Podcast Overview
Maybe I’ll hear back from your AI Assistant… Sorry Michael is Busy”
Welcome to another episode of eCom@One, today, Richard is joined by Michael Duxbury from Dot Digital, a true veteran with over 15 years of experience in CRM, email marketing and martech. In this insightful conversation, they dive deep into how AI is transforming the worlds of email and CRM marketing, from boosting segmentation and personalisation to automating content creation, reporting and campaign workflows. You’ll hear how AI can supercharge both seasoned marketers and resource-strapped teams, uncover the real adoption trends (and misconceptions), and explore the exciting future of cross-channel marketing, including the latest on WhatsApp and next-level list-building. Whether you’re trying to decode the AI hype or searching for actionable marketing tactics, this episode is packed with practical advice and forward-thinking strategies to help you stay ahead in a rapidly evolving industry.
Time Stamps:
00:00 AI in CRM and email marketing
04:25 Embracing AI in Business Operations
08:40 AI in consumer purchasing journeys
11:22 AI usage in email marketing
15:41 AI simplifies multi-channel marketing
17:02 AI’s impact on marketing agencies
20:05 Using AI to Generate Emails
23:19 AI in customer segmentation
27:47 Using analytics and marketing tools
29:34 Using AI tools for digital marketing
35:03 Using WhatsApp for targeted campaigns
38:23 Introducing tone of voice feature
40:53 AI-driven marketing assistance
43:36 Email CRM marketing evolution
45:27 Discussing Steinbeck’s Cannery Row
Richard [00:00:00]:
Obviously a lot happening with AI and email, but what would you say is probably the most overhyped thing in AI at the moment?
Michael Duxbury [00:00:06]:
Experience tells me that the thing that you think is overhyped ends up being underhyped and vice versa as well.
Richard [00:00:15]:
Do you see AI sort of completely taking the whole process?
Michael Duxbury [00:00:20]:
AI can help you do what you've been doing, but it can also help you do something that you've never done before.
Richard [00:00:30]:
You know what's maybe some of the biggest changes you've seen in that time?
Michael Duxbury [00:00:34]:
Email keeps on going and it keeps on performing and looks like it will continue to do, you know, even with AI, maybe even because of AI.
Richard [00:00:46]:
Hey, everyone, Richard here. Welcome back to the E Commerce One podcast. Today I'm joined by Michael Duxbury from Dot Digital. Now, Michael spent over 15 years in CRM and email marketing and marketing tech, so he's got a front row view of how this space has changed and where AI is taking it next. In this episode, we unpack what AI is really changing in email and CRM marketing. Where brands are already getting value and why the biggest opportunity right now is using it to support better segmentation, personalisation, reporting and campaign creation without losing that human touch behind it. So whether you're trying to get more from your CRM email marketing, improve your customer journey, or make sense of where AI actually fits in your marketing, this episode is packed with useful insights. Now do me one big quick favor.
Richard [00:01:33]:
Hit the subscribe or follow button wherever you are listening. So you never miss an episode. Let's get into it. Hey, Michael, how are you doing?
Michael Duxbury [00:01:43]:
Hi, Richard. Yeah, very well, thank you. How are you?
Richard [00:01:46]:
I'm very well, very well. It looks very sunny where you are. It looks like I'm in some sort of dungeon or something.
Michael Duxbury [00:01:50]:
Yeah, no, it's South London. I mean, glorious in South London today.
Richard [00:01:53]:
So did we just. We're actually in England and it is sunny.
Michael Duxbury [00:01:56]:
Yeah, we're in England and it's sunny two days in a row. What's going on?
Richard [00:02:00]:
Wow. Fingers crossed for the weekend, eh?
Michael Duxbury [00:02:04]:
We'll be waiting by then.
Richard [00:02:05]:
I just text, I said before you, before you came on, I text my wife and said, right, I think we might have a barbecue at the weekend. So I might have jinxed it for
Michael Duxbury [00:02:12]:
all of us now. Probably. Probably.
Richard [00:02:15]:
Well, thank you so much for coming on the show. I know it's. We're going to talk about a topic and a couple of topics that they're very close to our agency's heart here and our team, so very excited to have you on. Thanks for coming on the show. Now I know you've been in sort of CRM and email marketing and tech for quite some time and I think more than 15 years at dot digital.
Michael Duxbury [00:02:38]:
Yeah, more. I mean longer than I care to remember, to be honest.
Richard [00:02:41]:
Yeah. That is a long time, isn't it? It must be an absolutely fabulous firm to work for.
Michael Duxbury [00:02:47]:
It is a great company to work for. Yeah, yeah.
Richard [00:02:50]:
So obviously 15 years, I mean that, that is a long time, isn't it? In one company and obviously one company doing, you know, a specific thing and obviously a lot of change, no doubt and a lot of seen a lot of things as, as we all have if we've been in the industry that long. But you know, what's maybe some of the biggest changes you've seen in that time?
Michael Duxbury [00:03:07]:
Well, I mean, so I've been working on digital for all of that time and it is essentially unrecognisable now as opposed to what it was 15 years ago. Of course it was very much an, you know, an old school email marketing tool 15 years ago and that's did.
Richard [00:03:22]:
Yeah.
Michael Duxbury [00:03:22]:
And now we describe ourselves as a cross channel cdxp. So we're both cdp. We do the data side of things and we do the X which is the experience side of things. So we essentially now do everything that allow brands to communicate with their consumers using the data that they hold on the channels that are most effective to them. So the, the, the product is, is hugely evolved compared to 15 years ago.
Richard [00:03:46]:
Yeah. So 15 years ago was upload a lesson, write some text, send an email.
Michael Duxbury [00:03:51]:
Maybe that's maybe beautifully simple 15 years ago and obviously it works.
Richard [00:03:56]:
It worked.
Michael Duxbury [00:03:58]:
You know what the funny thing is is that what worked back then does still form a cornerstone of what most marketers do today in an email in that sense has been remarkably resilient in that, you know, email itself has actually changed very little in the 15. In those 15 years lots of other things have changed around it, but email keeps on going and it's keeps on performing and looks like it will continue to do, you know, even with AI. Maybe even because of AI.
Richard [00:04:25]:
Yeah, yeah, I think it's, it's enabling companies to do more, isn't it? Where maybe they're maybe a bit afraid, unable, not sure or not not using, you know, amount of people we look at there do a tech stack audit and subscriptions or no subscriptions, but subscript, even paying for software that maybe isn't getting used, you know, email software for example. And obviously AI helping now to ensure that tech is getting used a bit more potentially. And that is one angle of course. So yeah, so you know, I remember like I think maybe 25, 20 odd years ago, you know, emailing and very much, you know, building lists of, you know, when I was in the computer business, selling computers before this, my first sort of gig and it was very much, you know we built, I would say we almost built that business on email marketing because you know, it was, it was a small amount of customers but they were very high value because computers were very expensive back then. So we were sell computers but the ability to send an email and then ultimately then get an order very, very quickly. But we were taking orders on the phone at the time. So they would send the email and then they would phone up. So we would set the emails and you wouldn't see the orders come in at that point at the beginning we did later on, but the phones would ring and I guess it's still, still an element of that.
Michael Duxbury [00:05:39]:
Still do that today. Yeah, funnily enough, I mean we're talking about me being@dot digital for 15 years. Actually been there longer because before digital was digital we were a web design and development agency and we built e commerce stores for people, we built websites for people, particularly in the, in the not for profit sector. So you know, I used to be involved in building e commerce stores for the likes of WaterAid and Fair Trade foundation, charities like that. And it was the effectiveness of email marketing and our clients at the time asking us to build something that sent out these, you know, these newfangled HTML emails that you can track opens and clicks on. So, and so we, so we built it, we built it for Carity and then we built it for another one and built it for another one and eventually we said to ourselves we should probably stop rebuilding this every time and turn it into a product. And that is essentially that when Dot Digital all began so actually was born out of, of E commerce and E Commerce.
Richard [00:06:39]:
Using it for yourselves for.
Michael Duxbury [00:06:41]:
Yeah, and using it for ourselves and for our clients. Yeah. And then eventually it took over and actually the web design and agency, we, we let that go and we, we just became top digital and grow up from there.
Richard [00:06:51]:
Yeah, that's great, that's great to hear that. Thank you for, for sharing the sort of founding story. Founding story, yeah. I didn't know that. I didn't know that. I hear, I hear similar stories quite a lot. Right. You know, founders of, you know, of agencies then you know they built, start building a SaaS and then that SaaS becomes, you know, something that's way more than the agency and goes on.
Michael Duxbury [00:07:11]:
I think Wishcramer's example is this base camp, right. Or 37 signals. I think they had a very similar story to us at a very similar time. For your listeners that know Basecamp, maybe they use Basecamp.
Richard [00:07:22]:
Yeah. So obviously that's going back 15, 20 years ago, you know, and 15 years ago in this case. But obviously now the, you know, email and. And CRM marketing. There's so many different things we can. We can discuss, but I'm very keen to sort of hone on specifically some of the AI piece. You know, obviously AI, we've. We're probably not done an episode for the last year now without mentioning the word, and obviously some very specific episodes.
Michael Duxbury [00:07:49]:
I don't think any podcast has had an episode that much like AI.
Richard [00:07:52]:
I think I'd struggle to do one now if I can. There's like a, you know, keyword bingo that I can't say. It's going to be so obviously a lot happening with AI and email. But what would you say is probably the most overhyped thing in AI at the moment?
Michael Duxbury [00:08:09]:
You know what? I've been. I've been in and around technology long enough to know that it's a dangerous game to play, making predictions like that. And experience tells me that the thing that you think is overhyped ends up being underhyped and vice versa as well. So I'm frequently wrong on that. Unless it's in the metaverse, which we can probably all agree that that's gone beyond being overhyped, because that's already dead. I don't know. That was an easy one. But outside of that, it's a tricky one.
Michael Duxbury [00:08:40]:
But when it comes to AI, and this is not specific to email or marketing, this is just in general, I think probably the things to look at are not. It's not the technology, it's the solutions that people are talking about that happen to use that technology. And there's a. There's a use case that is used time and time again. You'll see it in people's product videos, you'll see it everywhere. And that is essentially AI taking over purchasing journeys on behalf of consumers, whether that is buying a holiday, and then I want to go somewhere hot for two weeks for family of four, go and book it for me. I mean, AI can do that, right? That bit's not overhyped. It can absolutely go away and do that.
Michael Duxbury [00:09:24]:
Or whether it's a lawnmower, I want lawnmower. It Needs to be cordless, go and buy it for me and it turns up the next day. Again, AI can do that. I am not yet convinced though that that is what people, that's not what consumers actually want. And I think that can be dangerous to overlook the non technical side of things like this because I think with some notable exceptions, generally speaking, people like, dare I say it, enjoy being involved in the purchasing experience.
Richard [00:09:53]:
They like certain items. Yes.
Michael Duxbury [00:09:57]:
I mean holidays is nothing to win. I like looking at pictures of sunny places and deciding which sunny place. Yeah.
Richard [00:10:03]:
And there's so many variables, isn't there, with the holiday and so many things you don't know because you've not been to the place. Whereas if you're buying something that you
Michael Duxbury [00:10:10]:
buy regularly, that stuff and things that you, you don't know upfront so you wouldn't know to type them into a prompt, but that you learn as part of, of the journey of buying the thing that crop up so your requirements can evolve and change. So I think when you look at a solution, you've got to look at the technology side. Yes. But you've also got to look at the human side and try and figure out which ones are going to, going to take off. But again, I can't stress enough, it's a dangerous game to play because no one can predict the future.
Richard [00:10:39]:
I think this is it. I think there's obviously so much, isn't there? And if we just step away from email, from an AI as a whole, in, in E commerce, for example, you know, no matter what topic we're talking about, we could list a thousand things to look at, you know, pretty much or without hundreds of tools and so forth. And it's just having that, I think that sort of mindset of right, we've got to, we've got to be inquisitive, we've got to, we've got to put time. But how much time, how much resource do you put in as a company, as a founder, as a team of marketers or you know, tech people within the business, what do you allocate? Have you got any thoughts on that? Sort of what maybe the founders and the people with the purse strings that are listening about their sort of approach to approaching AI?
Michael Duxbury [00:11:22]:
Well, yeah, certainly, but I mean from a, from a product point of view to build on what you were saying there? We do all those things but also we've learned we have to really look at how people are using the AI tools that we build. And what we've learned is that they don't always Use what you think they're going to use and they don't always use it in the way you think they're going to use it. Now that's true of just about anything, but it seems to be particularly true for AI. For example, AI and email is actually a really interesting example because AI can either enhance and augment or completely replace nearly every part of, of email marketing, from the strategy to the ideation to the content generation to the sending to the deliverability to the analytics to understanding the analytics and then bringing that into a loop and doing the whole thing. And maybe this comes back to predicting the future and it being quite hard. That may happen, it may not. But there are component parts that exist today. What's really interesting, when you, when you read the blogs and you listen to the podcast, there's a lot of talk about using AI and segmentation and personalization, and there's lots of really good, intelligent, powerful things you can do that.
Michael Duxbury [00:12:39]:
But what we're finding in reality is that although people are using those things, they're using AI to achieve those things. The most common thing that people use AI for an email is essentially content generation and for it to act as a bit of a sidekick to augment what, what they're, what they're doing, which is actually one of the least advanced use cases of AI, but, but it's probably one of the most accessible. And so people, we, we definitely see people start there and then they start to graduate on.
Richard [00:13:09]:
So they're starting as, like using it as an assistant, you know, and getting there. Yeah, getting ideas, getting first, maybe second drafts of things and ideas and especially on the content side, like you say. But then that's that we probably could have the same conversation two years ago that, that bit, you know, really. So it's like the advancement more recently and the things that you're able to do with email platforms are the ones that are maybe not getting the adoption more so at the moment.
Michael Duxbury [00:13:38]:
Yeah. And it's that some of that may just be time and people becoming familiar with those tools. It might also be that users themselves need to figure out how they want AI to work for them. You may have bosses that want one thing, but you've got users on the coal face that want something else. And then what we're seeing at the moment is that people are very, very open to AI when they feel like it is helping them, if they feel like it's helping them do a better job or do the best job, then they will absolutely use those things. So that's why for example, when someone is editing an email campaign, the number one thing when it comes to AI that they are most likely to do head and shoulders above anything else, is to ask AI for what we call a campaign commentary. So it could critique what they've written and make suggestions. That's the number one thing that they do.
Michael Duxbury [00:14:35]:
And that again, that follows that same pattern of I want AI to be there to help me assist. Yeah, exactly.
Richard [00:14:41]:
Yeah, yeah. So, okay, so we've got obviously quite a few advancements, should we say, as a bit of an understatement, that are sort of under the hood, that maybe a lot of people aren't aware of, not using. Do you what are some of the other areas then that maybe are underutilised quite a lot that maybe digital, you know, have, that aren't getting used?
Michael Duxbury [00:15:02]:
Well, it's not that they're not getting used, everything's getting used just a greater or lesser degree. And you can see different patterns over time. I think one of the key things is when you are multi channel. If it was just email used to create an email and you were done, it's nice and simple. Now you're multi channel. And generally speaking, a brand these days will be mixing their channels up. And generally you see them doing that based on where they believe the biggest value. So if you have a customer that has a high customer lifetime value or predicted high customer lifetime value, mix the channels up, you can afford to spend a little bit more on those people.
Michael Duxbury [00:15:41]:
So you might send them an SMS or maybe a WhatsApp these days, in addition to an email instead of an email. Which means that for any given campaign, whether it's a one off product launch or whether it's an ongoing abandoned browse or an abandoned cart campaign, you aren't going to have one piece of content now, you're going to have lots of pieces of content. And it may start on your website for different channels. Exactly. So you might have up to five of them. And so where AI can really help is make you create once and it generates the different versions for you. So for example, you can create an email, so you can create an SMS or an MMS from an email campaign. So you write the content once it figures out what you're trying to do and creates that content for you as a marketer, you can go in and you can tweak it, of course, you can edit it, modify, and make sure you're happy with it.
Michael Duxbury [00:16:26]:
But again, it's just acting as a, an assistant and doing that heavy lifting for you. So you're not there copying and pasting code and figuring out which bits to edit essentially.
Richard [00:16:35]:
Yeah, yeah. I know our team use that sort of functionality a lot and obviously if you're not using that functionality it's like, you know. Yeah, it's pretty crazy, isn't it, that if you're still doing that, things like that manually. Really?
Michael Duxbury [00:16:47]:
Yeah. I mean if I, if I, if I was setting up a multi channel campaign, which I don't do a lot, but sometimes we sort of simulate things, I wouldn't, I wouldn't build a multi channel campaign without using that tool, for example. It just wouldn't enter my mind to do so anymore.
Richard [00:17:02]:
So where do you see it? Obviously the platforms have evolved incredibly. We use several as an agency with different clients, obviously we've got our preferred options, but obviously different clients come to us with different platforms and so forth. We've got an array of platforms that we work on and an array of tech that's obviously within those platforms around AI and so forth. And certain things are better than others, shall we say, and certain platforms and so forth. I think it's like where do you see, do you see AI sort of completely taking the whole process, you know, at the moment, obviously as an agency and as an agency is like, is it. Should I be concerned that, you know, obviously we manage, you know, a lot of email campaigns for a lot of clients. You know, what do you see with s sort of, do you see AI completely taking over everything? You know, where does the, where does the human. Where does the agency.
Richard [00:17:59]:
Where do the, where do the teams of marketers fit?
Michael Duxbury [00:18:02]:
Yeah, it is a really good question. It's a predicting your future question. Again, it's hard to say for sure. I think it absolutely could. Again, it comes back down to what people want. I think there's always going to be brands with marketing teams that want finite control over it. But I think where what you're saying comes into its own will be very small company, maybe those without or with small marketing teams, or maybe with known marketing teams. They will then be able to lean onto AI to automate a lot more of the, of the basic.
Michael Duxbury [00:18:38]:
Of the whole flow. I can see that happening, particularly because there are basics and you'll know this more than anyone that every brand will want to get set up. The things you do first, right. Which are your welcome program and I mentioned them already abandoned cart, abandoned browse, all those kind of things. Those can absolutely be automated by AI. They could identify who to send the content to, they could create the content and they could Split test, make edits and do that, the full circle. Allowing marketers at that point to not have to worry about the basics. They actually just get given the results and how it's performing.
Michael Duxbury [00:19:19]:
And every so often they might want to step in or they might want to override something and go, you know what? I want to try crazy experiment to see if this increases my results or not. But otherwise they can leave it in the background because no one wants to be forever twiddling with an abandoned car. Email. That's not what people want to be doing, but what people do want to be doing. And what we see time and time again is the creative thought is the ideation. Marketers like that bit of the job. They do not like trying to debug HTML. That bit boring for most people.
Michael Duxbury [00:19:50]:
And so I can see AI taking the heavy lifting out of the basics, allowing marketers to concentrate on that creative
Richard [00:19:57]:
time, more around the strategy, understanding the brand, more thinking, not thinking time, but more time to create strategy.
Michael Duxbury [00:20:05]:
Exactly that. And then using AI to help those ideas come to life. Right. I mean we're, we are working on, we have an MCP server at the moment, as you might imagine, and we are investigating at the moment the ability to build emails, digital compatible emails, so emails that you'd then be able to edit within our editor and know who you are and your data and your brand via the MCP server. Which is another way of saying that you'd be using your AI tool of choice. That might be Claude, it might be ChatGPT, and you may have that plugged into other MCP servers with the other tools you're using, might be HubSpot, the CRM, whatever it might be. So, and those are about you and you can come up with an idea. You know what, can you just, can you create me that campaign for me? Like, what could that look like? Or maybe even just can you create three different versions for me? It will go away and do that and then you go to your digital account and they will do that and you can go in and you can edit them and you can tweak them from there and you can adjust the copy and you can change the imagery.
Michael Duxbury [00:21:03]:
But you're not starting from scratch.
Richard [00:21:05]:
No.
Michael Duxbury [00:21:05]:
So I think it. Being involved in every step of the journey. Yes. Taking over completely. Probably not just yet, no.
Richard [00:21:15]:
That's good to hear. We're okay, guys. But now I think, I think we
Michael Duxbury [00:21:22]:
are, I think we're okay.
Richard [00:21:23]:
I think we'll always be okay where we're at the end of the day. You know, I'm sure we'll be doing another episode in three. If I, If I contact you in three years for another episode and you don't respond, then.
Michael Duxbury [00:21:33]:
Yeah, you might get response from our AI agent.
Richard [00:21:35]:
Yeah, yeah, Michael is busy now. Oh, sorry.
Michael Duxbury [00:21:41]:
Maybe I'll be busy on the beach because he's doing everything for me.
Richard [00:21:44]:
Yeah, I think we're on the same page with that. Yeah, I think there's, you know, it's obviously using it, but also some very smart human beings that need to have input on a lot of things rather than just leaving it.
Michael Duxbury [00:21:58]:
So fundamentally, AI is a tool. It's a technology and it is a tool. And like all technology and all tools as a race, if that doesn't sound too big, we need to decide how we want to use those tools to the best effect. And I think augmenting what we do and making what we do better is.
Richard [00:22:20]:
And as you say, those merchants that may be this slightly smaller side of the merchants, you know, that are sort of maybe having challenges to get something off the ground and so forth, it's a massive enabler, isn't it, to help them, you know, for minimal or no cost if they're already subscribing sort of thing. But it's not using and not, not, not implementing.
Michael Duxbury [00:22:42]:
Yeah, maybe not knowing what to do. And again, it can, essentially, I can act as a little bit of a tutor and say these are, these are things you should be doing both, you know, generically, like, this is the basics. You should be doing this. But also, I understand your business. I look, I'm looking at your data and here are some bespoke things or custom things that I think you should be doing and acting as a teacher. Yeah. Oh, great.
Richard [00:23:06]:
So segmentation, I think that's a topic we don't do. I don't think we've done an email, email episode without talking about segmentation. You know, how can businesses use AI to really, really sort of step change their segmentation?
Michael Duxbury [00:23:19]:
Yeah, it's a really interesting question because segmentation as a science art form, maybe a bit of both, has been around for a long, long time. Right. You're not going to buy a tool like ours that doesn't allow you to do it. And AI, you can do two things. I think AI can help you do what you've been doing, but it can also help you do something that you've never done before. So in terms of segmenting on, if you think about what you do now, you segment on the data that you have on someone and you segment on the behaviors that you know about that person. So pages that they browse on your website, pieces of content that they've interacted with or not interacted with products they bought, what have they done in the past? AI can use that data and it can probably maximise the usefulness of it, and it might be able to find very valuable niches within that data for you to create campaigns and target. But where it gets really interesting, in my opinion, is that you can do all of those things, but it can also start predicting the future, essentially, which has been really hard to do before.
Michael Duxbury [00:24:20]:
So it doesn't just look at what's happened in the past. It can say, okay, well, based on what I know about this person and based on what I know about other people that have previously looked like this person and what they have gone on to do and what others have gone on to do, then actually I predict that this individual person is probably going to be requiring this type of product now or will be in a week, or they may have a very high chance of purchasing this product over here. I think that's where AI is going to come into its own.
Richard [00:24:50]:
That's very good, isn't it? I think that's the, that's very exciting, that sort of prediction piece. You know, there's, there's tools that have done it in the past in a way, but not, not, not in that
Michael Duxbury [00:25:02]:
depth sort of thing. All the breadth. Yeah, exactly. Just the sheer scale of it. And the data it can use is far more. Because you're right, I mean, go back five years or so and five, six, seven years maybe. I mean, want to give you an idea. We launched product recommendations back in 2018.
Michael Duxbury [00:25:19]:
I think it was about 8 years ago now. But it wasn't using any of the current generation AI models. It was using homegrown, homespun machine learning models. And in some situations we still use those models because they do still outperform, because they're very neat. But where the new AI models really excel is. Yeah, as you say, just the depth and the breadth of the data that they can use and the types of predictions they can make.
Richard [00:25:43]:
Very, very exciting. So obviously, as a, as a, as a brand and obviously somebody that's been in the, the trenches for a long time, you know, in, in email, obviously work with a lot, a lot of brands.
Michael Duxbury [00:25:55]:
You're making me feel very old, Richard.
Richard [00:25:57]:
You know, sorry, I won't half on about that. I feel like, you know, we, we're in a similar, you know, we've been doing this a while, both Both of
Michael Duxbury [00:26:03]:
us, of course, both of us.
Richard [00:26:05]:
England out of school, straight into.
Michael Duxbury [00:26:10]:
Obviously you've, you've come across a lot
Richard [00:26:12]:
of the, the other tech that's out there and the, and the different tools and tool sets and so forth. I think obviously you know, your, your own dot digital platform and so forth. You know, fabulous platform. You know, we, we've got a lot of clients on it and so forth. But any other tech that you'll recommend to our listeners to sort of explore around email marketing? Or is. Is it just we only need you only one tool? You know, is there other things that
Michael Duxbury [00:26:36]:
I think you don't just need one at all? Mostly because we sell more than one tool. No. And I could be cheaply and I could recommend Alia. So we recently acquired Aria, which is an audience. Maybe we'll come on to that bit. So I resist the urge to recommend that. There are so many tools and I think even the classics that are still valuable. Google Analytics for example.
Michael Duxbury [00:27:02]:
Yeah, every so often we speak to a customer and they're not using Google Analytics or you know, a customer success rep speaks to them and they're not using Google Analytics. So step number one is like, oh, don't worry about digital for a second. Let's get yourself set up with Google Analytics because it does a great job of marrying.
Richard [00:27:16]:
It is amazing, isn't it, that people still don't. Or it's even more amazing when you log into their digital account and they've been paying you guys for two years and they haven't logged in for two years.
Michael Duxbury [00:27:26]:
Yeah, well, yeah, we work hard to make sure that doesn't happen.
Richard [00:27:29]:
But that doesn't happen. But you know, it helps it with, with platforms. It can happen. It does happen. We know, we see it. You know, we see it. They've sort of go in and set things up and then this. But that's, you know, no, no fault of he bought it.
Richard [00:27:42]:
It's just, you know, we're all guilty of certain subscriptions but it's. Yeah, yeah. Oh, let's get this.
Michael Duxbury [00:27:47]:
It tends not to happen with your, with your core things like you know, your e commerce system or your, your cms if you're using your CRM or your. For your marketing automation platform, they tend to be the ones that people log into all the time, of course. But yeah, Google Analytics has always done and continues to do a really good job of joining all your data together. So you can see, see trends of people moving from your website and obviously you can track emails on it. Not everyone knows that you can track emails and link them back to overarching campaigns within Google Analytics. That's super powerful. There are also of course many, many tools that can, that run adjacent to the E commerce functionality. So for example, if you are, if you're in a lot of meetings and you are taking notes and actions out of the meetings, there's lots of tools that automate that now.
Michael Duxbury [00:28:37]:
So Granola is a great one. For example, it runs in the background, it's really innovative and how it works, it doesn't join your Zoom calls as a guest, which I always think is a bit, it's a bit weird because
Richard [00:28:49]:
meeting AIs, they do people nowadays.
Michael Duxbury [00:28:51]:
They do, yeah, yeah, yeah. And we've probably all, all seen the memes, right. When someone joins a meeting and they're the only human there and the rest of the Zoom call is AI bots. But what Granola does is it just interfaces with your audio directly. So it's just always listening to what's going on. Making sense of what you said, writing up notes for you, identifying actions in them to you.
Richard [00:29:11]:
I've actually not heard of that actually. That's. That specific one. Yeah, that's good.
Michael Duxbury [00:29:15]:
It's a great example again of AI working in the background as an assistant for you and just kind of not getting in the way but helping you and letting you concentrate on the things that you want to be concentrating on. She's not writing up meeting notes. Let's face it, I'm not affiliated with them. Go and check it out. Yeah, okay.
Richard [00:29:32]:
Any others?
Michael Duxbury [00:29:34]:
You know, there's so many. As I say, let's talk about Alia for a second because ALIA is, we should never ignore top of funnel when you know you're doing, when you're doing digital marketing, particularly audience growth that comes from a tool like aria, which is essentially it's asking for pop ups, but in a really intelligent AI powered way is growing your list with, well, first party data, but essentially zero party data. So data is being given to you as a brand, really express permission to be, to be used what you want to use it for, which is the most valuable type of data. It's the data that you really, really want. So if you're not using ARIA or at all alia, then you really should be. Where ALIA comes into its own is again it comes back to AI working as an assistant. It can set up split tests on your behalf for you, it can monitor results for you, it can make changes to those split tests. So it can experiment on its own in the background.
Michael Duxbury [00:30:37]:
And report back to you on the results. Once you've got a pop up live, it's not static, it doesn't just live there forever until you remember it and go, oh, I should probably update this to capture this extra information. Let's start capturing mobile numbers now because I've distorted it. It's doing that for you. It's running in the background and it's experimenting with things like that and giving
Richard [00:30:56]:
you valuable but smart triggering that you'd refer to.
Michael Duxbury [00:31:00]:
Yeah, smart, yeah, exactly that.
Richard [00:31:02]:
Yeah, yeah, yeah. I think it still amazes me the sort of, lack of, lack of, sort of forms and pop ups etc, you know, it's a, it's obviously if you're building the list and. But you're building a list then with the intent side of it, with the
Michael Duxbury [00:31:15]:
AI and the experimentation with it, you've got to experiment. I mean having a pop up that just says give us the U managers 10% off, you'll get some people do that. You will and it will be better than nothing, don't get me wrong. But different things work better for different brands depending on where you are. So again with like Alia does a really good job. You could, you could gamify the experience, for example. Yeah, the spin to win. You can do other things.
Richard [00:31:39]:
Spin to win had a very popular couple of years, isn't it? I think. Spin to wins.
Michael Duxbury [00:31:43]:
Yeah, exactly, yeah. And there are others and you know what techniques like that or tactics like that work for some brands and not
Richard [00:31:50]:
as the one or two specific ones that you've seen in maybe specific niches. You know, we obviously, we've got a lot of niches. We cover off any specific. Obviously you've got the spin to win, it's been very, very popular. And then you've got, you know, obviously different types of pop ups that. Different. Any sort of, you know what I've
Michael Duxbury [00:32:07]:
seen, it's actually really smart. Again, this doesn't work for every brand. It depends on what what you're selling is not just to ask for essentially pii, which is what's your name, what's your email, what's your mum again? Give them to me, give them to me, give them to me. And asking, asking all the time is actually to ask slightly different questions. Ask about them. So for example, skincare brands sometimes do this. They ask about you as a person, as an individual and what is your skin type like? Yeah, what, yeah, what are you trying to sell? What do you want? Exactly that. Yeah, exactly that.
Michael Duxbury [00:32:43]:
So it means that they are. It feels a bit More personal, but it, and it is literally more personal because you can then tailor your responses based on that information. So if someone has oily skin, you don't make a product recommendation that says, hey, here's a product for dry skin. And so it works top of funnel, but it flows all the way through. Yeah.
Richard [00:33:04]:
And then in terms of segmenting those, those signups is obviously very, very powerful, isn't it? Straight away, you're not sending them like you just said.
Michael Duxbury [00:33:11]:
You know, I always think that when, when you're doing a tactic like that, it has to work for you as a brand, but you have to think about the consumer as well. They are on the other side of this. So the best things work for both. They work for the brand and they work for the consumer. And that's a really good example of that happening there. It's just the brand gets what they want, but also the consumer gets a better experience, fundamentally. Yeah.
Richard [00:33:33]:
So I think if you're still with us and you're thinking, you know what, our pop ups, we've had that spin the wheel on the website for three years. We need, you need to have a chat with the guys at Alia. Yeah, Experiment, experiment, experiment, Mix it up. And you know, that sort of smart triggering that will change.
Michael Duxbury [00:33:50]:
Yeah.
Richard [00:33:51]:
So something I'm, I'm interested in personally is WhatsApp and WhatsApp marketing seems to have really, you know, really taken a good, you know, real sort of slice of the market, you know, what sort of things are you seeing there and any sort of, any sort of tips to our listeners around what to, you know, how to get the most and best out of WhatsApp marketing.
Michael Duxbury [00:34:12]:
Yeah, WhatsApp is a really interesting channel. It is young. So in a sense actually if you go now, you'd probably still be benefiting from first move advantage, to be honest. Because, you know, WhatsApp is not a busy place when it comes to brands for most consumers. But the biggest tip is probably just to just get started. It's easy to overthink these things. Sometimes we talk about AI or not so easy to overthink and just get started, for example. And WhatsApp can be very cost effective.
Michael Duxbury [00:34:40]:
It's a slightly unusual bidding model. We don't, you know, pay for conversation, etc. But it can work really well. So for example, you know, if you are collecting email and mobile number when people sign up, comes back to experimenting, experiment with sending a welcome WhatsApp rather than an email or both. Experiment with including their incentive, whether it's a 10% off or whether it's something else via WhatsApp.
Richard [00:35:02]:
On WhatsApp.
Michael Duxbury [00:35:03]:
On WhatsApp. Otherwise we see it working really well for quite targeted campaigns. So not. You probably don't want to be sending WhatsApp every other day like you might do with an email. But if you have a new collection being launched, for example, or you run a VIP club, this is where this is particularly powerful. You can communicate with your BMP club over WhatsApp. Bill's a bit more special at the moment. It gives them access, come behind the scenes access to a sale, 24 hours and, you know, all that kind of stuff.
Michael Duxbury [00:35:37]:
When we see brands do that, that's where we can see engagement really, really peaked. So think about what experiment, don't overthink it, but also think about what you're doing. That. What is, what could we do that's a little bit different? What, what do we do that's going to be high value to a consumer?
Richard [00:35:51]:
So maybe send that. Cool things you're doing on email, you're just not doing on WhatsApp. But you're probably going to get better conversion on WhatsApp. Well, you're going to get better opens for sure.
Michael Duxbury [00:36:01]:
Yeah, yeah, yeah, yeah, yeah.
Richard [00:36:02]:
Why?
Michael Duxbury [00:36:02]:
I mean, some people see it's not just opens, it is click through rates all the way through to RI on products, depending what they do. Really some really good results on WhatsApp.
Richard [00:36:12]:
Yeah. Something we need to do more of. You know, as an agency, I know we've got some clients using it, but some are a little bit less open to using it, so. Yeah, yeah.
Michael Duxbury [00:36:22]:
Again, advantage.
Richard [00:36:24]:
Yeah. Well, that's the AI thing, isn't it? Yeah, absolutely, absolutely. So we've covered quite a lot of ground there. You know, we've talked obviously a lot about AI and where, where that's going and where it is and, you know, sort of testing and trying and trialing things and don't be afraid and you know, and especially those guys that maybe aren't doing a lot with their email. You know, it's going to give you a real boost, get you up and running, but ultimately, obviously it can do a lot more than that. Um, but you know, where do you see is, is there anything you could share maybe from that, you know, that you're really excited about over the sort of roadmap over the next sort of 12 months that maybe Dot Digital's got coming, coming through or some things that you're aware of that you're really excited about?
Michael Duxbury [00:37:04]:
Yeah, so we, we've got a bunch of Teams that are working exclusively on AI and AI experimentation. And of course the, you know, the best thing about my, one of the best things about my job is I get to see all the crazy things that they're doing and come up with ideas as well. But we, we do this fully in the knowledge that not all of those things will work out and not all of those things will, we will bring to market. Right. So just like I'm saying, hey, if you own an E commerce brand, you need to experiment. We're experimenting too and we have had things that we thought, you know what? It's not, it's not quite right. It's not, it's not there yet. You know, maybe the technology is not quite there.
Richard [00:37:39]:
I can hold my hand up.
Michael Duxbury [00:37:42]:
Yeah, it's what it's all about. There are other things that we were slightly less sure about but actually have proved out, proven to be not very useful. But also tech up has been huge. So an example of that recently is that we extended, I talked already about how the generative AI side of the platform is, is the most used for us. It's gone up about 180% year, one year growth. But it had, it's essentially given very similar content to all of our customers. Not always. So for example, if you did a subject line, it is also always looked at your previous subject lines and worked out what worked and what hasn't worked and adapted.
Michael Duxbury [00:38:23]:
But otherwise it was sort of using the same technology for everyone. What we launched last month was what we're calling tone of voice. So you can now when you're onboarding, you can give digital your, your, your brand guidelines or your tone of voice types and you may, even if you're a big brand, you may have a giant PDF with everything written down to the nth degree and you can just feed that in and it learns from it. Or you may be a much smaller brand and you don't have that at all. But you do have a website and you, you know, you may have paid someone to write your copy for. You may have worked on it yourself, you know, overnight getting it just right. And you can learn from that. You can essentially just give it anything that you have that demonstrates what your brand sounds like.
Michael Duxbury [00:39:05]:
That means that everything that it generates from now on, whether it's a subject line or it's rewriting an email or generating an SMS like common email as well. So everything that creating uses that kind of like that, you don't have to worry about it, you just use it, use it in the background and We've seen extraordinary take up of that in just. Yeah, because I think just show brands are really like, we love the tool but it has to be, it has to work for us. Yeah, yeah, we're seeing that a lot.
Richard [00:39:30]:
It's trained on their tone, it's trained on that. It's very specific brand, similar to what we do in SEO to be fair. You know, we create products in different tools and we train. It
Michael Duxbury [00:39:41]:
can come up with very different outputs depending on the brand. And that'll be important. Yeah, yeah. Well otherwise what we're working on now, we are very much experimenting with agents and we're quite old fashioned maybe and that we don't want to, we don't want to release stuff to market unless we know it's bright. So we are, we're still working on our agents but we're working on different ones. We're taking, you know, for example, there will be a segmentation agent which does a lot of what we've been talking about, which is really exciting. So it takes your data. Yes.
Michael Duxbury [00:40:11]:
But it also predicts the future. You can prompt it and say actually and I give me quite specifically, you can say give me everyone that's bought from our Lincoln branch and then ask through Reese and it and just go off and do it. Or you can be a lot more like give me anyone that is showing a propensity to buy this type of product in the next three months and it will go away and do that as well. So one's rearward looking, one's forward looking. Yeah, I mentioned it already. We're looking at content creation agents either in use in dot digital or via an MCP server. So via your tool of choice. And the third one we're looking at is an analytics agent and a reporting agent.
Michael Duxbury [00:40:53]:
So again, works in two different ways. One is like you can prompt it to say what was the open rate of my last campaign? And it will give you that easy. But you can also get it to work in the background and make suggestions and say, so you can log in one day and it'll go, hey, we've noticed that. Or I've noticed that actually your performance is dipped in the last six months and we think it's because of this. Therefore we recommend you try this. And then ultimately would you like us to do that and then show you set that up and would you like create this? Yeah, yeah, yeah. And so the big question then is you've got these agents and that's all really useful. Do they stay as separate agents? Do we end up with one big agent is doing work on your behalf.
Michael Duxbury [00:41:36]:
Again, predicting the future. It's very hard. So we, it's a, yeah, we will wait and see a Where the technology leads us, but also where our users lead us.
Richard [00:41:45]:
Yeah, very exciting, isn't it? It's literally, I think the most exciting time I've had or been in involved, you know, just, you know, what you're saying there just seems very similar to what we're doing here. But obviously from an SEO performance perspective, SEO, ppc, so forth, you know, building various agents, you know, and then the obviously test, test, test, test, test, mcp, you know, it is, it's the Wilbur in, isn't it? It's the, it's the whole new language that's just appeared in the last 12 months.
Michael Duxbury [00:42:17]:
Exciting because the possibility as a cliche, but the possibilities are essentially endless. And if you look forward even more, we're building emails already. We are seeing, for example, Apple email summaries. We're essentially seeing AI reading an email and summarising it for you. So we could be in a position whereby AI is creating an email. AI is involved in sending the email and identifying the audience with that email. And then AI is also reading the emails. And at the moment a human is jumping in and reading that summary and making a decision off the back of it.
Michael Duxbury [00:42:52]:
But of course, what's the future? The future might well be an AI reads the email and then does something off the.
Richard [00:42:58]:
And then just, yeah, summarises.
Michael Duxbury [00:43:00]:
And you might be able to say, I'm really interested in this pair of shoes. Keep an eye out for when a 10 offer comes in and it, when it reads that there's a 10 offer on an email, it'll go, head straight away and purchase those shoes.
Richard [00:43:10]:
Yeah, I mean they sort of do that now. Not, I'm not referring to email, but you can do that with, you know, with Record and ChatGPT, you know. Right. Look out for this. Tell me when this is this and yeah, exactly.
Michael Duxbury [00:43:21]:
So you know what? As marketers, it may well be that the emails we send become a core part of that. I mean, we may need to change strategies off the back of it. Sometimes we may want that to happen, sometimes we may not want it to happen. We need to, we all need to learn how we cope with that.
Richard [00:43:36]:
Bottom line is email CRM marketing is going through, you know, a transition, I guess is maybe not quite the right word, but it's going through a very, very exciting time. And I think, you know, those that are listening now, you know, it's you know, really, really embrace your CRM, your data, you know, and obviously there's some, there's some tools available now that are going to make things easier, a bit easier. But also, you know, they're sort of helping to make stores more profitable because they're, they're, they're sort of opening up, you know, the way we talked about segmentation, you know, it's better now. You know, the ability to do more things, the ability to integrate more channels,
Michael Duxbury [00:44:14]:
it can make marketers lives better by allowing them to concentrate on the things that they want to concentrate more on as well. We shouldn't, we shouldn't lose sight of that.
Richard [00:44:22]:
Yeah, totally. Well, Michael, it's been an absolute pleasure having you on.
Michael Duxbury [00:44:27]:
Thank you very much.
Richard [00:44:28]:
I like to finish every episode with a couple of questions around. Do you have a book that you recommend to our listeners?
Michael Duxbury [00:44:34]:
Yeah, I may have gone slightly rogue here because I've gone for fiction.
Richard [00:44:39]:
I think that's fine. That's fine.
Michael Duxbury [00:44:40]:
The book I'm going to mention is not actually the book I'm going to recommend. I'll come on to that. But. Okay. The talk of AI and technology and what I'm seeing happen and some of the fears around makes me think of a John Steinbeck novel, which is the Great Rock, which is ostensibly around the Great Depression, but it is also essentially about a book about technology. Right. It's the tractors and farming rather than, than anything else. But it's essentially a book around technology.
Michael Duxbury [00:45:10]:
But I thought that might be a bit close to the bone, so I picked another John Steinbeck novel which is Cannery Row. So I think if you want something at the end of the day to turn off, read that. It's one of the most probably evocative books I've ever read.
Richard [00:45:25]:
Just say the name again, please, Michael.
Michael Duxbury [00:45:27]:
Cannery Row is the name by John Steinbeck. And it's essentially about, it's about a cannery. I think it's a sardine cannery. Um, it sounds odd and it is all set there. It's just about the people that are intrigued already. Yeah, there you go. But it says a very evocative book. If you want something to switch off, don't be.
Michael Duxbury [00:45:46]:
Switch off. That's my recommendation.
Richard [00:45:47]:
Yeah, no, I will, I will link that up in the show notes. I will get that ordered. I, I'm always intrigued when we get thrown a sort of a booklet is. Yeah. That I've not heard of. Or not. Not, not. We've done about 250 book recommendations now, so.
Richard [00:46:01]:
So that's definitely a new one on me. Brilliant. Brilliant.
Michael Duxbury [00:46:04]:
They all listed somewhere.
Richard [00:46:06]:
They're on the different episodes. Yeah. Each episode's got them listed.
Michael Duxbury [00:46:09]:
Yeah.
Richard [00:46:09]:
Yeah. We should do a one pager with them all on, to be fair. I'll get somebody. I'll get. I'll get my AI assistant to do that. I'll think about. Somebody might slap me. Well, thanks for coming on the show.
Richard [00:46:22]:
For those that want to sort of find out more about digital, more about yourself, Michael, what's the best way to do that?
Michael Duxbury [00:46:28]:
Ask your AI tool of choice. I'm sure it will tell you otherwise. You can go to dot digital.com or Michael Duxby on LinkedIn. It's probably the best place.
Richard [00:46:35]:
Brilliant. Yeah, brilliant. Well, thanks for coming on the show. I really appreciate it.
Michael Duxbury [00:46:38]:
Thank you very much. It's been a pleasure.
Richard [00:46:44]:
If you enjoy this episode, hit the subscribe or Follow button. Wherever you are listening to this podcast, you're always the first to know when a new episode is released. Have a fantastic day and I'll see you on the next one.
Sign up to receive email updates
Enter your name and email address below and I'll send you periodic updates about the podcast.