A Conversation With Sean Duffy of Reignite

Over 20 years ago, email standards added an HTML 'body' to the plain-text 'body' of messages. So for over 20 years, email has been technically equipped to fetch images on-the-fly from remote Web servers. That little technical accident is the foundation for a niche part of the email industry. The most frequent label for that niche is "real-time content". But as Sean Duffy — founder of Reignite, a company in this niche — and host Matthew Dunn — founder of Campaign Genius, also a company in this niche — discuss at great length...there's more to it than that!

Modern email marketing is a complex game; the marketing messages in your inbox are likely to have required multiple people multiple days of work before hitting Send. That's true even when — like most emails — thousands of people get essentially-identical content.

To make the content more personal, more relevant, more interesting takes technologies like Reignite. By using data and rules to produce content — like personalized images — these platforms give email marketers a route to better engagement with less work.

As an experienced email consultant, Sean is quite objective about the challenges of real-time content. For better or worse, it's not a mainstream practice. Is that because email marketers are stuck in old habits? Because the technology is too complex? Because it's still too technical? This discussion gets "into the weeds" on all of those issues, and more.

Personalization gets quite a bit of airtime (or a thrashing). As common as the term is, it's still squishy and undefined, and email industry practice shows it.

If you've thought that email could use an update...this conversation is for you!

TRANSCRIPT

[00:00:09] Matthew Dunn: There good morning and good afternoon from my guests. This is Dr. Matthew Dunn hosted the future of email and my guest today, Sean Duffy, my, my friend and colleague and fellow soldier in the trench wars of real-time content. Sean, you the founder of re-ignite and also an email consultants, correct?

[00:00:32] Sean Duffy: Yeah.

[00:00:32] Yeah. So the reignite came out of the consultancy really is neat to be able to get stuff into an email without having to work with every single ESP different API APIs or limitations. So that, that sound we've come about really?

[00:00:47] Matthew Dunn: So, so for the 90, probably 98% of guests who got a real time, what the heck are they talking about?

[00:00:55] W we'll do a brief orientation here. What Sean's company and Mike company campaign genius, both do in varying ways is is try to make the current. That shows up in your inbox get there more easily more recently in a more timely fashion or even God forbid to have it reflect conditions that have changed since the thing was sent fundamentally the, the fetch of image content from email is kind of the backbone of this, right?

[00:01:27] Sean Duffy: Yeah. Yeah. And, and we talk about it as being the real time email industry, but I think we've talked about this before. That's a bad batch for what we do, because the real-time element is a teeny tiny part of what we do. There's so many other reasons why, and I think that that causes us headaches when people start talking about MPP and all these other things that we're going to be massively impacted.

[00:01:51] Yes. Because as an industry, we haven't been able to probably market ourselves properly experienced batch to that, that room to.

[00:01:59] Matthew Dunn: Terminology. Yeah. Yeah. I mean, this, this is going to be the inside baseball episode of future of email, but darn it it's important. Now, let me, let me frame this again, for those of you listening are going, what the heck are they talking about?

[00:02:13] Shaun's Shaun's based in the UK Leicester, I believe correct? Yep. Yep. And I'm I'm way across the pond and across the country in the Pacific Northwest of the U S campaign genius customer base, mostly you S you, you could nominally say, well, Sean and Matthew were competitors, but Hey, that's not how email works.

[00:02:33] Be it, the real-time piece of it is such a relatively small industry that I think it's much more important for all of us to get better at our game and build awareness of it then than to worry about whether or not we're going to run into each other in the market, because we never had.

[00:02:49] Sean Duffy: No, no. We're primary UK based.

[00:02:52] You're primarily US-based, I'm sure we both take revenue where it comes

[00:02:55] Matthew Dunn: to. Yeah. Well, Nick nature, nature of a startup for sure. Do a, do a bit of history, cause we've never talked about this. You said it came, you said rig night came out of the consultancy, but how long ago? And do you remember the glimmer of, wait a minute?

[00:03:10] I think we can do this. That got,

[00:03:13] Sean Duffy: yeah. Blimey. So my background is 20 years. This year I've been in the email marketing industry, which yeah. Well, is that congratulations? I'm not sure sometimes, but I sort of started off a small UK ESP where we, we were, there was four of us at the time and we launched an ASP and got quite successful.

[00:03:34] And I was there for sort of 10 years when clients side got really frustrated, having been in an ESP. And effectively we could do anything we needed to do from a client. If we couldn't do it, we're going to build it to then, okay. I need to do all these things mainly around personalization for footwear retailer, and a lot of data analysis or transformation things.

[00:03:56] And that just wasn't the solutions out there. And we didn't want to go work with a big enterprise kind of customer. So that's where sort of sacramentum is the name of the consultancy where that sort of sprang from, which was all around. How can we bridge that gap between you've got an ASP which maybe have lots of bells and whistles and you have a set of strategies and plans, but how can you sort of breach any gaps?

[00:04:17] There are really, and getting very sort of hands-on and so a lot of what we did on that. Was around taking customers' data, transforming it to something useful for them to do segmentation and lots of sort of personalization around building bespoke, recommendation models and things. And then we come to, okay, let's send that email and how do we get all this individualized content for each individual recipient into MailChimp into bigger ESP and how do we do that in a manner, which we don't have to reinvent the wheel each time.

[00:04:54] And clearly we've, you know, we've been around the theme of space for quite a while. So we were able to see what other people were doing and, you know, so it wasn't our idea in the first place. So we sort of stumbled upon, well, this is what the likes of say move blink are doing, which of course are the very first in our industry to really do this.

[00:05:12] And our initial kind of play was we actually run them in real time. We pre generate all the images that we haven't needed. So we would take a product catalog of 40, 50,000 products and create all the different images that we'd need for them. That, that was our first, first stab in the dark. But then of course you evolve over time and probably a few years into the sort of the business.

[00:05:34] We, we we'd got it to the point where we could, we could serve it live and it's much more flexible. And as technology over time has been, become more accessible. I think it's fair to say, it's the, the reason why you and I are able to operate in the market nowadays is because the likes of Amazon web services, Google cloud, they make it so much easier for startups to come in, come into the market.

[00:05:55] So we did that and that, and that was primarily really to serve personalized product content into emails, but then of course you see, okay, what if we can do that? And we can also do some of the other real time things, whether it be store locators might be weather forecasts. It might be making decisions based on where they are right now personalized images or all of those things we sort of started starting to sort of play with.

[00:06:18] But very much we are our part, our sort of, I guess, motivation is all around individualized content for recipients. Yeah. The real classic real-time stuff is very much kind of secondary and it's because, well, we could do so let's, let's also do that at the same time. It's not our, it's not our passion project.

[00:06:39] I would say some of those things,

[00:06:42] Matthew Dunn: well, I'd say one way, one way of explaining it to people. It's like, there's a, there's a box in your email, maybe rectangle shape. It may be square, but there's there, there's a box that's that that's going to get filled in with with the graphic. We can fill it in at the last minute with darn.

[00:06:59] Anything, and it tends to be the biggest design element, single design element in the email, and the fact that so frequently it's filled with off the shelf last minute, stock photo pick is such a wasted opportunity in terms of communication and brand and motivation and moving people to act. And all the other things you're trying to do with marketing content.

[00:07:23] Like why, why is, why is email so visually down and boring? Can we change this please?

[00:07:30] Sean Duffy: Yeah. And it's, it's when you start telling people that they can use their own funds it, or at stylings that there could be that the visual element is really important. I think that's that's what real time doesn't necessarily describe so much.

[00:07:42] Matthew Dunn: Yeah. Yeah. Agreed. Agreed. And Sean's also, you know, you're also alluding to something both both of us have had to grapple with in last gosh, it's coming up on a year, you know, apple, apple, Announced in June a year ago, 20, 20, 21, that they would start handling images in email differently. They would prefetch and not reload those.

[00:08:05] And that's had an effect on many other aspects of email marketing, but for sure, for the niche of realtime content, I think see, see if this resonates with you. I, I wouldn't when people who I know are really good in this space go, oh, well, I guess time content is dead. I want to smack them upside the head.

[00:08:25] Yeah. I mean,

[00:08:26] Sean Duffy: w we're both members of only influences we joined the Thursday calls and there's been a few of those times where I've kind of, there's, there's a few of us in the real time space on those calls. And we can kind of look at each other and sort of, you know, roll, roll the eyes and things. It's a.

[00:08:42] It's coming from a good place. I think that that sort of comment, but it's perhaps not so educated. Yeah. And I mean, there's a, there's a few things about that. Well, first of all, as we've talked about, real-time, isn't a main thing anymore. It really isn't, but it's the batch we've given ourselves. The other thing is the way apple does the prefetching, what we should really call it.

[00:09:03] Prefetching either. It does some fetching that could be for they've been emailed, but most likely it's afterwards. And again, our friends that moved blink, who's got by far more data than you and I can sort of throw together. They also came out with the, the research piece that showed that virtually all of the click-throughs that they track at the same time occurred very shortly after the open, which suggests that, you know, the opens that Apple's prefetching are late in the day tends to be.

[00:09:38] Sort of nine o'clock, 8:00 PM on woods. You ha you get this secondary spike. Yep. On that people are trying to come up with an email anyway, so this is probably, I don't know, I sort of finger in the air about 85% accuracy. If you could create an algorithm to sort of work out what's a real open or not on apple.

[00:09:58] So, so from that point of view, you know, it's still fairly, it's mainly real time anyway. Yeah.

[00:10:05] Matthew Dunn: Yeah. Yeah. And, and as, as more of a broader statement that you made that, that the real-time or last minute changes, the content is not at all the only game in town when you've gained the capability to manipulate that, that significant piece of email content.

[00:10:25] And yet it's the moniker, it's the sector label we've hung on ourselves.

[00:10:30] Sean Duffy: Yeah, because that was the unique thing. Wasn't it? When various people started coming out with these technologies it was the thing which grabbed people's attention. You know, count downtime as the ability to, you know create these maps or weather forecasts based on which used to be based on IP address, which no longer really possible either.

[00:10:48] And yeah, everybody's sat up and paid attention. But when we drill down to where the real value is, it's things like. It's around the automation piece. So a client doesn't have to manually sit there and code that content. It's awesome to produce for them, which then means you can have infinite versions of that content which means that you save so much time.

[00:11:09] And we have a, we have a client where they send a, an email every single day with our TV listings in, they don't touch it. It's awesome. Actually could use for them. And it does have some real time elements in there as well, in terms of, you know, working out the hour of the day, they're in currently to highlight things.

[00:11:25] But, but you know, that 7 3, 3 graphic designer days a week. Wow. So it's, it's significant that that's that's one area, but then the real value is how do I get some of these complex personalization projects done when I've got an ESP, which doesn't really have user-friendly personalization tools. I would need to sort of have a coder involved, you know, from the it team.

[00:11:51] Kind of isn't really available for most people because they're all going to come back and say, yeah, yeah, you can have that resource for a week in three months time. So it's also about giving marketers, I think, control over that, that personalization and breaking it down into a project they can manage and not have to go for the extra resource.

[00:12:12] Yeah. Yeah.

[00:12:13] Matthew Dunn: Which is an increasingly expensive resource. You know, this, this just just occurred to me, perhaps it's a fair, fair comparison. Web websites because they're delivered when a page is opened and because there's a, there's a legit language that runs in the browser websites in theory would be a lot easier to personalize than email.

[00:12:40] And there are many, many web delivered vehicles. LinkedIn, Facebook are obviously examples that are personalized to the nth degree, but interestingly, I guess most websites aren't personalized in to any particular level. Would you agree?

[00:13:01] Sean Duffy: Yeah, I would. Yes. Yeah. So I, I would agree to it nine times

[00:13:07] Matthew Dunn: out of 10 where I'm running, where I'm running with that thought that just occurred to me.

[00:13:11] Cause you know, same, same is true of email, personalization and email. Let's leave. Dear Sean, dear Matthew, out of the mix. That's not really personalization anymore. Most email messages, personalization, light, or not at all. Because, and, and back to your thought about the developer, because it's it's actually a very complicated thing to pull off website or email.

[00:13:35] Impactful personalization requires such a degree of control of data and, and such a degree of complexity in the delivery platform. That at the moment, at least it's still kind of a developer level job.

[00:13:53] Sean Duffy: Yes. But also I think because there isn't a huge history of marketers managing that they don't have the experience.

[00:13:59] And if you haven't done something before, it's really difficult. So I do find it difficult because I've done it a lot. It's like, I've learned that skill now have a lot of experience. So, so I think even the understanding of the hoops, you have to jump through it. It's a big switch off because you know, where do you kind of start?

[00:14:20] So I think that's an issue, I think certainly, and again, in the email space the existing tools. The market has relied on that. Haven't really tackled personalization in it and use a friendly way. It's, it's, it's a tick of the box that yes, you can connect to a feed or you can create dynamic content rules, but you know, you might even have a user friendly interface with that, but that's not scalable.

[00:14:44] Some of them will have glorious AI solutions, which are usually fairly generic and you can't attain to uni.

[00:14:53] Matthew Dunn: And again, you have, they have, if the data house isn't in order. Yeah. It's hard. Right. And, and by saying data house in order, we're saying like that's a really, really big statement which data about who, oh, well, we've got their email address and maybe their first name and maybe that's accurate.

[00:15:13] Like, what else do you have a little bit, but not a lot

[00:15:18] Sean Duffy: because ASP, so really have a lot of places most of the time, just to put. And if they do have somewhere then to marry it up with rules yes. Content X to customer Y. Yeah. And what are the rules? It's quite quite limited. So yeah, it's, it's it's a hard job.

[00:15:36] It is a hard job. But that's, that's, that's why our, again, that we call it a real-time email solutions. It's one of the reasons they exist, they solve a lot of those complex custody issues. You know, it's challenging for us sometimes because that customer often has expectation that the ESP should be able to do with everything.

[00:15:54] Yeah. And I think increasingly they're sort of finding that that's not really the case and you know, not just the likes of ourselves, but you've got other people in the, sort of the ecosystem of email marketing world, which have sprung up because email marketing vendor. Can't can't have everything well, not that they call themselves even marketing more that multi-channel omni-channel in

[00:16:15] Matthew Dunn: cloud.

[00:16:16] Sean Duffy: Yeah. Yeah. Well, there's those sort of things. Yeah. I mean, a lot of them don't even send their own emails. Right? Yeah. That's the crazy thing. So, so you've got deliverability platforms, you've got content rendering platforms. You've got standalone email editing and content management systems. So I know you interviewed viewed Elliot from taxis, a prime example, you know, that sprung up because CSPs don't do a good job themselves at creation

[00:16:39] Matthew Dunn: of email.

[00:16:40] I literally just adding onto that, literally this, I rarely have two conversations for this podcast series back to back, but 7:00 AM my time this morning, I had Roland McCormick from chameleon on, as a guest. So like I talked to him an hour before I talked to you and that's all attacks. He chameleon he is let us make the.

[00:17:00] Collaborative work involved in the design of email messages, better, easier, faster. Why? Because, because the editor built into the ESP may not be doing that job to the level you want or otherwise they wouldn't have a market. Yeah.

[00:17:17] Sean Duffy: Yeah. And, and then you've also now this whole range of CDPs or customer data platforms, which has sprung up, which really blurred the lines between who was an ESP, who was a CDP, but CDPs.

[00:17:31] And so it was so buried. I mean, you've got something like segment.com, which is, I guess, to explain to marketers, it's probably closest, closer to Google analytics in terms of. Management system in a sense first and foremost is probably how people would sort of almost recognize recognize it. And then you've got so many of the guys that much near like an old CRM or single customer view platform where, you know, so, so there's a real mix of things out there, which because the email channel has sort of evolved, I guess, in terms of its matured and, and

[00:18:06] Matthew Dunn: the, and the, the, the, you know, the person at the other end of the message has evolved in terms of the, you know, attention, they've got the channels that are hitting them, the expectations that, that, you know, that they've got.

[00:18:20] I mean, the funny thing is I'm not sure. And back to your statement about email marketers, not really having personalization deep personalization is sort of the habit, if you will, or a a recognized. Cor a pillar of the practice. I posed this question on the only influencers list you mentioned maybe a year ago.

[00:18:41] I said, essentially, what is personalization? Is there an actual in-depth accepted model that says, this is personalization. I got zero, zero response, not a single answer. And that's probably the top collection of email marketers you could get. Yeah. And we don't even know what personalization

[00:19:02] Sean Duffy: is. No, no. And, and I think this is the problem is there's sometimes there's tick box mentality around.

[00:19:09] We need to do more personalization. So we'll, we'll, we'll do that token gesture and tick the box, but it doesn't move. If you do it in a certain way. Not certainly not, not to the extent you need for the amount of effort it takes. Yeah. So we've always thought about throwing around the term individualization, but I think there's enough terms out there, but don't want to go to create another one.

[00:19:32] So when someone else will get the wrong idea from anyway. Yeah. Because I've already seen other people use that term and it's yeah. It's just, it's just

[00:19:41] Matthew Dunn: as long.

[00:19:43] Sean Duffy: So I would say this there's a hierarchy isn't there there's personalization, which is dear first name or merging data into an email. There's that level.

[00:19:51] That's going to get you kind of this far and it's, but it's fairly low effort. Yep. Then I guess you've got, you could, you could count. This is the bay I've had on with a couple of oys segmentation could be an occupant personalization, but you were limited to how many versions can you create? Yeah.

[00:20:11] Usually a handful because you're manually doing these things similar to that. Then dynamic content rules where you will maybe tailor different sections, that email. So if this happens, show this, otherwise, show this again, you've got to write those rules. So it's only going to be as personalized as as many rules you can write and maintain.

[00:20:30] Then you've got you into the world of algorithms and, and things where you are, you aren't getting to the point of individualization. And this is the bit where there's probably a lot of stuff happening in the moment. Yeah. But as you touched on with AI is it's is only as good as the data behind it and the algorithm.

[00:20:50] Yeah. And there's a lot of functionality is perhaps been produced, not just in personalized in the continent email, but in predicting the best time to send an email predicting which customers are going to churn those sort of AI solutions where you kind of get feeling. How it's come about is a developer, the marketing cloud provider.

[00:21:12] Should we say, it's giving you that there, that you found to probably played around with one of the off the shelf tools from Amazon web services or, or so on. And they'll generate you this generic sort of AI model, which is just the pre-programmed algorithm, which can take your data. And everybody's going to get a spot out in the same way to come with the same results and, and it's promoted as being bleeding edge or whatever term you might want to call it, but it's not proven tech and it's not tailored to you as your business.

[00:21:48] So every business is. Using that technology, whether you're say an airline discount retailer or luxury goods company, we're using the same algorithm, how can that possibly get you the optimum sort of personalization to the customer?

[00:22:05] Matthew Dunn: And if we don't have a good, we don't have a solid definition of personalization the old day, I asked him, you can't you can't manage what you don't measure.

[00:22:15] If we don't have a structure for that, that we can measure, then how the heck do we measure the impact of the thing? So we're sort of back to it. You know, it seems promising let's try it. Oh, gee, that helped a little, but there's no real accepted. If you do this, you'll have X percent different results. I'll give you one example.

[00:22:36] I'd be curious about your reaction to this. I've booked stuff through Airbnb more than a few times, not in the last two years. Very much. Thank you. COVID but planning a trip for the summer. I went on Airbnb and did just a little bit of a gee. I wonder if we could get a place in this small town at, up in the tip of Washington, two, three days later, I'd get a really nice looking email from Airbnb.

[00:23:02] Like, Hey, looks like you're looking at Republic, Washington. Very, very, very personalized. Like, not only was it to me, my inbox, my name that town that time. Why? Because they had the. Yeah, right. And, you know, I showed clear intent by browsing that city. Obviously they bothered to connect the dots, record the browse or the potential booking activity.

[00:23:27] Push it back over to the email system. We big company, very good at their top at their game. That's not a trivial amount of work and I'm guessing they wouldn't do it. If it didn't help him move the dial. Right. Did it remind me? Oh yeah. I didn't book that trip. Sure. Useful, but man, that's an exception.

[00:23:47] Sean Duffy: Yeah.

[00:23:47] I, I, I think it's, I think there's a quite common one. The remarketing piece is where we're personal stationary levels. You use not cart abandonment stuff,

[00:23:56] Matthew Dunn: cart abandonment.

[00:23:57] Sean Duffy: Sure, sure. There's a few problems with that though. A lot of people won't measure it correctly, so you may well have gone back and booked directly anyway, even didn't get that email.

[00:24:08] Yeah. And a lot of the providers out there, the specialist can't abandon the drivers out there nowadays with. John you've commissioned for that sort of email service. So they'll go and do it all for you. And they'll say anybody, which in your case made a booking on the back of that. Having after receipt email, we going to count 100% of that revenue as we generated that, right?

[00:24:29] So therefore you might pay say five to 7% commission on that on a standard retail site, you might expect the counter Kanban natural return rate for people which abandoned the cart, but then come back and buy later. Let's say it's 14%. If you send the emails, you might get 16% return with me, you know, a classic series of three emails, but over a few days if you say the 16th sentence, you will real revenue.

[00:24:58] That's a complete lie if you're using your click through funnels or just to match them who who's got that email it's in reality, it's that 2%. So one of those emails are great. They're often measured on, but also they don't generate that much scale. You know, some of the numbers you see bandied around by people on the perhaps not incremental, should we say?

[00:25:17] And this is where a lot of people I think get personalization wrong is they look at those sort of easy to use sort of easy journeys to sort of put them in, not abandonment, browse, abandonment, that sort of thing. Maybe next purchase, but there's not much scaling those. You know, the, the key thing is what about the 99.9% of the emails that you sent the business is usually announced?

[00:25:38] Like how can you scale those? Yeah, but with personalization, that's where the money is. That's where the money, if you make it, if you make a 10% increase, then. That's going to blow away every sort of trigger that you can, you can probably do it. Yeah. In terms of what incrementing delivers.

[00:25:54] Matthew Dunn: It's also, it's worth making a site observation that cart abandonment messages also have a rude assumption that the customer is stupid.

[00:26:04] Yeah. I did leave it there. You're absolutely right. I chose to leave it there. I didn't need you to tell me I left it there. Thanks. Might piss off. No, we're not going to Republic wash. She did. Don't bother me. Right. And so okay,

[00:26:21] Sean Duffy: exactly that. And you know, the I, I just think that actually, if you just sent any email at that point with your logo, which, I mean, if you look at most businesses, usually emails, you look at whether a lot of the click-throughs happen logo top left the NAF bar that you put in it's because, you know, DentaQuest was just the one in which really.

[00:26:44] Went on a lot about, about this, but it's true. You don't even have to open the email to be influenced by it and be reminded. Yeah. Yeah.

[00:26:50] Matthew Dunn: You're absolutely

[00:26:50] Sean Duffy: right. You're right. So, so how much of, how much of it is really the content you put in that car recovery email? How much of it is actually just because they saw your brand in inbox in the first place.

[00:27:00] Right, right. As well as the course then likely to come back in the first

[00:27:04] Matthew Dunn: place. Yeah. Yeah. I get, I get, I get more non-linear about I'm thinking. Hmm. And I'm not picking on Airbnb. I'm just using it as a, as a proxy for this whole set of issues at the moment. But if data come back with parks, hikes, bike trails, whatever in that area, as opposed to the really linear, you looked at this and we want you to book it, that, that might've helped me make a decision that I obviously didn't make when I was looking at the cart, so to speak.

[00:27:35] Yeah. Right. Yeah. Yeah. It, it it's, it's a tough complex and very ironically immature terrain for marketing. Like the, the content that actually moves people to act seems to be the last priority on the list. We throw technology at everything except the content a lot of ways.

[00:27:55] Sean Duffy: Yeah. And there's one other big barrier to personalization for any, my marketing is evil.

[00:28:00] Mark is generally used as a push mechanism to support the company's short-term goals. If we have promotion, we need to push it. And yeah, but that becomes a barrier because actually that's not customer centric content you're putting out there. Yeah. So you're you, you as a marketer to have that first hurdle for, it becomes, how do you, how do you do personalization when you know, you can't.

[00:28:26] You want to in control of the content going out in the first place. Right, right. You know, and everybody's after big numbers of to send something to, because big numbers means more revenues, the theory isn't it. But we all know that actually it's only a small percentage of less, which is going to get to respond to that particular offer.

[00:28:42] And you'd known advance. You could work out who's the most likely to respond to this offer on say camping equipment, you know, well, this is second the customers, which we know have an interesting camping equipment or, or the second customer should look like the customers we have, which buy camping equipment, if that makes sense.

[00:29:00] So yeah, that, that's it's another area that hurdle they have to overcome. Yeah.

[00:29:06] Matthew Dunn: Yeah. And, and the, the, the complexity of making an email message render, display, you know, look, look good at all. It is such a source of friction in that chain of decisions. We ended up back at where I think you were starting about some of the production efficiencies and flow efficiencies that the so-called real-time sector bring into the email marketing game that aren't really possible with the editor HTML, the conventional tool set, if you will.

[00:29:43] Yeah. Or,

[00:29:44] Sean Duffy: or there's there's things they are possible, but they take a long time and they take, they take another person to take a coder. And the more people we add into that chain you're exponentially increasing the amount of time it takes to create tests, sign off an email. Yeah. Just the communication string gets

[00:30:03] Matthew Dunn: so long.

[00:30:04] Yeah. Yeah. What is what I read somewhere and maybe you'll have more accurate data about this, but it's something. Average of two weeks, the cycle, the cycle of how long it takes to say, we're going to send marketing email X before it goes out the door. It's not done that the day of it's frequently planned out way in advance, reviewed, signed off revised, you know, lots of cooks in the kitchen, lots of inputs, et cetera, et cetera.

[00:30:32] Sean Duffy: Yeah. And it's why emails are often sent at 5:00 PM because there's got to be set, then I'm leaving the office. It's not going to get sentenced. Yeah. W what I was in the email vendor, well, we, we had a managed services team and, you know, the poor person whose client, they were looking after Friday afternoon, 6:00 PM.

[00:30:54] And they're still in the office, which they're waiting for that, that final sign off. And then of course the client signs it off. There's some sort of error in there. Not necessarily something we've introduced, but some content we shouldn't point right. Match to the website Monday morning, you get that phone call that went out wrong.

[00:31:10] Well, you know, yeah. That's the sort of impact that those sorts of processes have stressful

[00:31:16] Matthew Dunn: for everybody stressful for everybody. Now, I, I, I've got, I've got to tell a tale and compliment the company. You already mentioned big dog in the real-time space, movable ink. I went down a couple of years ago to a road show event that message gears message gives was making a tourist sort of introducing himself to the world.

[00:31:38] And they, they work closely with moveable link and they had a speaker come in from the email group at the outdoor retailer REI who happens to be based in Seattle, south of me here. And the REI guy had this rave tale to tell about. The REI daily DZ deals, email after they brought real-time content into the mix.

[00:32:05] She's like, instead of someone being stuck with you're on the hook for Tuesdays daily deals, you're going to be here till midnight, dealing with the last minute changes in this and the sand, et cetera, et cetera. I said, we all got to go home at the end of the day. When we, when we brought real-time content to bear on that content.

[00:32:25] Set of decisions that have to go out the door in a daily deals, email. I was very encouraged by his story, even though he's talking about a different company like this could work. Yeah,

[00:32:35] Sean Duffy: yeah, no, no, indeed. You know, I think we see that as well. I mean, talk about the, the TV shopping channel with their TV guys.

[00:32:42] Now they don't touch it, but they found, they found plenty other things to get busy with. So they're still stressed out in terms of, because you're not based on that anymore. So here's his load more tasks that we can, we can sort of leave with you, but I'll

[00:32:53] Matthew Dunn: tell you the camp, the campaign genius story, and that same in that same domain, we have a, we have a customer who's a custom home builder.

[00:33:02] So w when they send out an email, they're not, they're not hoping to sell, you know, bananas or, or television. They're talking about million dollar homes, but especially with the housing market in the U S right now, That's that they may be gone. They're probably gone. So they really want their email content to be inventory accurate.

[00:33:26] So what we, what we do for them is we actually pull the live inventory and pricing and availability from their website at the moment of open. So that the email that customers get our houses, they could buy. Yeah. Yeah. And like, okay. That seems like a good thing. Even one sale. And I think they're probably doing just fine in terms of, in terms of the cost of the technology.

[00:33:53] Sean Duffy: Yeah, I think so. I think another example would be. If you normally, the way a marketer would pick eat products, read email. Yeah. Is there kind of look well, let's make sure nothing is, it's got penny stock behind it and then spin out. But, but that means you end up with this fragmented stock, a particular sale environment.

[00:34:11] So I'll go back to the footwear retailer where you know, you'd have one or two sizes left and you would put them in an email because not everybody's got that size. So you'd look at the conversion rates and Google analytics or the sale emails. And it'd be like north 0.2% from click to conversion because people clicked, oh, it's only in science for a left, but if you can.

[00:34:35] This type of technology with some rules behind it, as well to say, only show if it's in stock in that customer size, all of a sudden you've opened up all this extra stock, you can't get rid of otherwise. Yeah. You know, out to the market and it's more effective because it's more relevant to each, each individual,

[00:34:51] Matthew Dunn: the visual.

[00:34:52] Yeah. No, that's fine. That's a really good example, actually. And God forbid they cross record. My past, you know, jacket and shirt purchases, and go probably got big feet. Let's hit, let's hit him for this particular boot sale. Why? Because we got a lot of size twelves on the shelf. Yeah. That's that, that may be somewhere in the personalization model, foot size, right?

[00:35:17] Somewhere down complex

[00:35:19] Sean Duffy: model. It's a common thing. Isn't it. We also have this term called underserved segments. So a classic one maybe travel that might be people just travel on their own or in retail. That would be people big or small sizes. They're worth quite a lot per customer, but there's not many of them usually.

[00:35:36] So as an email marketer, you focus on Mr. Average, generally, don't you where most of the, most of the money is, but if you can start creating emails, which don't take you any efforts or, you know, it's a standard template, maybe best you change the subject line, you change the hero. Yeah. You've got a system which ultimately pulls through the content for that email.

[00:35:56] Yeah. You didn't have to touch, you can start increasing the revenue of getting from those tiny ball, but small segments quite, quite easily. So there's, there's that sort of option as well. And this is why I say real time is too bad terms. Sometimes

[00:36:13] Matthew Dunn: personalization. I'm sorry to cut you off Sean, but you just said something super insightful.

[00:36:18] And I'm trying to figure out how to capture in my head. It, what we think of is personalization. Like this guy, this guy, this guy may have more business impact when it's when it's when it's identifying the, not the segments, that's not quite the right word for that. The, the, the, the, the cases of value, there may be one, there may be 20, there may be 500 of them, like your solo traveler example and tailor.

[00:36:48] What's said very much to what makes them different. What makes them special, unusual you know, a particularly good customers for some aspect of your business. And that has very little to do with what their first name is. Yeah, yeah, yeah.

[00:37:05] Sean Duffy: Hmm. And I think that's also part of the problem with personalization projects is marketers don't necessarily imagination that those things can be done true.

[00:37:16] True

[00:37:17] Matthew Dunn: at the time to do it.

[00:37:19] Sean Duffy: Yeah. What? Yeah. Cause I haven't done it before. They don't, they don't know it can be done. Yeah. Mark is often do the things they, they know they, they can do already, they have done in the past. And I think where we find it's much easier to get business on board with personalization is you sort of turn it round and you start with the business.

[00:37:39] And then personalization just happens to be the means of solving that problem. So again, in travel that's things like, do you have a problem with late availability, you know, properties you can't sell? Yes, we do. Well, let's create late availability content automatically that sits in every email, which is going to help you nice sell through some of those properties you otherwise would in retail.

[00:38:00] They might be a sale model. We have lots of fragmented stock, whatever the problems are. Usually personalization and email can consult those. And again, the danger starting real time content is it's you starting with. It's not solving any, any problems specifically for the people. Yeah. I think that's where another reason I hate that term.

[00:38:24] Matthew Dunn: I did, I do too, especially since, especially since apple through that what's your term spanner in the cognitive spokes. Although I appreciate apple taking on the performance cashing job for us. Like, thanks guys. You took a load off, you took a load off our cloud and made things really snappy, but whatever.

[00:38:48] Yeah, it's it's gonna be a while. It's gonna be a while till we see increasing adoption, I am of real-time or whatever the label is. We, we come up with as a substitute. I am encouraged to see th th the company that's done this for a long time and, and, and sort of, and has the, I think the biggest market share moveable.

[00:39:05] On a heck of a healthy course, they get to ring the ring, the bell at NASDAQ a few weeks back and celebrate I think they've darn near doubled revenue in less than five years. So like giddy up good for them. For the 99% of the market that says, Ooh, I can't afford that. Maybe they'll start talking to rig nine.

[00:39:24] Sean Duffy: Well maybe in the campaign genius. But I think there's also something that goes back to the, the value is again, realtime content, just doesn't you talk to somebody say, even if it's $50 a month fee, if it's not delivering any added value and it's just seems an additional hassle that would have to deal with.

[00:39:46] Then it doesn't go anywhere. So you can't just be, you can't just be more cost-effective should we say then, then our friends that move a link?

[00:39:55] Matthew Dunn: Yeah, yeah. No, it should not. Yeah. That's not a reason to do it. You're absolutely right. Yeah. You're absolutely right. And, and, and the kind of true automation, you know, I'm fascinated by your TV guide description like that, that true automation of the end content.

[00:40:14] That's, that's a heck of a value proposition. The fact they don't have to worry about that schedule. It's a big deal. Yeah. Yeah.

[00:40:26] Sean Duffy: It is that they just have to remember to fill it in which on occasions they, they have gotten to which, which, you know, means it doesn't appear anywhere, but on their website even.

[00:40:35] So yeah, but yeah, it's, it's, it's about thinking different. Because everybody, every market is brought up in the concept, how we do email. We have this blank email when we have to fit it in manually each time. Th this is what the textbook says for us to do,

[00:40:49] Matthew Dunn: but we've got, this is how our ESP works.

[00:40:52] The big piece of it, too.

[00:40:54] Sean Duffy: Yeah. And this is, this is what I've done previously, my last few jobs. And you tend to see that a lot of Yuma marketers will bring the same tactics to tweak the job. They go to every two, three years. And you know, I think, I think certainly my career I've always just tried to take that step back and, and really not be influenced by what I've done before.

[00:41:16] I mean, I use that as a reference, but you have to really get into the nuts and bolts of what, what is the specific problem that we have as a, as a business that w w w we're trying to solve? I don't think those, those template approach. Really very effective. Again, they can look like they, they do cause the thing which really gets, gets my back-up quite a lot is it comes out attribution, you know, and people claiming X, millions of revenue and they, you know, there's a lot of in the DTC space is a direct to consumer there's.

[00:41:49] A lot of people do use a certain ESP who has a certain attribution model. Should we say is I wouldn't necessarily suggest is that robust? Which is anybody who has opened this email, which goes on to buy over the next seven days. We count that order and that's used across cart, abandonment, welcome emails.

[00:42:09] Next purchase emails, all those emails, which those customers can didn't send them an email they could perform massively well anyway, and I think a lot of people do get carried away. Some of the headlines, they seal nose and they take the simplistic strategies across. And don't really understand that, you know it's not really delivering incremental benefit for the business and I'm starting to see some of that now with NPP.

[00:42:37] Some people, some clients I speak with that are open rates. They're amazing. Not now, now it's rolled out across so many places. And they haven't yet got to the point where they're sort of, they sort of know why, but they're not telling anybody further up the change.

[00:42:56] So there's going to come point where you're going to have to, you have to, you know, say, hang on a minute, you know why is opens going up, click Sansom? Well, you know,

[00:43:10] but yes, Well, while they can, that they're using is a good news story. And yeah, it's in no one's interest long-term for that, for that to be the

[00:43:18] Matthew Dunn: case. Yeah, no, I, I would agree with that. I had I had a I had a guest on this show a few weeks back, peer lifted from 1440 media, not an email marketing in his case, the company is in the journalism business.

[00:43:33] 1440 is a daily by E via email news source. And when we were chatting Pierre and I were chatting about MPP, I think after we stopped the recording he noted that the concern that a publisher like 1440 has with the breaking of the feedback loop induced by MPP is that over time he's going to lose track of.

[00:44:04] Who's actually reading like the, the engaged subscriber which is, that's a measure of value for a publisher, right? It's like, yeah, I don't actually, I'm going to have a hard time knowing I said, huh, I w I actually think we'll be able to help them with it. I think there's ways to answer that in the cloud of data that you get, but it's not going to be a trivial, not going to be a trivial exercise.

[00:44:28] Sean Duffy: No. And I think part of the frustration with MPP is what I thought might happen has happened is, is the ESPs perhaps stepped up their game. Oh, they done their side of the

[00:44:39] Matthew Dunn: bargain punted complete, sorry, be American term punted. Right? Like, come on. Most of them are done either nothing or, or damage with a, with a blunt measure.

[00:44:49] Oh, just leave them out. Bad decision mathematically.

[00:44:54] Sean Duffy: And I have some sympathy. The fact that it was, I think it's quite difficult to have a strong prediction during the beta period of how is this really going to be implemented. But I think we've had long enough now to sort of stay well. You really should have got something in action better to, to sort of, to, to help with this.

[00:45:15] I mean, I, I know ESPN switch, you know, they, they report out what email clients are and it's still down as unknown, you know, which, which is horrendous for apple iOS 15. But yeah, I just think that they should have done more in terms of what real estimated open rate

[00:45:32] Matthew Dunn: T to be, to be fair and, and very, very inside baseball that an accurate answer is extraordinarily.

[00:45:45] Tricky. We've done it. I've got the patent filings, but I gotta say like, wow, that this, this, this is hard. This is a hard problem to solve even adequately. So yeah, and the ones that have done nothing in a, in a funny sense, they may be, you know, throwing up your hands and saying, don't look at open rates, which seems to be the kind of prevailing email marketing mindset.

[00:46:08] I, you know, I, I, I think it's, I think it's a unwise overall because you that's, it is a different signal, right? Then the, I click the website, the iRead, your cotton-picking message, or, you know, people read this message more than that message. That is a different signal. And it's not an, it's not a, a non valuable piece of feedback, even though it's less accurate, it was already an accurate.

[00:46:36] Now it's even less accurate. That doesn't mean its value is.

[00:46:41] Sean Duffy: No, but I still think that's a level of accuracy. You'd get to knowing that a lot of these opens happened late on your day, long after the email has been sent all of the, these, these aren't small companies we're talking about here, we're talking billion dollar companies.

[00:46:55] And I think it's sometimes a symptom of companies which maybe been acquired that ESP and maybe in a mall, you know, some of those are better than others. Should we say in terms of innovating? Yeah. But, but I think, yeah, the general shrug and have it go in, even if you don't get it rights, at least have a go.

[00:47:16] And I think the general attitude of not helping him, it's the one place that really needs to be is in the MSP dashboard because yes, agreed, agreed like yourself. You can, you can create solutions outside of it. And

[00:47:27] Matthew Dunn: things, but it belongs in blogs, belongs in with the other.

[00:47:31] Sean Duffy: Yeah. Yeah. And on the, on the salty side of our business where we're doing something with a very big ESP, which actually allows us to write extra data to the tables, we're going to, we're going to flag which ones we think are definite false opens and which ones we think are possible, you know, good open stock because it gives us a level of, it's not perfect.

[00:47:54] It wasn't perfect before anyway, but in terms of when it comes to, has this customer open an email and asked three months, it gives us a bit better handle because, you know, was it 60, 70% of your audience want to open an email? If you've got your best, best way of handling, a lot of deliverability issues is getting rid of the, the inactives and stuff.

[00:48:12] I don't know. That's why I just feel that the lack of support from ESPN. Disappointing, but not unsurprising.

[00:48:21] Matthew Dunn: Yeah. Well, Hey, listen, if you've got, if you run into ESPs who don't want to tackle this problem, but, but do want to solve it, like tell them to call us because we actually in conversations, we're in conversations with a couple of ESP is about helping them about helping them do that with that, just that one, narrow, you know, one area aspect of the data.

[00:48:41] I also think the reason it's been difficult for the ESPN platform, the vendors in the email space to grapple with this is why is developer and specifically data talent is. Such a, such a precious commodity at this particular moment, like your dev staff is a hundred percent busy already guaranteed.

[00:49:07] Before apple made that announcement. They were already booked out for years to come and you were trying to hang onto them and not lose them to someone who was going to pay him twice as much. So throwing, throwing a very complex problem on their plate. Yeah. It's tough to add to the stack.

[00:49:24] Sean Duffy: Maybe developers like complex problems.

[00:49:28] Don't they? I think

[00:49:32] some do the good ones. Do I suppose there's other streets just want to work nine to five? I get that that's the same in any profession, I suppose. But it's also the, also the ability to move things around in your development schedule as well. I just saw so many of the webinars from ESPs saying, this is what you need to do marketer with MPP.

[00:49:50] And it just seemed very little about how we're going to help you. This is your problem. We're not necessarily sharing that enough. You know, so maybe I'm being a bit harsh, but yeah. And, and, you know, I guess it gives you a great opportunity because you've solved the problem. So,

[00:50:05] Matthew Dunn: Well, w w yeah, w w we, we, you know, we've got, we've got a, we've got a solution to, to the level of what I think are useful and reliable, useful insights.

[00:50:19] It's not the same measure we were getting before, but, you know, we, we took a different not to delve into that too much, but we, we took a political polling approach to the opening. Problem rather than a straight math approach. Cause straight math doesn't work with, with false, with, with with a falsified, a flood of falsified, potentially falsified signals.

[00:50:42] But yeah, it, it, it, it was, it was, it was interesting. And my God, back to what you said about AWS and so on, if the cloud compute ecosystems weren't at the level of maturity, they are now, it would have been complete impossibility because the data crunching is a freaking nightmare. It's just a Bismal, like just ingesting that, right.

[00:51:11] You know, when a campaign is going, a million opens or 5 million, one weekend, just, just being able to ingest it and not lose stuff for starters was hard. And. Y, you know, watching SQL boxes, sorta blow smoke for four or five minutes on a single query. I'm like, Ooh, I don't think we can get there. Get there this way.

[00:51:31] Sean Duffy: I mentioned that the cost to put that infrastructure in the old way would been, I mean, let, let Mrs. A really good example. They, they need to exist as a business doing what they're doing because of AWS. I don't know if you remember the olden days when you only had a couple of options. I think it was pivotal to the rusty and turn past doing those rendering in different email clients.

[00:51:53] And Paul who's who's from finance, not, not far from where I grew up. He, he invented the solution on a couple of old PCs. But for him to scale to the required level, you know, it would, there's no way he could. That's why people, the rusty and return path there was so expensive that even us as an ESP, we couldn't afford to use them.

[00:52:13] Wow. Interesting. Oh, hang on AWS. They're able to scale up all these virtual machines, running, all these different things. They didn't have capacity issues. They didn't have to build up here in order to cater, you know, in most of the time they'd be down here. And I think that's happening a lot in the email space solutions.

[00:52:31] That's so much more accessible now price, but like ourselves, like those guys, like, like actually a lot of email sending services is it's become a lot more cost effective because of the cloud. And that that's been, I think, a huge transformation in the last 10 years.

[00:52:45] Matthew Dunn: Well, as, you know, as a startup like reignite go go 10 years, ballpark, 10 years back, you would've needed a VC and five to $8 million in data center buildup.

[00:52:56] And then you could start actually doing the work. And now you take out a MasterCard and start reading the AWS documents. Basically instead.

[00:53:08] Sean Duffy: Yeah. I mean, Aus, isn't always cheap if you, if you do it wrong, you can have some access if you say so, it's make sure you set up those alerts, but yeah. And it would have also been a lot of hassle.

[00:53:22] It was, it was the time, I mean, back in our ESP days, email center, we we had all our own servers and I think that's still probably the K having been bought by another company now. But yeah, we have scaling issues, both bandwidth and service, but it was the fact that they had to be micromanaged almost along the way.

[00:53:41] And you're right. Those, those big boxes we had to cater for this when actually most of the time we're doing this. But the amount of times that someone would be having to go down to the data center, some would be up at night, just grading servers. Yeah. I know the guy and he was so stressful.

[00:54:00] Matthew Dunn: Oh, quick.

[00:54:01] I knew cloud computing was going to happen story. I was a CEO of a venture funded out 2005, 2006. Unfortunately the technology was a decade ahead of the market. It was it was a music identification and relationship stack, meaning find me a song like this. And it actually algorithmically found you a song like that which has a lot of complexity to the, to the tech stack.

[00:54:29] But we ended up cutting a deal with, with a good sized record label. And all of a sudden we had a couple hundred thousand songs that needed to go through the algorithmic analysis, which was about as long as it would take to play the. Full CPU to analyze the song three minute song. I need three minutes of a desktop CPU to analyze it.

[00:54:51] So like multiply that by 200,000 and go, oh crap. Yeah, we get to do, and by sheer luck, I was keeping the feelers out. And I read about these, these goofballs at Amazon that had virtual, we're making virtual compute instances available. So like very, very baby steps of AWS. And I went, oh crap guys. Instead of buying tens and tens of thousands of dollars worth of Dell servers in rent, this capability for CPU crunching in analysis from Amazon.

[00:55:26] And we by guy did it in oh five, but bandwidth was so scarce. Then that uploading that flip in catalog of 200 songs securely because it's crown jewels for the recording. Yeah, that was the bottleneck. Like the guy who's babysitting the, the internet pipe instead of the server, but it was still a lot less expensive

[00:55:48] Sean Duffy: now.

[00:55:48] Amazon used to didn't they used to accept you post in a USB

[00:55:52] Matthew Dunn: stick. We were so early. They didn't have it yet. Yeah, because

[00:55:57] Sean Duffy: I think it was a case in South Africa where somebody for laugh they, because the bandwidth was so poor, I think they, they sent in via a carrier pigeon or something. I mean, now you can start loading

[00:56:07] Matthew Dunn: it.

[00:56:08] They'll they'll, they'll sit in a fricking semi. Now, if you have enough data to transfer, Amazon has a semi served. We will send a semi to the park at the corner and plug into your data center and transfer all your stuff.

[00:56:21] Sean Duffy: Yeah. Yeah. And, you know it's, it's incredible how the world has changed, but I think possibly sorts of scary that, that.

[00:56:28] So much more dependent on three or four companies for our entire internet infrastructure. And we do notice that if Amazon goes down, she has done a few occasions. The internet does

[00:56:39] Matthew Dunn: as well. Yeah, it does as well. AWS we're we're Google cloud mostly which is a choice I'm happy with after a good number of years still.

[00:56:49] But same, same reliability problem. Right. They went, GCP had an issue a few months back and I was like, oh crap. I can like, literally you can't do anything. It's not as quick. It's an earthquake, technologically speaking, all you can do is duck and cover.

[00:57:06] Sean Duffy: Yeah. But I think the nice thing is you're sharing that problem with so many other people and from a client perspective in, in our, again, the ASP days, if our servers went down for some reason, because the data center went down.

[00:57:18] Yeah. They would notice anybody else has gone down because very few people have gone down in the world. So it's purely your problem. But in this new world where oh eight us has gone down people are much more understanding about that sort of thing because probably their websites go down as well.

[00:57:35] Matthew Dunn: And I do find myself wondering I'm a skeptic, I'm a skeptic kind of bunch of sort of intellectual historical levels.

[00:57:42] But with the, with the current buzz around cyber currencies, I'm waiting for someone to go, oh, what do you mean. Can't get to my money. Oh, crap, right?

[00:57:55] Sean Duffy: Yeah. Well, there's plenty of times as evidence of people which have forgotten the key. Oh yeah.

[00:58:02] Matthew Dunn: In the trolling, the city dump for the missing hard drive, that kind of stuff.

[00:58:06] Sean Duffy: But those are the things, but the, I mean, it's, it's off topic, but that, you know, I think you know certainly I know the bank of England is, is looking at cryptocurrency payments and things now, not in the sense that you and I knows about Bitcoin. But we're, we're in the, our office is in the same place as someone who's working on that technology right now.

[00:58:25] I mean, I don't understand any of it, but it sounds great. So we're just not, so you don't want to see,

[00:58:30] Matthew Dunn: Yeah, it's it, it's, it's an intriguing, it's a new trigging frontier. I, I, I, I hate to see us warming up the planet just to find mathematical skills.

[00:58:40] Sean Duffy: Yeah, in fairness, this is what they, they don't do.

[00:58:43] They don't want the planet then model UN I don't fully understand. So again, I'll just not either, but yeah. It's environmentally friendly way of doing things apparently. So it's not like Bitcoin, which is going to you know ended up taking the world's electricity many times over,

[00:59:01] Matthew Dunn: Yeah, that's not going to work and that's not going to work well, I know we'd wander a field.

[00:59:05] I should probably free you up to, you know, to go help people solve real time content problems. But I knew this would be awfully fun to talk, Sean.

[00:59:15] Sean Duffy: Good to catch up finally.

[00:59:17] Matthew Dunn: Yeah. Yeah. So if you're listening to go, what are they talking about? Real-time and personalization re-ignite hq.com is a good place is a good place to start.

[00:59:28] And Sean Duffy CEO of rig night has been a wonderful guest. Sean, thanks for the.

[00:59:34] Sean Duffy: Yeah, no, thank you. Speak. See

[00:59:35] Matthew Dunn: Matthew. All right, man. Boom, boom. We're going to hit stop. Where's the cursor. There it is.

[00:59:42]