something like that. I don't know. It's it's table stakes at this point. I think that was, just, you just have to at least have the option to do it. It's just, I mean, it's built into everything anyway. It's so easy to like, I mean, I love the fact how, I think you took a very data driven approach, just showing the data of why push is effective as a channel and driving engagement and business results. And then you have to show that to leadership so that they say, okay, that we have to invest in this. [00:14:39] But I think that what's surprising to me and it sounds like to all three of us is the fact that you had to do that in the first place. Like, why did you? Why did you have to, what a waste of time, but anyway, well, what can you do? [00:14:52] Jeff: Well, I think the hard thing is because push is, it's not a great sales channel. Like I don't find that you can send a push and somebody is going to [00:15:00] buy something from you. [00:15:01] Darrell: Right. So it's removed from. And this is a problem that we've had for some other things of it's hard to prove that the engagement lift eventually turned into the monetary lift. Right? And I like, and I'm not, I don't sit on the data science and stuff like that. So I don't know where all those, like the flywheels go. but it's just the fact that it's it's a second, it's like our first or second derivative away from the dollars. That is a little bit harder to prove. [00:15:25] that also seems very, that also seems like a very standard, a default part of B2C app, mobile app marketing, which is doing things to increase engagement to their, to therefore increase revenue in the future. So it's, I find it even strange too, that. and maybe I have it wrong, I haven't done like mobile app marketing is not part of my portfolio. [00:15:51] So, so, but it just seems to me very logical that if you have an app on a very competitive app marketplace, that you would want to [00:16:00] do things to increase engagement. [00:16:01] maybe it's not as black and white as that, and like you really need to go through these business discussions and do the evaluation [00:16:08] Jeff: Yeah, I mean, it could have been, and I'm, again, I'm not the product managers, but like the, it's like the engagement, maybe there is, cause they look at different parts at different quarters and maybe. Like they're looking at onboarding for the first quarter, although, engagement on the first week is definitely, could be driven by push, but they also don't want to turn people off either. [00:16:31] Right. Just by being annoyed. So there's always that balance. I don't know how product managers balance like frequency and stuff, in terms of turning people off or dissuading them from using the app. Cause I, like you said, you get the push notifications from us. a lot of people don't want to receive a push notification one or two times a day. [00:16:46] Yeah, that's why Apple requires that home screen to be able to opt in and out of them. I joined a company where they were doing push already, but it was very poorly done. Like you could tell that the person that was asked to [00:17:00] do it, was just trying to get this off their plate really quickly. [00:17:03] Phil: And it was literally the same message every single day, every morning. Basically prompting the person to log back into the app and do this one action. Right. and I built a team of life cycle managers who were able to break up that onboarding funnel into a couple of different stages and we called it like activation. [00:17:23] And there was a huge drop off from the number of people on that first day within the first couple of hours that didn't do X, Y, and Z. And so the push after they opted in for it. Was a hundred percent dedicated to just doing these two checklist items. And eventually we had a checklist in there after. [00:17:41] And so to your point, I think it is like harder to demonstrate that engagement metrics lead to revenue. I think that a lot of companies have done like the data science work to prove it. And I think that that's part of the selling story of life cycle. Like we improve activation and that helps retention. [00:17:59] And [00:18:00] both of those are levers. to monetization. that's the whole kind of growth flywheel there. if someone's not active and using the product, they're not going to upgrade and add more things to their plan. So I think it plays a big role here, but you talked about like selling this to PMs, like you mentioned PMs a few times in there, I would say a big part of this and bigger companies and [00:18:22] How to Time Martech Project Approvals With Engineering --- [00:18:22] Phil: What's going to be on the roadmap or not for me, sometimes I've had to actually sell some of these projects to engineers and not the PM. So like maybe an engineering leader, right? I'm curious if you've had that at,at Flipboard or even calm, could you walk us through Typical conversations with engineers when you're trying to like get buy in for this new, like marketing campaign or something that you're missing. [00:18:45] And how often does that get you into build versus buy discussions with engineers when it comes to Martech? [00:18:52] Jeff: well, okay. So riffing, continuing to riff on the sales ideas. the average salesperson closes like 1 percent of their deals, [00:19:00] right. Or something like that. Right. It's a lot of reach out outreach. I would say it was the same for me where I was like, just like throwing ideas or, things that I wanted to achieve and maybe I got 10 percent or 1 percent of them, whatever the number is, right. [00:19:13] It's, which is fine, right? [00:19:14] Darrell: I'm the same way. [00:19:15] Jeff: just, you're going to get rejected and just get used to it. Right. [00:19:18] Darrell: I, [00:19:18] Phil: I think. The one thing that we did was that, the ML, being able to pull MLREX for, filling in the, to fill in the, an email or a push, right, and just say, okay, let the machine go grab a thing for this person and present it. [00:19:31] Jeff: that one was something that we had done at Flipboard and, pretty successfully. And, again, I think it was just getting it in front of, A person like, cause we, again, new engine leadership. He came in, and they want to make a splash. Right. because like they want to have a win within their first hundred days as well. [00:19:47] Everybody wants to get look cool. so I, I came with them that way, that idea just was like, Hey, let's, this worked at a different company and we used it. And it's it reduces a lot of, it's something that's evergreen. We use it all constantly [00:20:00] and, let's work together on it. [00:20:01] I'm getting lost in my thought there, but It was like right person, right time, we were doing a little bit more ML like in, in other parts of the app. So they had already the infrastructure was already in place and it was going to be You know, like a sprint of one person's time to get it up and running. [00:20:15] and yeah, it was just like, how do I sell it? well, something we already have to do, they already done, right. You get to expand on something that they've already built. Right. Cause that people always love to reuse something that I've already built or repurpose, right. Cause you're getting more return on the amount of time that they invested in it. [00:20:33] yeah. And then what's figure out, maybe you figure out like, how does what I'm, what I need expand on something that they've already built because then they can iterate on it and it could be smaller because if I just went in and asked for a, like an ML thing from out of the blue and they never done it, there's no way, right. [00:20:49] Cause it's going to, that's like a multi year, probably, I don't know, like a year's long project that. That's,coming out of one, one group is just not, it doesn't serve like the whole company, right? Like it doesn't serve a [00:21:00] bunch of other reasons. I think they're [00:21:00] Darrell: Yeah. the timing thing is so critical and you're so right. I, Because if you introduce, and I've done this both ways, like I've introduced a new idea when the rallying cry for the company was to simplify, simplify. And then you have me presenting to leaders saying, Hey, Let's start a hackathon, and it just fell so flat because it's just it's not the right time. And one of the things that I learned,which is what you said resonates with me is when there is a rallying cry across the company, or like a really big theme, like AI is a really big one. [00:21:31] if you come in with an idea and tie it. With the theme or whatever, like big,big agenda is happening right now. You're much more likely to get buy in. and so, so I can see why, if ML is a big thing for engineering, they want to introduce machine learning across everything. [00:21:48] You can come in and say Hey, here's this marketing initiative that requires machine learning. Let's do it. and naturally everybody's going to be on board. Whereas if you're like going against the grain, which kind of sucks to say, but if you're going against the [00:22:00] grain, it's much more difficult, especially as the company's gets bigger and bigger, it's, it's already hard to get people rowing in the same direction. [00:22:07] So if you position your idea as something that's going to help us go in the direction, same direction faster, it's much more likely to get adopted. So yeah, kudos to you. that's exactly how I would have approached it too. [00:22:20] Jeff: Yeah. But you got, it's like lying in wait, right? You got to be that tiger in the bushes and just okay, I'm on the lookout. Something's going to happen. They're going to do something, let's, attack it when the time is right, which I think I've been doing wrong, like for the push notification stuff, like wrong people, wrong time, right? [00:22:35] They didn't want it. so you're not going to win, right? But eventually the tide turned. [00:22:39] Phil: Yeah, I feel like it depends also on like how the engineering and the data team in this case is set up from like a cycle perspective. Like I've worked at companies where they're like boxed in and there's maybe a PM that's responsible for figuring out what the engineers are going to work on in this like cycle or whatever, and that's like. [00:22:59] They've [00:23:00] already built a roadmap two years ago and they're still following that and you're the marketer coming up with an idea and they're just like, dude, get out of here. we've got two years worth of tickets to go through, like good luck getting in here. But I've worked at companies where there's a shorter cycle and there's like a planning cycle to this. [00:23:16] And the roadmap is a bit more in flux and it's dynamic. And I would have a counterpart on the data and the engineering team, and we would have quarterly. Sessions where those people would meet with different stakeholders in the company. I was one of their stakeholders on the MarTech team, and they would look at requests that I would have. [00:23:36] So they'd look at like projects that are on my wishlist and then we prioritize them. I like explain the use case, the potential impact, they put this in like a rice model, whatever they used. And then they go off and in a week or two, they tell me like, We're not going to do all of these except for that one. [00:23:53] Or like this one, we're maybe going to do, or we're going to push everything back, but at least I felt like I got a say in [00:24:00] trying to, and it wasn't like, like I was less of a tiger in the bush waiting for the right time to push this. And it was more of like, all right, I have this quarterly opportunity to make a push and make a pitch for this one idea or this other idea. [00:24:13] And maybe I come to that pitch with support from. This other senior leader who also wants this, who is sharing that same KPI as me. So I feel like it depends a bit on the maturity and the sophistication of the data and the engineering, like cycle process. [00:24:27] Jeff: Yeah, I, that definitely, that sounds, I agree with that works,but you're still like, you would have had your laundry list of things. Right. And it lined up. So I think we're talking about the same thing, right? Like you were, you had written it out. It's not like you wrote like an entirely new set of things after they rejected the other 10, right? [00:24:45] Or the other nine, right? You kept it and you waited. And it was there, right? I think that's just important, have the ideas, have them written down, and be prepared to,to discuss them and back it up with something that,is gonna get, pitted against, probably 50 other [00:25:00] things, in that quarter. [00:25:01] Darrell: Yep. [00:25:02] How Three People Run Marketing Lifecycle and Automation for Calm’s Billion Message Machine --- [00:25:02] Darrell: Good point. [00:25:03] ​ [00:25:03] [00:26:00] [00:27:00] [00:27:17] Darrell: so next question, why don't we talk about like team structure a little bit? what's the division of skills? Like what kind of skills that you think that you need on the team? To succeed in doing, messaging at this scale. gosh, I saw it, it's like in the hundreds of millions, right. [00:27:35] And in terms of like email and push. So that's so crazy. so, so what, how did the evolution of your team go? Did it start with just you and you hired some more people? How'd you, how did you choose around, around who you would bring onto the team? And what does that look like? [00:27:47] Jeff: Yeah, so before we had a Lifecycle, head of Lifecycle that started before me and her, she was a, like a big time, like email advocate and she brought me [00:28:00] in because she has done Lifecycle a few times and had the problems of can't get engineering resources, can't get like X, Y, and Z to patch things up. [00:28:08] So she was very specific in trying to find an operational person that could patch all the holes. like a duct tape and bandages guy. Like that was basically where I thrive. That's what I love to do. so then when it was me and her, and then we brought on an HTML, like a designer guy, I think as like a full time or he was as a contractor before. [00:28:30] So we had three people for probably two or three years and we were jamming out. [00:28:33] Darrell: And you're using a home homegrown tools or like internal? [00:28:37] Jeff: we have a, platform we use,CDP or not a CDP, whatever, email ESP or whatever, like a customer IO or, one of those. [00:28:45] Darrell: Gotcha. Gotcha. [00:28:46] Jeff: and we were building stuff, one off campaigns constantly, and eventually, we were You know, when you get to a thing where like every email that we would send, we send a weekly email on Sundays and it's like this digest of, come, here's what you want to come in and see, [00:29:00] it would take six or eight hours per week to get this thing together. [00:29:04] Darrell: So, cause we had to get like pull in content from other people, figure out what's the feature. So we had to look at, where's all the friction coming from and productionalize it. So like he developed a, my, my design teammate, he designs and codes. So he developed a, like a Lego pieces, like a design system where we prebuilt all the HTML, and then I templated it out. [00:29:26] Jeff: And now we're getting, we're producing emails in, probably an hour or something like that, maybe half an hour, depending on whatever, right? and we, it, we had to line up, the content people into, the cadence of, okay, you gotta get us this information and, or even just batch it up, a month in advance. [00:29:40] So, this kind of stuff, it, it took a long time to get, Get to that spot. But we're there, finally. And then like over the time we added a couple more Lifecycle people, a couple more,some people come and they go. But for the, the bedrock is,a lifecycle person, and then, the designer developer, and then I'm doing the [00:30:00] tooling to ease the tension between or just give him the capacity to handle three, three or four marketers to like throwing stuff at him all the time. [00:30:09] And, if you get to a sit, if you engineer the, a thing properly, I, we could probably, we've probably scale our different types of campaigns that we can do, probably two X last year. For like year over year, [00:30:21] Phil: are you like a bit like a service org also, are you fielding requests for campaigns from a bunch of other teams on top of like the things that you guys have on your roadmap also? [00:30:33] Jeff: Yeah, we do. but the, we do, yeah, we're like a service to worry, but we tried to not be, as much as possible. so there's often pushback, from the team, but I leave that to the life cycle managers. They're good at that. I'm not,I'm the type of person who just like, all right, fine. [00:30:48] I'll just do it. And,at some point, like somebody has to do, those last second asks,and it's whatever. It's just make it as painless as possible. [00:30:56] Phil: It's cool that like you were brought in to do some of the [00:31:00] patchwork and not have to have the friction with using internal resources that are external to the lifecycle team. Like I like the angle that. That you were hired into. we talk a lot about this on the show, this idea that like, how do we get engineers excited about marketing is oftentimes easier said than done. [00:31:19] But, we've chatted with another one, this year also, but you're also in the camp of essentially a web developer who went into the email route and you touched on this a little bit, but [00:31:29] Why Email and Push Can Be Standalone Products --- [00:31:29] Phil: Every engineering team that I've worked with and a lot of those engineers, like initially they saw Martech and marketing technology is just like sending emails, sales on mass, like a lot of like spam, like one bad Apple creates a lot of that negative perception about marketing. [00:31:45] I'm curious to ask you if you could share like a specific project, or,something that happened where you're able to transform that perception for others, or when that happened for you, when you made that transition, like, how do you reframe life cycle [00:32:00] marketing to get developers excited about some of the technical challenges that lead to growth, but that are technically marketing and not product. [00:32:07] Jeff: Yeah. Well, I mean, it's technically marketing, but I view, like the emails and the push as product or platform in itself, right there, we send as much, we'd probably drive as much traffic as. as the website does. Probably more, actually, because it's 100 million people or whatever. even Flipboard, 300 million, billion messages, right? [00:32:28] There's like real, that's a product. That's not just marketing, right? You got to treat it with some respect because it's driving a lot of things. and it takes,in order to get it to like an evergreen state where we were just, sending like these digest emails, that's also a product. [00:32:43] It's a webpage, right? every email Is a collection of data coming out of our backend that has to be formatted and turn into something. Right? So it's not just, it is like it is engineering in a lot of regards. but what was I trying to say there? It's, That, I think that was the part where I was working on [00:33:00] webpages and you had to wait for people to come and visit them and they don't always come, right? [00:33:04] Like even if, like you could have an SEO or something, like even if you're driving whatever, a million visits a day, if your webpage is in the reg gate, people aren't gonna see it, right? Like you got to get that thing in front of their eyes, right? So, Like from an email,we could prove out some things too. [00:33:20] Like you can use, probably faster to produce like that, for example, the ML thing,Hey, is it picking the right thing for this person? Well, you blast an email into their face or, like into their screen or a push notification that's like presenting the piece of content, like a news article. And if they open it, that's a way faster signal than, waiting for them to open the app and hoping that they'll see it, right? It's just you can prove things faster. And I think that's where, the,where I made the transition. It's cause I, it's hey, I can get this thing in front of this person. [00:33:52] And if I made a mistake, It was an email, so that, most people, it disappears, it's it'll just get buried by a thousand other things by tomorrow. [00:34:00] So I can try again. Or you get a new chance, every couple days, probably, to,make an impression. or just try to make your, take your shot at whatever you're trying to experiment with. [00:34:08] It's just faster. It's a very fast cycle. [00:34:10] Phil: That is a cool way to think about it. Like email and push, even though they are technically like channels leading people to the product, they could be treated as a product in and of itself. And I feel like that's some of the pushback you get from engineers. They're just like, I don't get excited about getting up every day and like logging into an email tool and like blasting emails to a bunch of people. [00:34:32] Like I'd rather. work on this feature that's going to be seen by a hundred million people in the app itself. But to your point, it's yeah, but those people just don't like all wake up one day and want to log into the app. That's where email, that's where push comes in. And they're almost like a product themselves because they are standalone. [00:34:51] Like that push notification in the phone. I Like there's a lot of cool stuff that you can do to trigger that and get people to, to log in and want to do [00:35:00] something, let alone email. Like you said, like an email is essentially a like static version of a webpage and sometimes you can make it dynamic also, but yeah, I really like your perspective there. [00:35:11] Jeff: it does. I mean at some point like it has once you have that we have those evergreen like weekly ones they are a product that people miss they notice when you didn't send it to them on Sunday at 5 p. m We'll get complaints like hey, I didn't get my email on like where is it? [00:35:25] Send me another one like set make sure you get it to me. Alright, so it's like that's a I don't think that everybody has that problem. but it's a good problem. It makes me feel good because it's oh, people are noticing that they didn't see it. also a, yeah, it's just, it's a nice feeling to be like, they missed it. [00:35:39] Phil: It's a cool spot to be in. I think like most people that send emails, the purpose is to drive people to a landing page or their website or log into the product, but it's like what you just said is if your email is good enough, it can be a standalone product like those like Sunday digest emails. And,I'm a calm user, but I was also [00:36:00] a huge Flipboard fan, back in the day. [00:36:02] I haven't used it in a while, but, I loved it. And I, instead of going on Google or seven different news sites, I would just open my Flipboard app. And instead of opening the app all the time, I would just start reading the digest that I got from the emails themselves. So it's almost like email was the Flipboard product because I would digest the stories through my inbox as opposed to always logging in there. [00:36:24] So I think that's like a bit of a privilege thing though, because like Daryl and I. We both spent a lot of time in the B2B world and B2B, like we don't have always apps for the end users to use and email is usually just a tool to, get people at the next stage of their sales journey or go on a website and fill out a form. [00:36:45] I think in the B2C world, there's a lot more fun and flexibility with email for sure. [00:36:50] Jeff: it's hard. I've been trying to replicate the experience for another,and it's, it doesn't, not everything lends it to itself, to like an automated well, yeah, like weekly [00:37:00] thing. [00:37:00] Building Technical Excellence in Marketing Operations Leadership --- [00:37:00] Darrell: So I have like a, this is a kind of a random question, but I feel like you have a really cool job. And it's like doing marketing operations for B2B, for B2C at a really big scale and, just a cool company. Like for those people that like want to follow in your footsteps and want a role like yours, do you, should you be super technical? [00:37:21] Like you, do you consider yourself very technical above average? Like how does some, what skills do you think you need? What skills have you found super important to be successful? At your job. [00:37:32] Jeff: I am. Yeah, I would say I'm definitely super technical. I think it's like what Phil was alluding to is like, how does an engineer get interested in Martech from that lens? Right? I like it because we rely on these platforms that have to do all of the pager duties. Answering all the fires, right? [00:37:50] I'm not responsible for the 99 percent uptime of the email service. Somebody else's, and we're paying them a lot of money to be the person, right? So I get to do all the [00:38:00] fun stuff, right? I'm making these emails. I'm doing the experiments, all of the, analytics, blah, blah, blah, blah, blah, like all ends up in some other platform where we're looking at the results and stuff like that. [00:38:10] So I've managed to strip away all of that's like the drudgery of unit tests, scaling,penetration testing, whatever, it's very fast paced. it's very creative. And it's,it, for us, it has a direct impact on the bottom line immediately. Like you send an email and like thousands of dollars come back. [00:38:28] It's it's fun to see. [00:38:29] Machine Learning Marketing Projects Can Actually Cost More Than They Make --- [00:38:29] Phil: You mentioned, the ML recommendations project, a few times in our chat already. And I want to dig into that. one of the things I enjoy the most about working at bigger companies, I've only had a few, stints, but it's This idea of like more resources, specifically data engineering, data science, and even though you're not always getting your stuff on the roadmap, like it is cool to have access to that in your typical startup. [00:38:52] Like you're lucky if there is one data engineer on the team, let alone like finding people to help you build like custom ML models to [00:39:00] do the type of stuff that, that you're describing there. but still though, for many folks at those big companies, and I'm curious to hear if that's the case for you at Como. [00:39:07] Also, like those. ML resources are not always shared and not always dedicated to marketing or lifecycle. And sometimes you just have to figure out a way to do it without them. What are your thoughts on this? And like, how do you go to market when you don't always have access to those resources? [00:39:26] Jeff: Well, I think the ML, the effectiveness of the ML itself from the regard of picking the recommendation wasn't like the mind blowing,result that you might hope. Right. In terms of like the amount, like the cost per user to generate those recs. And then the return of the engagement that we got probably didn't pan out. [00:39:47] It was like an interesting, it was an interesting, like a nice technical achievement that we were able to pull off. But like I was talking about before, like in turning of getting that engagement and then turning into dollars down the line, I [00:40:00] don't think it paid off. so we've maybe, moved away from that a little bit, but we can still reasonably personalize according to at least some signal that we're getting from, your prior usage, which is what the ML was looking at anyway, right? [00:40:13] So it was basically like, Hey, Daryl happens to like soundscapes and, he uses them a lot and Phil likes sleep stories, so he uses them a lot. So You we could approximate it and it's not going to be like the ultimate thing, but we can say well, this person uses soundscapes more than the other things we'll send them. [00:40:30] Like the same thing of what they already used. It's just cheaper and faster that way. so there's some yeah, there was like some cost benefit. Like we had to figure out in terms of, whether it worked aside from the technical achievement. But then in terms of getting the resourcing, that one was like, they had already built it, right? So I was just trying to get like an into it. I don't know that there was anything, any special magic that like we pulled off. if you want, at the end of the day, it's ML is a personalization, at least from this regard. [00:40:56] So it's just if you can personalize it at a broad level. that's better than [00:41:00] nothing. And then the more you can prove that the personalization is impacting things, then, the more likelihood that somebody is going to start paying attention to what you're working on, I think. [00:41:09] Phil: Very cool. So takeaway is almost like if you don't have access to a fancy propensity model to tell you the next best action that a user should be taking, sometimes there's a lot of value in just looking at that first party data in your product. Analytics tool that is helping you figure out the direction of potentially the next best action for some of these things, but just looking at what Daryl has interacted with recently, or, stuff like that. [00:41:36] Jeff: Yeah, absolutely. Yeah. It's just showing them something that they've already seen before. It seems to help, like it seems to work. [00:41:42] Darrell: There, there's, there's this, [00:41:44] ML Algorithms Make Bad Marketing Decisions Without Human Input --- [00:41:44] Darrell: I don't want to say it's a rumor, but it's like a trend at a lot of forward thinking companies. And that is that market marketers are no longer create, no longer selecting audiences for their [00:42:00] offers. So meaning that, Hey, I'm a marketer. I'm going to come up with an offer or message. [00:42:05] The system and the data and the algorithms will choose my audience for me. and I, and some forward thinking companies are like moving toward that where historically marketers are like, Oh, I want to send to these people. But now, especially at more technical companies, they're, we're taking that out of their hands. [00:42:22] Does that resonate with you at all? Is that, do you feel like that's a good direction? or do you still feel like marketers should pick their audiences? [00:42:29] Jeff: I did talk to a vendor about that and it was interesting. It was like a, definitely,A mind shift that maybe we're not at least I'm not ready for Like I couldn't understand like I got the idea. I think there's there's some It's nice to be able to know Why something's like working for those people and it's like more explainable to my like leadership, for example, right? I think because we had evaluated something like that and it was we just it wasn't like we couldn't [00:43:00] Develop a story except for to say well the ML picked something and it worked like you don't always know what's going on Like you don't, not even always, like we never, it's like a black box. [00:43:09] It's hard to like, when you're just sending something and it seems to be working, but it might be working for the wrong reasons, right? It's I think I had mentioned in our little pre interview or something like, like I accidentally sent 19 million emails that were basically blank. And we got a massive open rate, right? [00:43:25] Darrell: So if the ML, or the AI is going to be like, Oh, we should send more blank emails. I don't think that's the answer, right? That's genius. [00:43:32] Jeff: so you have to know, there's something there to like, to have a, [00:43:35] Darrell: so [00:43:36] Jeff: yeah, like it, it could be optimized for the wrong things. And if you're not looking at it and you don't know what's going on and you just assume that it's doing the right thing, then you probably, you might end up screwed. [00:43:44] that's, I think there's some value in like knowing who you're trying to talk to because that was another thing was like like we were talking about the podcast I listened to like I listened to a lot of magnetic marketing and just like marketing journals like like the machine might know what they want but I think [00:44:00] there's like a it's like you just got to know like it has something has to know what the user wants and if you don't know if it's a black box and you don't know what the users are responding to like on your own like how are you going to write anything for them Like, or to like to build this copy or to get to have the, make this connection with the end, like the buyer or the end user or whatever. [00:44:19] Right. I don't know that the, that we can trust the, those like the automatic stuff to be saying the right thing that's like brand safe and or even brand aligned, right. To to always do the right thing or at least not where we're at right now. [00:44:33] Darrell: No, that's such a real example. I never thought about it that way, but I just. Your example of 19 million blank emails and AI is optimizing now for blank emails. I think that just, it just really just goes to show that, yeah, there's always going to be a place for human, the humans of our tech. Gotcha. [00:44:50] there's always going to be a place for the humans of Martech. One other thing too, I will say is maybe you can tell Phil because Phil is of the belief that [00:45:00] engineers and data people. don't really like to be in marketing. They think it's like a tour of duty. So you're never really going to get great technical people working in marketing. [00:45:10] So can you tell Phil Jeff that he's wrong and that you really liked being in marketing? [00:45:15] Jeff: Well, absolutely. [00:45:15] Darrell: just kidding. I don't want to put words in your mouth, but if that's how you feel, tell us. [00:45:19] Jeff: Yeah, absolutely. Enjoy it. I think, I like, I don't know. It's like some people, it depends, right? Some people like doing like certain types of things, but if you're like an engineer who likes. If there are, engineers who like speed, like of execution, high iteration, impact to the bottom line, or measurable like impacts, and don't like, you're not getting paid just to change the color of a button and wait six weeks for the, for it to go stat sick, that red was better than blue, then I think, You might be interested in marketing, right? [00:45:48] there's, yeah, that's the pitch more engineers in marketing. I think that's the way to go. [00:45:52] Phil: I like it. [00:45:53] How Engineers Actually Want Marketing Teams To Pitch Project Ideas --- [00:45:53] Phil: What do you think is the most common lie that marketing teams tell themselves about why [00:46:00] engineers won't support their initiatives? Like how can marketing teams better understand and respect engineering constraints and priorities? when it comes to pitching ideas to engineers. [00:46:10] say is like that they don't have time. Engineers like engineers have some time, right? They're there's, I'm not saying, maybe not at a meta or something like that, but like at startups. I had some time. I made time, like we're I think that's the beauty of we're like, we're creative people, right? [00:46:26] Jeff: so we have ideas. We try to, people build their own apps like at home and stuff like that. So if you can build a, build something internally, that's fun too. Yeah, so that's probably the lie in terms of like, how do you get them to come over to the other side? I, just thought what was I like, come up with if you have something, try to build it a little bit or, or start like prototyping it a little bit yourself or develop like a mock up or something at least that you can communicate it because they might get interested in your project. [00:46:55] They might like, you're like, oh, actually I am already working on something that's very similar [00:47:00] to that and I could just throw in five extra lines of code and you'll be supported. You don't know, sometimes, you'd be surprised that in some cases, something that you think is really huge might only take 10 minutes to complete. [00:47:12] And something that you think is really tiny might take 10 months, right? so you just have to like, have it there and be able to like, discuss it right. A little bit. [00:47:22] Darrell: Yep. Yep. No, and I,I totally believe that. and, actually my brother is a product manager and even they have to sell to engineers. And he tells me that the way he does it is he literally puts on, he built an entire PowerPoint show about like how, why this feature is going to be exciting and turns it into This is the future guys. [00:47:40] this is the future. And,you're going to help us get there and that's, and you have to do it. You know what I mean? like most of work and life is just sales and influence. So anyway,we're running out of time. So I wanted to talk to you, [00:47:51] Finding Your Marketing Career Ikigai Through Technical Flow State --- [00:47:51] Darrell: Jeff, you're a technical leader, email pro, but you're also a dad. [00:47:55] Golfer, Home Chef, Light Programmer. One question we ask [00:48:00] everyone on the show is, How do you remain happy and successful in your career? How do you find balance between all the things you're working on while still staying happy? [00:48:07] Jeff: So, like I mentioned before, it was like, I like, I'm a tinkerer. I like to do things, like to build things. And there's this term of Ikigai. It's find out something that you're passionate about and then you'll also get paid for. I think this, like this combination of marketing and engineering of the speed, that's it. [00:48:23] I like doing this. So it's like the balance is there. I'm not saying and I tend to work on the things that interest me or that, well, I'll get into a flow state and I'll just realize Oh, it's five o'clock, six o'clock. It's Whoa, what happened to today? Those are the things I like to do. [00:48:36] It's not good career advice, right? I'm not saying if you're trying to be a VP or a director or something, X, Y, or Z. Chasing Ikigai stuff isn't going to get you there. Maybe, probably not. but at the, at that point it's just I enjoy working. I like, I do this once the kids go to bed, I might continue to do more stuff like occasionally for myself. [00:48:56] Like just because it's entertaining. And, it's just I try to align my [00:49:00] career with my interests or my like things that I enjoy doing. And here I am. [00:49:04] Phil: Super cool answer, Jeff, really appreciate your time, sharing those war stories, behind, two of those awesome, really cool companies, clearly they're lucky to have you, I think any life cycle team out there needs a Jeff on their team and, yeah, really appreciate you, you sharing those stories with us, man. [00:49:18] Thanks so much for your time. [00:49:19] Jeff: All right. Thanks for having me on. It's been a lot of fun. [00:49:22] Darrell: Thanks, Jeff.