The Bootstrapped Founder | 256: Brennan Dunn

This is a masterclass in personalized marketing. Brennan even convinces me to ramp up my own efforts!

Listen in as I uncover the power of personalized marketing with Brennan Dunn, author of 'This is Personal.' Brennan offers valuable insights into how indie hackers and solopreneurs can best learn about their customers and utilize that knowledge for revenue generation. He explores why personalization works, how audience segmentation should be done, and where data collection should start, providing you with a wealth of knowledge to transform your marketing strategies.

Brennan on Twitter: https://twitter.com/brennandunn

00:00:00 Personalized Marketing for Indie Hackers
00:05:02 Segmenting Audience for Better Marketing
00:10:19 Importance of Personalized Questions in Data Collection
00:20:35 Segmentation and Survey Funnels in Marketing
00:32:10 AI's Role in Personalized Marketing Future
00:37:59 ChatGPT for Data Analysis and Writing
00:47:02 Personalization and E-Commerce Recommendations
00:52:04 Personalized Book Recommendations for Marketing

The blog post: https://thebootstrappedfounder.com/brennan-dunn-this-is-personal/

The podcast episode: https://share.transistor.fm/s/3f36312e

The video: https://youtu.be/oPNhJvcvb0A

You'll find my weekly article on my blog: https://thebootstrappedfounder.com

Podcast: https://thebootstrappedfounder.com/podcast

Newsletter: https://thebootstrappedfounder.com/newsletter

My book Zero to Sold: https://zerotosold.com/

My book The Embedded Entrepreneur: https://embeddedentrepreneur.com/

My course Find Your Following: https://findyourfollowing.com

This episode is sponsored by Acquire.com

(00:00) - Personalized Marketing for Indie Hackers
(05:02) - Segmenting Audience for Better Marketing
(10:19) - Importance of Personalized Questions in Data Collection
(20:35) - Segmentation and Survey Funnels in Marketing
(32:10) - AI's Role in Personalized Marketing Future
(37:59) - ChatGPT for Data Analysis and Writing
(47:02) - Personalization and E-Commerce Recommendations
(52:04) - Personalized Book Recommendations for Marketing

Creators & Guests

Host

Arvid Kahl

Empowering founders with kindness. Building in Public. Sold my SaaS FeedbackPanda for life-changing $ in 2019, now sharing my journey & what I learned.

Guest

Brennan Dunn

Writes 2x weekly at https://t.co/FtjqZcn9SJ, working on https://t.co/wtdPTFIdeO, co-founder of https://t.co/YbxEfQUz0y, smitten with @lauraelizdunn

What is The Bootstrapped Founder?

Arvid Kahl talks about starting and bootstrapping businesses, how to build an audience, and how to build in public.

Arvid Kahl
Do you know your customers? Do you know what they need and when they need it and how you can offer every single one of them exactly the right product at the right time? Well, Brennan Dunn knows. And he's going to teach you just how to do that. Hey, I'm Arvid and welcome to The Bootstrapped Founder. Today, I talk to Brennan about personalization and how to best learn the most you possibly can about your present and future customers. And then how to make money from that, as well. This episode is sponsored by acquire.com. More on that later. Now, here's Brennan.

Brennan, welcome back to the show. Last time, we were chatting a lot about email and automation at scale. And today, it's getting personal.

Brennan Dunn
Yeah

Arvid Kahl
You've written a book called This Is Personal. And I'm super, super happy that you did because it's exciting. And it's all about personalized marketing. How has the author journey of this book been for you?

Brennan Dunn
It's been very different from what I'm used to. So like you, I'm kind of my background is in software and kind of like digital products and everything. And I was telling you before we hit record, I started this process really in 2018. It's 2023. It's finally in three weeks gonna be like an actual product. So I'm very used to like, iterative building and kind of this idea that you never actually finished anything. It's like even with a course, you could still add content, you can revise things. This whole idea of like, it's a bit like I guess, like making a movie where once it's done, it's done. So yeah, it's been very, very, a lot of new challenges, a lot of at times frustration with how slow the whole industry is. But overall, I'm glad that they're kind of the proverbial finish line. So yeah, it's been very interesting, so to speak.

Arvid Kahl
It's quite different from the creator life, right? Or somebody, a marketer life where we have control over all the steps leading up to the result, feels like publishing is this kind of thing. You have to at some point, just get to somebody else and hope for the best. Right?

Brennan Dunn
Well, it's that and it's also it's such a like, I don't know if polish is the right word, but like such a process driven, intentional thing that I've built, but it's my cheapest product. To me, it's just wild like. Yeah, it's been quite the experience. I'm glad I've done it. We'll see if it happens again. But, you know, at least now I can tick off the bucket list, like done with the publish book thing, so.

Arvid Kahl
Yeah, that's awesome. I'm so happy for you. I mean, I've been following this journey, a few writing the whole thing and looking kind of behind the scenes every now and then. And it's really, really enjoyable just to also see how the process works because it is quite different I feel from the regular creator process where you just like experiment all the time and you try different things and you do like AB testing modulations and that kind of stuff. And you run those experiments over time, you get the data, you work in backend. The feedback in what we do with the internet, right? The digital kind of marketing, it's almost immediately, or at least you can get immediate data. And you can get data series over time and kind of compare them. With books, it's kind of fire and forget and hope for the best, right? And the audience is also so much more diverse, like your book probably is going to end up in like airport stores and bookstores all over the world. And I hope certainly to find

Brennan Dunn
I'm not sure about airport stores, but hopefully

Arvid Kahl
Well, they'll have to be in airports. If I have to take a copy and take it to an airport, I would make that happen. But you know, it's gonna be just exposed to all different kinds of people. You have no control over who it is, which is kind of the exact opposite in a way as a medium to personalized marketing. Right? I kind of want to go into the book because I think the topic is spectacular. It's something that I personally struggle with a lot because I have no educational background in marketing, whatsoever. So that like regular traditional marketing or fancy newfangled, you know, relational marketing or whatever you might call it. To me, all of this is just highly experiential. And it's nice to see an expert like you put all their insights and knowledge into something I can hopefully soon find at an airport store. So you know, let's get into personalized marketing. And the question I kind of want to lead into with is because I'm an indie hacker, I come from a software driven background. And I, personally also am a solopreneur. So I do all the things, right? Marketing is but one of the many things that I do. How important is it for me as somebody who's just myself doing all the things to think about personalization in my marketing, my outreach, my emails and that kind of stuff?

Brennan Dunn
Yeah, so a common question is, I think a lot of people compare personalized marketing to something like A/B testing, where it's one of those things that work at scale. But when you're just starting out or you're dealing with a limited set of customers or a smaller audience, it doesn't make as much sense. I mean, my thinking here is so with indie hacking, so like, let's say, let's pick SaaS, right? Let's pick on SaaS. So early on the idea is, before you spend a ton of time on a sales page or marketing website, you should get on calls and do demos and you know, so on and so forth. And that's kind of the conventional wisdom, right? Like, you should learn as much as you can from your audience. And then, hopefully, individually, the kind of like, convince them to buy your thing, figure out what objections they had and all that good stuff. So that's sales. I mean, that's sales one on one. It's very one on one, it's low scale, like you can only book out an hour at a time. And personalized marketing, the way I look at it is not things like putting people's names on an email or something like that. Instead, it's trying to think, okay, what are we doing when we do like a sales demo of a product? We are, I'm listening to who you are, you're technical, you're in SaaS, you're a German living in Canada. I can talk to you about like, you know, I mean, that's more of an extreme example but like, I have an idea of who you are. And then I'm going to talk about different benefits that I think you would care about. And so I think, ultimately, we should all be thinking, our audience is probably not a giant monolith, like there are different types of people in it. Some care about certain things we do, feature wise. Some care about other features, benefits. They want to hear different case studies, examples and so on. So I think ultimately, it's a thing we should all be striving towards. But early on, I don't think you should drop everything and be like hacking away at writing conditional content in your sales emails or anything like that. What I do think, though, that people should do sooner than later because it's harder to do, once you already have an audience is to have some go beyond just first names and email addresses. So most of us have like an email address opt in form and we get their first name, we get their email address and that's all we know. And then we'll tag them or something that they buy from us. But I do think it's really important to try to uncover at a macro level, the different identities needs of your audience. So it could be as simple as you know, with Create and Sell, which is my email list, when they join, hey, what's your focus now? Is it on building an audience? Is it on selling to your list? Or is it on automation? And then from there, I ask questions about like, how experienced are you with your email platform? What do you as an email platform and so on. So I think even if I don't do anything with that data, I think it's important to have that information. So you can see, like, what kind of content should I be, you know, focused on. It kind of lets you do, really just kind of market research on your own audience and get kind of percentage based numbers that you can then use to make decisions about acquisition, product creation, what newsletter content you should write about or whatever else. So I like doing that early. Like I started doing that when I did create and sell and launched it to the world at first. Immediately, I was segmenting, but I wasn't doing anything with that data. So I do think it's important to kind of enrich it to the best of your ability, your early audience with who they are and why they're following you.

Arvid Kahl
So I'm just gonna throw in a couple of thoughts at you at this point as a solopreneur, who wants to do this, but has a couple fears. Because we tend to fear reactions of people that we interact with, right? From a developer background, I very much struggle with talking to people in any case, right? Or interacting with people in any case. So my immediate fear is, am I asking them too much to start segmenting themselves at that point? Because I haven't built a relationship with them just yet when they sign up, right? Is that the right point? What can you say for that?

Brennan Dunn
So the good thing is we because the primary use case at this point for my software company, RightMessage is to do post opt in subscriber surveys. I have a lot of data on like, not conversion rates, but engagement rates of the surveys. And typically speaking, there are some surveys that do really well. And some that don't do well. And the common theme between the ones that do well and don't or the common theme, I should say about the ones that do well, are really a few things. So the first thing is that they make it clear this is not just like market research data that we're using to put on a PowerPoint somewhere that we're going to show to a board. We're getting this information or I'm getting this information so I can better serve you. So if I know that you use ConvertKit, I'm going to talk to you about my ConvertKit course probably but if you use ActiveCampaign I'm not, right? Like so you make it very clear. And usually it's a one sentence thing that says hey, thanks for joining. Do you mind taking 15 seconds 20 seconds, make it very low work on their part, sharing a bit about why you're here so I can better best get you what you need. Because I think there's a lot to be said about people ultimately want what they care about, right? So even if you're not yet sending hyper personalized like these people get this newsletter because they fit this profile and these people get that newsletter. Even if you're not doing that yet, the fact that you're saying I want to be able to better serve you that as well. So across the board, we're seeing between an 80 to 85% response rate. So completion rate of the surveys for new subscribers when doing that. And the other thing that people do well, I think to get those high rates is they don't make the questions hard. I think the mistake people make is it's like, a three sentence question. And every answer is like another two sentences. And it's like the mental overwhelm of trying to figure out like, well, which one is it? Where instead of something like what Jay Clouse does, where it's like, hey, so we're gonna best give you what you need. Are you on LinkedIn? Yes, no, cool. Are you on Twitter? Yes, no. It makes it very, almost conversational. Right? So it's very easy to answer. No one really needs to think it's not like, how would you solve, you know, world hunger. It's very specific, choose one of the following options. And he's getting a ton of data about that about who his audience is. And it helps him better understand not only who they are, but what products he should be offering them. So, you know, if you like, with Justin Walsh, if you don't use LinkedIn, LinkedIn OS probably doesn't make sense for you to use, or at least not yet. So yeah, I think as long as you just make it, you know, the questions need to be easy. And you need to make it clear that this is for them.

Arvid Kahl
So you don't really start asking why questions or you don't want to prompt them to actually go deep into something. You just want to kind of allow them to self categorize into binary or like multiple choice situations?

Brennan Dunn
Yeah, I like one of many, especially for personalization. It's hard to do. If you have an open ended text input, you can't really I mean, I'm sure you could pipe it up to ChatGPT and like, infer things. Great. But for the most part, it's more work to fill in a textfield than it is to click a button. And on top of that, if it's like which of the following do you like red, green, or blue? They choose red, it's very easy for you to then say, hey, if you chose red make the background red or something. So I just, I prefer one of many. Yeah

Arvid Kahl
Okay. Yeah. That makes sense to me. It feels like again, easy, right? If you want it easy for people to be able to quickly do it and to get to where they want, which is you knowing what they want, them knowing what you have to offer and kind of combining this into a successful transaction. How much is asking too much? I always wonder with personalization, like, can you over personalize? And I guess like, if I have 100 people on my email list and I send 100 different emails that are all personalized to their use case, then I have kind of 100 pristine and handcrafted emails. That's a bit much but even just in collecting data, how much data do you actually want to collect before you kind of saturate the willingness to respond to data?

Brennan Dunn
Yeah, I mean, so the good thing is, if you do make your questions easy, there tends to be at 98 to 99% rate of going to the next question, right? So if they answer the first question, if you make the question easy to do and as long as you're making it, again, you're setting the expectation that this is going to take like 20 seconds, then the response rates tend to be almost like everyone who starts will usually finish. But yeah, I mean, in terms of like, there's two things, right? Like, you don't want to ask too many questions because then you're, I mean, unless there's really something in it for them. So if you position it as like a personality quiz or something like that, sure. They get something at the end that says you are, you know, house gryffindor or something right? Like, then that's the win that they want. But, yeah, I tend to keep it to at a minimum. You should be uncovering two things. One of which is why they're currently interested in what you have to offer. So in my case, that's what is your like what are you looking for audience growth or sales or automation? And then a who question is, so the who for me is, what email platform you use? And how would you self assess your experience with it, like how good you are with it? And that gives me a good idea of like, this is somebody who is a self identified complete beginner with ConvertKit. And they want help with audience growth. That tells me a lot about them. Whereas somebody who's pretty already advanced and they want help with automation, they could probably both benefit from me, but they're going to benefit in different ways from what I have to offer. And on top of that, I can also look at that top line data and say, well, I have a lot of people on my list who are complete beginners. I need to create more beginner type products. I don't really have anything for them. And that's actually what I'm doing exactly now because I'm looking at that data to help me fill in the blanks of my own offerings.

Arvid Kahl
That is super interesting. How do you do that? How do you figure this out? Because looking at all that data, again, the more segmentations you have, the more data you have and I think it can quite overwhelm you if you don't really know what you're looking at. So how do you personally look into this data to find these trends and then see what you can do with them?

Brennan Dunn
So I have ace up my sleeve, however you say it, that I've been using and it's been really interesting and this is a tool. I don't know if you've heard of it, called SegMetrics, but SegMetrics is it pulls in all your email contacts data and it superimposes like Stripe or whatever data on top of that. And what it does is not only gives me things like that, you can kind of get from the ConvertKit dashboard, like average subscriber value, but it also gives me average customer value, time to purchase, all that good stuff. But I can also break that down by any dimension I want. And I can see, oh, ConvertKit people for me are worth about $80 per subscriber, but somebody on ActiveCampaign is worth about, like five, right? So that tells me like either ConvertKit people have a lot more disposable income and they really like what I'm doing versus Active Campaign or I'm dropping the ball with them. So like what I've been learning is, as my list has grown and become less and less CovertKit-y but my core offering is a course on ConvertKit, which is a bit of an issue. So I've been looking at that hard data about people's experience levels. I ask people what they've done and haven't done when it comes to like, are you currently doing any form of segmentation? Do you want to be doing segmentation? And I asked these questions throughout kind of their experience with me. And it really made it clear that there was a high demand for some sort of training product that will help people get a basic like segmentation, one on one understanding regardless of what you do and regardless of the email platform you use. So that's actually, yeah, I mean, that's what I've been doing. Now, I'm working, no surprise, but a little mini course on exactly that. And it came from looking at that data, looking at it in SegMetrics and realizing that people who answered a certain way, were just not worth much. And it really became a question of, well, why aren't they worth much? Well, it's due to this gap in my product offerings. So I really like, you know, that kind of approach of looking at, instead of just saying, I made this much money with my audience or I have these many subscribers, being able to individually reduce it down to exact conversion rates of subscribers, exact valuations of subscribers, that is super powerful once you start doing that.

Arvid Kahl
Do you sometimes think of new potential questions or new answer options to add to your existing, you know, personalization questionnaire as a consequence of looking into this data?

Brennan Dunn
Yeah, so not for that data as much. But one thing I'll often do is I'll have like an other option. And then if you choose other, I asked you to put in your own words like what other is. So over time, you get a lot of data and then you can start to think okay, is do I need to add another option? But I'm always wary of doing that because, again, if you have 20 options, that's a bit extreme. I usually try to limit it to like three to five buttons. But I've seen over time, like how there have been holes in my own understanding of like, who my audience is, that I could then go and backfill by either removing questions that we're getting, like a 1% response rate in favor of questions that we're getting a lot of other response rates that, you know, that I then added.

Arvid Kahl
I wonder like, in this case, since you change things, does that make it harder to look at the data as a whole? Because, you know, if you start with certain kinds of questions, you get certain answers and then you add new things that kind of shifts the response pattern in there or do you just look at like certain timeframes within your data?

Brennan Dunn
Yeah, so the thing I learn with SegMetrics is I can give it certain time, time windows and look at it that way. And I also sometimes will just change the label. So case in point, like with one of my businesses, Double Your Freelancing. We were getting a lot of people choosing other and we asked people what kind of work you do. And they be like developer, designer, marketer, writer, or other. And if they chose other, they could then tell us what other was. And a lot of people would put like branding or logos and stuff. And I'm like, that's design, like, why aren't you just choosing design? So then modify the button to be design/branding and pretty much all those others went away. So that's an example of where I didn't really add something new. I just kind of rephrased one of these existing options.

Arvid Kahl
That's very interesting. Like how a little word like a little difference there can just impact the data so much. It's funny, it reminds me very much of my political science studies. We had a lot of like quantitative data analysis and all that kinds of stuff in there too. So that's kind of would you have to do if you do surveys with people around the political topics and stuff. It feels like it's the exact same math. Obviously, it is, like underneath these kind of surveys, responses and the kind of coding that you have to use, encoding and decoding of the answers into certain kinds of categories. It's very interesting to now see this thing that I never expected to ever use again, to be part of my creator journey. And one thing that stands out for me is that I did not expect to ever have to use this in my creator journey. And I think a lot of people might also not expect or even think of using that kind of deep qualitative and quantitative data analysis in their lives as a creator or as a founder, as an entrepreneur, whatever they might be. It sounds like a lot of work. Let's just say that. So it sounds like you've spent a significant amount of time on that. Is that right?

Brennan Dunn
I don't know. I mean, to be honest, like, so we've been doing recently kind of started doing consulting again with this little agency thing. And we've been doing pretty much for all of our clients setting up segmentation survey funnels. And it doesn't take that long. I mean, really, what it involves is, I think most creators have, we know like things like our list size and less growth rate and things like that. And then we kind of have a good understanding of who our audience is based on emails that came in from like, replies to our newsletter or something like that. But we don't really have like, we can't say 40% of my audience is a beginner or hasn't started a business or something like few of us can do that. So that's what the clients that have come on board with us have wanted. And really, what we've done to do that is either if the person we're working with has a good understanding, they just haven't, like captured it, then we will do like version one of the survey based on their understanding, but we've worked with some people who don't know anything. So what we've done is, we just do a simple broadcast that goes out to the list and says, hey, we want to make like, we're working on the next kind of like our future content calendar. And we want to make sure that what we're creating aligns with you and your needs. Would you mind replying to this email with one or two sentences about a bit about like, who you are and what you're looking to get from us or me? And that's the email and then you get a bunch of replies. And then you basically just normalize all that into, okay, like, these are the top like three or four things that keep coming back. I mean, you'll have outliers, but you just kind of ignore them, you just focus on what are the common things that people are saying and that becomes your V1 segmentation strategy. So I mean, yeah, that upfront work can take some time. But once it's up, it's up. I mean, I haven't really changed the Create and Sell one that I have running now. I've tweaked it a bit, but I haven't really changed it in years, but it consistently is getting 85% of all new subscribers on my list telling me or sharing with me why they're here, what ESP they use. I've been able to see the growth of Beehive not on Twitter, but by virtue of the fact that people are now selecting Beehive instead of like, ConvertKit. So like, I'm getting, you know, hard data about what I'm also seeing trend wise online.

Arvid Kahl
That's really cool. Is there like a minimum size list for this to make sense? Or should you start like from the beginning the moment you start a newsletter or a marketing list?

Brennan Dunn
I mean, I think like, so again, with our clients. So the good thing about working with the clients we're doing is these all are people, creators, like hundreds of 1000s people on their list, so we're able to get like really good statistically significant data. But without fail, they're all getting about that 80-85% new subscriber segmentation rate. But when it comes to them sending out broadcasts to their email list saying hey, like, I'd love to find out more about you. Would you mind clicking this, take this quick survey? It's super small comparatively. It's hard to get people to click links to go to surveys but it's a lot easier to say when they just gave you their email address and click submit on the next page capture a bit more data because they're pretty engaged at that point.

Arvid Kahl
That sounds to me like it does make a lot of sense to set it up very early to get that immediate engagement from the start.

Brennan Dunn
Do it early, get the data and then you can use it later on.

Arvid Kahl
Okay, nice. That's a nice idea. I still struggle with it. I do now send like a follow up email and I suggest some kind of stuff but I don't segment much because I honestly personally I don't think I've understood yet and I'm just saying this as I come to terms with my own incomplete business strategy here just why I should do this necessarily. Because I do still have sales. People find my stuff eventually. I also, I guess for my business I go in very broad in terms of I just want to have a lot of impact on a lot of people and then have word of mouth happening so I don't have direct sales much from the stuff that, I don't push my info products a lot. So I think I still need to be convinced that segmentation, effort in segmenting people is worth it for me. Can you help me with that, please?

Brennan Dunn
Yeah, okay. So I think like, I think we could probably all agree that if you were to go, if you were to niche down heavily and say, Arvid only helps people who are doing SaaS and they are pre product or something, right? Like, if you only dialed in on that and you made everything you do focused on that, the conversion rates or whatever, whatever you're tracking for people who fit that profile will be high comparatively than, you know, trying to have a wider net, right? So we all know like niching always works, right? Like, because with niching, it always works for the audience you're targeting because people want to buy into the thing or subscribe to the thing that is specific to them and their needs. So if you go too generic, you end up making putting the onus on them to kind of figure out how to apply what you're saying to fit their unique circumstance. Like even with a lot of specialty consulting stuff, a lot of the consulting stuff I've done over the years has been clients who have literally just wanted me to rehash a blog post that I know they've already read to fit them, right? And so you know that happens all the time. Like we all want the thing made just for us. So with that said, if niching works better, it doesn't always make sense for me to have like, I don't know, mastering ConvertKit for book authors, mastering ConvertKit for SaaS. Like that would be insane. So instead, the idea is what if we could then dynamically niche? So instead, it's like, they are getting a email that I send about the course. And then if they're a creator, maybe I put the quote from Dickie Bush of Ship 30 because he's a creator. And if they are a consultant, maybe I put the quote from, I've had quite a few people who have said, like the views with our clients to like make, to basically they've taken what I've taught them and build a really good little personal freelancer, agency business. So it's that kind of thing where like, the agency person is going to care more about how can this product help me get better clients maybe or help me like, sell my clients on more stuff by offering them this as a service. Whereas the, you know, the course grader doesn't really care about that, probably. So the thinking is, again, like, this is an example of just using simple if-else's to say, this person is going to see this testimonial with this case study, whereas that person is gonna see something different. I have both because both people have equally benefited from my product. And you're just basically saying exactly like, if I was trying to talk to you over the phone about buying mastering ConvertKit. I wouldn't explain, although I would just focus in on what I know about you and what you've shared with me to try to convince you to buy the thing, right? And so it's basically, I look at it as the middle ground between what most of us are doing now, which is basically online brochures where you have like your website, which is like a brochure, it's like the same thing for everyone. Everyone sees the same thing. I call it like billboard kind of marketing. It's the middle ground between that and what you would do if you picked up the phone and we're trying to convince somebody of either following your newsletter or buying your product or whatever.

Arvid Kahl
Okay, yeah. Okay, billboard makes sense. That's quite clear. That's how I feel. I do it right now. Just throw it out. And hopefully the right people to look at it at the right time, right?

Brennan Dunn
Yeah

Arvid Kahl
Instead of writing a message that reaches people at the right time for, not necessarily just in a convincing way, but in a relatable way, right? In a way that actually allows them to see oh, yeah, this is something in my wheelhouse. I know Dickie Bush, right? That kind of stuff

Brennan Dunn
Yeah

Arvid Kahl
Okay. If you rephrase it like that, it makes a lot of sense.

Brennan Dunn
Anything goes beyond that. Like, I think like, if you need to do the billboard thing, let's say, let's go back to Ship 30. Because they're actually using what I teach at Ship 30. So you have everyone from people who have never written publicly, to people who are already doing pretty well and they want better systems to do it better. Right? So you have that those are the two kind of extremes. But if you were getting their marketing stuff and they needed to say constantly, you know, regardless of if you're just starting out or are we really experience, Ship 30 for 30 blah, blah, blah, blah, blah, regardless of if you're just starting out or experienced, Ship 30. If you see that, like when you see that, if you're pretty experienced, you're going to be thinking I mean, can something that helps a complete newbie really work for me and vice versa. The same will happen for the newbie who's seeing like, big crater types who have used it or something. So it's one of those things like I think it's just, if you can say like, I know you've told me you're starting. You've told me you're a newbie, basically. Here's why this product, maybe it's the same product for everyone. Maybe, you know, here's why Ship 30 for 30, as a cohort based course, is going to equip you to start a really great ready habit and build your audience from scratch. And it's just thinking in terms that it's not about like having 50 million parallel sales pitches, it's just conditionally saying this part of the email, we might want to change up. And that's all basically we are going to do.

Arvid Kahl
Yeah, yeah, it makes sense. Like, if you say, this is for everyone, that's just general startup advice, right? If you build for everybody, you build for nobody. Because nobody is everybody at the same time. Everybody's somewhere on the spectrum. And that makes a lot of sense. I wouldn't want a product that is both great for the expert and great for the beginner because then I know it's somewhere. It's trying to do everything. It's the thing that doesn't speak to me. But if you can angle it, if you can look at it from a certain perspective, this part of it is great for beginners. There's another part not for you, not even gonna mention it, right? I'll mention it to you if you are an expert or when you become one. That is interesting. My question that comes immediately to my mind is, how can we make sure that this insight into the psychology maybe even of our potential customer is not abused to a point where they feel that the thing we're writing is almost creepy or intrusive, like we are over personalizing it to them? Like we know too much about them and it kind of shows. How can we prevent that?

Brennan Dunn
Yes, I think the creepiness becomes when or starts when you do have you seen like, what Clearbit reveal in platforms like that allow you to do were given an IP address, it will say we think they work at IBM and then like a little pop up will show up like a chat drift bot saying, hey, IBM or something like. That to me is because I've been at like a Starbucks, where presumably maybe some guy at IBM was working there, too. They think now Starbucks' IP is IBM and now me, I'm getting addressed as like an IBM type person. Like, I think that to me is where it's kind of the Account Based Marketing is the way that it would be described. That's where it gets a little creepy. What I prefer is just more like, I'm very open to people and saying, hey, like, when you joined, you shared with me that you use, I don't know, let me give you a hard example of that. So with some of the work we're doing now with Justin Welsh, like we're capturing information about where they are in their journey, like, are they just starting out? Are they already pretty successful with a team? And we're very clear about not just saying, like, you know, this works great if you have a team. Instead, it's like, hey, so you shared with us that you, you know, you work with a team. That's awesome. So here's why we think that this will be able to help you get you and your team to the next level. Right? So it's not about being coy. It's not about being, it's just about like, again, like showing that you're listening and saying, I'm listening. And here's a different set of bullet points for you that I think are gonna really resonate with you about why you may or may not be interested in this thing. And I think people do, I mean, there's been like professional Google of like, personalization stats and stuff. And HubSpot did a big thing where they had different, you know, all these attributed stats and one of them was about how people do are generally okay with sharing this information with a brand, assuming it's used for good. So assuming it's not, you're not just sneaking stuff in, you're not like labeling people's IBM employees when they're not. Instead, it's you've shared this with me. So here's why. I want to make sure you're getting enough email, you're getting enough view on webpage is like let's cut to the chase and not try to cover every FAQ, every possible edge case, let's just focus on you. So this is you. This is why you need this. And somebody who did that well, is we worked with Pat Flynn of Smart Passive Income a few years back with his black friday promo where he had a lot of products on offer. And we just did exactly what we're saying now, which is you fit this profile, you obviously haven't bought this product. We think you need this product. And here's why. And they did a proper AV test control because they weren't sure like would this actually work? So what they basically had, I think, I forgot what the whole back rate was, but some meaningful back percentage that got what their black friday campaign was going to be, which is just, here's everything we have on offer: discounts versus the really focused thing. And it was something like 2.38x more performance or 238% more performance for the personalized niche thing than their control group. So it's just again, it's one of those things that make sense why it works. It's not black magic. It's just saying, if you fit this profile and you shared this with us, here's why we think this product is probably better for you right now.

Arvid Kahl
That makes sense. I like what you said like listening, listening instead of just like assuming or enriching with data that you're not supposed to have, right? Like listening is just really okay, you have a need. You express it very clearly. And I have something for you. Whenever it comes to marketing that uses these kinds of things and speaks to people directly in both a personal and yet also an automated way, I always have to convince myself that it's okay. I'm just, I have been so burnt by the kind of outreach that is personalized, but it's really not.

Brennan Dunn
Yeah, they're not filling in blanks. I mean, that's like, hey, Arvid. Hey, first name.

Arvid Kahl
One question is like, what's the worst way you could potentially personalize an email?

Brennan Dunn
I think it's when you're putting in specific data and it's the wrong data. Like, if I was addressing you as Jenny, like, what's going on there, right? Like, I think that's the worst thing you can do. And that's why I try to make it like, again you shared this with me. Have an ability, you can use a tool like smart subscribers something to let people like remap things because people's needs change. Maybe now they're focused on this instead of that. So I think like, my big thing is, obviously, I want to treat my list like adults. So yes, this is an automated email. But I wanted to, you know, listen to the fact that you shared x with me. So here's a set of benefits that I think you're really going to resonate with. And making that clear, like not trying to be all like, you know, it's like you, I hate like, you know, like these automated webinars where it's the fake chat room. And they're like layering the ghosts of previous live webinar or fake webinars, like on top of the same timestamp. So it looks like it's like, really high engagement. I hate that stuff personally. I just want to be able to say, I can't talk to all 10,000 of you one on one. So here's my best attempt to try to make it clear should you get this or shouldn't you based on what I know about you.

Arvid Kahl
Yeah, yeah, I hate the fake stuff, too, because it's so obviously fake. However, and this is probably a whole topic that we could spend hours talking about. But it becomes less and less fake the better the technologies get that emulate real human beings, right? I kind of want to talk to you about AI. And its place in this whole thing, like ChatGPT you already mentioned it. That is, for many people an interesting avenue to like, kind of uniquefy emails that they send, like turn something that they know, give some context about the person and then have the whole email rewritten automatically by Chat GPT and then send out that, for example exists, but there are probably many, many other different ways. Even to collect the information, you could do it in a kind of talkative manner, right? Where the ChatGPT becomes the agent of collecting information and then spits out some kind of normalized data. Without any prior judgment from my side, I would like to hear your prior judgment on this whole thing because you've been working in this for a long while without these tools. So I wonder what do you think about AI, both the good and the bad for the future of personalized marketing?

Brennan Dunn
I think I'm generally a fan of it for kind of basic, I was talking to a friend about this the other night, like basic, like junior analysis type stuff like. So I talked about how I, when we don't know what the core set of questions should be, ask your audience for a one or two sentence thing. You can upload a CSV of his responses to ChatGPT and say identify, you know, up to four trends about why people are joining my list. And that's pretty great. I mean, I used to do that manually with a spreadsheet where I would like have all four or all the CSV responses. And then as I went through them one by one, I would say, oh, this is an agency owner. I'm going to create a new column called agency and paste it here. And then I'd have a big like, second column of all the agency owner responses. And then now this is a, I don't know, health clinic, same deal. So it's definitely made my life a lot easier for kind of the data analysis thing. I also think for things like writing, because one of the annoyances, if you will, of doing personalization right would be, let's say you want to, you have five industries you're targeting and you want to have a different headline on your sales page for each five. It's still, like the hardest thing is copywriting here because you need to come up with five variations. But you can easily go to ChatGPT and say, hey, given these five industries and this core headline, modify it for each and it'll spit it out and you run with that. You can even get it to write liquid code that you can then use to.

Arvid Kahl
Yeah, that's cool, right?

Brennan Dunn
Yeah, exactly. So it definitely made my life easier. Like I don't do a lot of content creation with ChatGPT. I really do much of it. It's really the beginnings of like the headline modifications I just mentioned, but I am a big fan of using it for identifying trends over large datasets.

Arvid Kahl
It's funny because I use it the exact same way for my writing. ChatGPT is my writing research partner. It's not my writing partner, doesn't write for me, but I throw topics at it. I tell it to give me like 20 versions of a certain headline until I find one that I like and then use parts of it to use for my own or I just ask it to look into a certain topic for me and give me a couple interesting but not all too common like perspectives on it. You can use it like a human being that you just, like ask a question and it gets your data, but you can feed data into it and have a look at it, too. I've been doing this too with my analytics and like, look at trends that give me certain insights. It's a research assistant, that's what it really is. It's not a good writer, right? The writing of the tool is so generic and so bland, sets the whole point I guess of ChatGPT even if you try to color it in a certain way, but I use it a lot in my early exploration and my drafting stages. And that's where it belongs, personally. So where does it not belong for you? Like, whenever it comes to writing, would you use it? Like, if you had all this data about a person and you send an email, would you allow it to rewrite the whole email without you actually getting to check it off and then send?

Brennan Dunn
Possibly, but not the whole email. So where I found the idea of what ChatGPT does well to work really nicely with is so in order to you to do and I'm gonna switch to just talk about the English language for a second, we're gonna get more complex in a second. But let's say you wanted to do something like you wanted to change up a sub part of a sentence depending on the email platform they use, right? So let's hypothetically say you wanted to say, all you need is a email marketing platform. Right? So keep it in English now. Now, if I knew that you wanted to use Active Campaign. I couldn't say all you need is a fill in the blank Active Campaign account. Because now a with a vowel following sounds weird. So when you think of the actual technicals of you know, personalization, for example, it's a lot of that. It's a lot of, but then you have these articles, like a and an, where there's very specific grammar rules around how that works. So I think that's the hardest thing I've run into in doing things like, we do this all the time for clients. Like a common thing you see at the bottom of newsletters now is here are x ways I can help you when you're ready. And we've been for our clients been making it. So these descriptions and our personalized and even the products being shown are personalized, depending on what you've bought or haven't bought. So if it's like, if it's one way that they can help you is here is one way is, not are, is one way singular. I can help you. And then here are two ways plural that I can help you. And it's that kind of thing, which is really tedious to get right. It's doable, but it's annoying. And I think what I like about ChatGPT is the ability to know basic grammatical rules. And it'd be great to be able to say like, hey, we're looking at put here are x ways where x is a number, modify their surrounding sentence as necessary to make it grammatically correct. Like that's the thing, I don't know how to do that yet with, like, during that same ConvertKit. But that would be the were the biggest, interesting use case for me right now.

Arvid Kahl
It's interesting to see, like, you're thinking about AI here as a tool, like with one input one output, right? It's not that AI does a job for you. AI is just kind of complicit and making the thing that you already create and structure more effective. I think it's a very important distinction. And it kind of I feel the same way, right? I want it to be there for a specific part. But I don't want AI to do my work. Right? And it would be wonderful if AI was kind of smart enough or ChatGPT was smart enough to actually draft like a full email very specific for this person. But if there's one cringe statement in there, right? That the whole thing is lost. And I wouldn't want to do that at scale. That's very interesting. It's an interesting perspective. And I think there must be a way for ChatGPT AI tools or something to offer this. And if not, it's a great SaaS idea to like

Brennan Dunn
Yeah

Arvid Kahl
Like a grammar, it's just really simple, really grammar checker, grammar corrector kind of tool. You know, it reminds me very much a Feedback Panda, our SaaS that we built and sold a couple of years ago because we had the same problem that you're actually talking about. We had student feedback templates, where you know, online teachers would write feedback for the parents of the student and that was a male and the female version. And obviously pronouns are different, right? If you only write one version, you kind of want to have the male or the female version as well. You don't want to manually write the same thing, right? Because you teach male and female students at least. In China, they have a pretty strong binary perspective on this still, but that was kind of what we were serving, right? Chinese schools and we had like American online English teachers. That was kind of the market we were in. So they had these templates, only one version with a he and they would manually go through and change like the his to her or the him to her. But the other way around was tricky because if you have her, you can't automatically know if it's a him or a his. You can't change her into him.

Brennan Dunn
Exactly, yeah

Arvid Kahl
So I built this complicated machine learning system that took like 18 hours to train like a statistical machine translation model in the background. That was probably the largest like technological thing that I needed to build for this really simple text based app because ChatGPT did not exist. Today, it would be one API call with the text and would come back with the perfect pronouns in the exact spot.

Brennan Dunn
And funnily enough, I did exactly that. Today, I mentioned the, like agency thing that I kind of started. I took my intake document that I've been using for years for my own clients. And it used to say, like, hey, I'm super excited to work with you. When you're ready, let me know, you know, that kind of thing, right? And I needed to go and modify that whole thing to be we, instead of me and it's yeah, it's that exact thing, which is like, oh, it's such a like, this is such a solved problem. As a coder, I'm like, why am I doing this manually? But then like, it's easier just doing it manually for this one off need. But, yeah.

Arvid Kahl
I think AI has a lot of potential with that kind of stuff where it has through its own training data, so much information that can make a snap judgment so easily. Where for us, it would just be a lot of if then statements, which is kind of what personalization is, right? So it's kind of also why I'm asking because I wonder if you can train. And that's maybe the speculative part in all of this. If you can train a system to understand what you want. And it then sets up these things that you're currently setting up manually and it looks into the data and creates new products, like branches or information for new products all by itself. So you only really have to say yes, sure, I want this or I don't. Do you see this coming? Do you see this coming in the more approximate future? Or do you think this is still a long while?

Brennan Dunn
I think we're already seeing that kind of like, if you look at like E commerce with the recommendation engines, they're kind of increasingly getting toward that, or at least the big ones, right? I mean, we all know Amazon does this. We all know, like, the enterprise players are doing this really well. But I think it'll probably start with more recommendation engine type stuff before it gets to dynamic product or messaging, positioning. I mean, the personalization tech now is kind of its hands on at this point. But I do think if there's a high degree of likelihood that it wouldn't screw up, then that'll get even, yeah, that'll become more of a real thing. I mean, we've even had to like I've built an internal tool that allows me to say, given all the different possible segmentation options, first off, tally up how many total combinations there are and then run this bit of text with liquids surrounding it. So liquid is the engine we're talking about for doing like, if-else's and personalization over email, paste it in, and then it'll yield the output for every possible combo. And what's nice is you can then just skim it and say, oh, a active campaign or this is grammatically off or oh, shoot, this looks awkward. And then you can go and fix it accordingly. Because it's really hard, frankly, unless you think you have a programmers mentality to look at, like a block of if-else's, especially if they're nested if-else's. And think like, what are the possible permutations that can come out of this?

Arvid Kahl
Wouldn't it be cool if there was like a liquid command just for correct grammar? And it would automatically

Brennan Dunn
It would be great. Yeah, you use Rails, right?

Arvid Kahl
We're using elixir, but yeah, I do some Rails in the past.

Brennan Dunn
Okay, well, there's a, I haven't used Rails in years. So I'm probably gonna botch this up. But there was a active method, I think it was called or active something function called, was it like PluralEyes, where you could say like, the first argument is a number. The second argument is a word like way or thing or something. And then it will output five things or one thing and it'll like do that for you. So I'd love to see, like, the equivalent of that being brought into a templating framework, like liquid. Yeah

Arvid Kahl
Interesting. I wonder if we're ever going to see like grammar correction on the client level, but the email client actually goes through the thing and just fixes the little mistakes and that would be an interesting feature too, because that would kind of absolve you of the need to your if-else statements all the time.

Brennan Dunn
If you think of it like real time translation services, it's kind of the same thing. It's just like translating from English to English, let's say or something.

Arvid Kahl
That's what ChatGPT can do as well which is kind of crazy. And that's a whole other thing, didn't even think of this before, but you could automatically like translate your full email into the local language of choice of your potential customers without having to do it.

Brennan Dunn
Yeah

Arvid Kahl
Is that something you would ever try? Or would you be quite afraid of the potential cringy translation?

Brennan Dunn
I'm actually still I'm doing that not with translation. But I'm doing that with on my website. So I'm write messages website, where if you're in a, I think and I set up for Canada because I think Canada does as well. So but like if you're in a country where it's an essence that have a Z or a zed, it's spelled that way so like, personalized with an essays rather than

Arvid Kahl
Wow!

Brennan Dunn
Yeah, so I've done that. And, but that's an example of like, it's more of a parlor trick there. Like, I don't think anyone's gonna care. Like, I don't think somebody in Britain who's used to like, the Americanization of everything on the planet, like he's gonna go to a website and be put off by a Z instead of asked for the word personalization.

Arvid Kahl
But they will appreciate yes, right? That's the thing. That's where this comes in. Right?

Brennan Dunn
Yes

Arvid Kahl
It's nice to see something done for you, not to you, but for you

Brennan Dunn
Exactly, right!

Arvid Kahl
That is very cool. Well, I'm very, very excited by the prospects of this, particularly in the future as it becomes more of a thing that understands what I want and does it for me, like, prepares these kinds of things for me instead of me having to dig into all these if then else statements. But I also see at this point, at least, how valuable this could be. And I'm really fortunate, I guess, to know you and know that you have extreme levels of insights and put them all in a book, which is also very cool. I'm really looking forward to reading the whole thing. One thing that I wonder about this book in particular and we talked about it at the beginning of this conversation, right? Books are kind of products you finish and then you sell them at scale. But your marketing approach tends to be a personalized one. How do you combine this? Like for this book product, in particular, do you have any ideas on how you can still personalize the sale for this? Like your sales efforts? How can you still kind of market it in a personal way?

Brennan Dunn
Yeah. So funny enough, I kind of beta tested this last Friday. I was at a local conference and I brought copies of the book to the event. And what I did is when I would talk to people and meet with them, depending on kind of where they were like, are they you know, like, what are they working on, like their business, whether current challenges and so? Like the normal stuff you do when chatting with people at a conference, I would recommend certain chapters to them. So I think that would be where I'm going to plan on doing this is depending on what they share with me, like if it's audience growth or automation or whatever else. I can say, hey, so I'd love to tell you about my book, This Is Personal. I specifically want to talk to you about why you're going to love what I talked about on chapter or page 75, where I get into onboarding new subscribers and how I think that what I cover both there and in the rest of the book is going to benefit you. And I think that's something that could be done like, yes, it's a generic kind of mass market product because you can't personalize a book. I can give recommendations for what I think they should read specifically, pay attention to specifically in the book. And that went over really well. You know, from my limited, like, five sample size. I did the conference last Friday. But I think that's what I'm planning on doing is exactly that.

Arvid Kahl
That's so cool. I just love how you have this deep insight into this whole approach so that you can even use it. You can even personalize the unpersonalizable. That's awesome.

Brennan Dunn
Like I said, it's still one size fits all. But yeah, the way it's yeah, it's a book recommendation to a friend. Think of it like that. Yeah. Like, why do you think you'd like this book?

Arvid Kahl
And the specific one, right? It's not just the book is great, but it is great. And here's the part that is good for you. I think that's just like if you look at any kind of marketing that you can do or that you do and give that information to your customer at that time, like this is not just wonderful. Of course. It's wonderful. That's why I did it. Right? But it's wonderful and wonderfully useful to you because of this particular part. That just to me explains why personalized marketing is so powerful because it contextualizes and empowers people right there.

Brennan Dunn
Exactly! That's what I've been trying to get across to you.

Arvid Kahl
It took 50 minutes. No, I really appreciate you like sharing all of this with me. And I bet a lot of people will appreciate reading this book and I hope they do and I hope they find you. So where do people go to find you on the internet? And where do people find the book?

Brennan Dunn
Yeah, so let's start with the book. I do have a vantage domain for it. So thisispersonalbook.com

Arvid Kahl
There it is.

Brennan Dunn
It just links to like Amazon and stuff. So that's the best place to get the book, comes out 17th of October. I would love you to reach out and let me know what you think of it if you do buy it and read it. But my createandsell.co, that's my twice weekly newsletter on whatever happened to be thinking that week about stuff like what we just talked about. I talk everything about personalization, but also just email marketing in general and what I think it can be done for how creators like us can use it to better connect with our audience.

Arvid Kahl
That's one of the few newsletters that I always read. So I really, really appreciate you sending that out as much as you do and as personally as you do. So I appreciate being on the receiving end of that as well. Thank you so much, Brennan, for being on the show today. That was wonderful. Thanks for sharing all of your knowledge with us today.

Brennan Dunn
Of course! Thanks, Arvid.

Arvid Kahl
Absolute pleasure.

And that's it for today. I will now briefly thank my sponsor: acquire.com. Imagine this, you're a founder who's built a really solid SaaS product, you acquired customers, and you're generating consistent monthly recurring revenue, the SaaS stream, right? The problem is, you're not growing for whatever reason, maybe it's like of focus, lack of skill or plain lack of interest and you feel stuck within your business. What do you do? Well, the story that I would like to hear is that you buckled down and you reignited the fire working not just in the business, but on the business and did all these things like audience building and marketing and sales and outreach. And six months down the road, you made all that money. You've tripled your revenue and you have this hyper successful business now. But reality is unfortunately not as simple as this. And the situation that you might be in is different for every founder facing this crossroad. Too many times, the story here ends up being one of inaction and stagnation, until the business itself becomes less and less valuable over time or worse, completely worthless. So if you find yourself here, already right now or you think your story is likely headed down a similar road, I would consider a third option, that selling your business on acquire.com. Because capitalizing on the value of your time today, that's a pretty smart move. Acquire.com is free to list and they've helped hundreds of founders already. So go to try.acquire.com/arvid and see for yourself, if this is the right option for you right now.

Thank you so much for listening to The Bootstrapped Founder today. You can find me on Twitter @arvidkahl. And you'll find my Twitter course and my books there too. If you want to support me and the show, please subscribe to my YouTube channel, get the podcast in your podcast player of choice and leave a five star rating and a review by going to (http://ratethispodcast.com). I would really appreciate it. It makes a massive difference if you show up there because that will show the podcast to more people. It will show up in their feeds. And any of this will really help the show. So thank you so much for listening. Have a wonderful day and bye bye.