How to scale your DTC brand with Tyson Drake: Growth strategies, data tips, and AI trends

Key takeaways:

1. DTC growth starts simple but gets smarter over time

Direct-to-consumer (DTC) doesn’t just mean selling online—it’s about finding the right mix of channels to reach customers. Brands often start with Google and Meta ads, but growth happens when you expand into new channels, products, or even markets.

2. Forget ROAS—focus on contribution margin

Revenue and ROAS are outdated metrics. If you’re not looking at contribution margin (what’s left after costs like shipping and marketing), you’re likely wasting money. Tracking this daily keeps your brand profitable and your decisions sharp.

3. First-party data is your best friend

Want better insights? Ask your customers. Post-purchase surveys and behavioral data can show you what’s working and what’s missing. Combine this with your sales and ad spend data to make smarter decisions that customers actually appreciate.

4. AI is shaking up marketing

AI is becoming a game-changer for DTC brands. From tools that analyze your data to AI-powered scriptwriting for ads, these tools are making it easier—and cheaper—to get great results. The tech isn’t perfect yet, but it’s evolving fast.

5. Scaling requires playbooks, not guesswork

Tyson’s secret? Tried-and-true playbooks. These strategies work across industries and help DTC brands tackle common challenges like plateauing growth or managing in-house teams.

6. Fast feedback loops are a DTC superpower

Unlike traditional retail, DTC brands can make changes and see results quickly. Use this to your advantage by tracking performance daily and adjusting on the fly. It’s the easiest way to stay ahead of the competition.

Read full transcript here:

Edward Ford:
Hi, I'm Edward from Supermetrics and this is the Marketing Intelligence Show, the podcast that empowers marketing leaders to work better with their data and make sure every marketing dollar counts. Now let's get into today's episode.
Evan Kaeding:
Hey everyone, and welcome to the Marketing Intelligence Show by supermetrics. I'm Evan Kaing, director of Solutions Engineering at Supermetrics, and today we're joined by an incredible guest, Tyson Drake. Tyson is the former CMO of the odi, A nine-figure D two C brand, and now works as a fractional CMO helping e-commerce brands achieve profitable growth. In this episode, we'll explore Tyson's approach to building robust growth strategies, measuring performance, and the impact of emerging AI tools On D two C marketing plus Tyson shares how brands can leverage first party data to create more meaningful customer experiences. So without further ado, let's dive in. Tyson, welcome to the Marketing Intelligence Show.
Tyson Drake:
Thanks, Evan. Really happy to be here.
Evan Kaeding:
Excellent. So, Tyson, with your experience, I think it'd be good maybe just to zoom out for folks who maybe aren't quite as well versed in the lingo, what does D two C mean to you and how do you define it?
Tyson Drake:
Yeah, so for me, D two C stands for direct to consumer and for me it started off as new thing that people thought would be exclusive to online purchasing behavior, but over time it's probably evolved to being part of a mux much broader marketing channel mix where direct to consumer is an existing is one of many channels that brands would use. So maybe a typical playbook would be if you are a new brand you sell only online to you have a website, you sell online, and then later on you kind of bolt on these other channels. So that might be retail, that might be wholesale, that might be Amazon, other marketplaces. And so direct to consumers started as online owning experience and it's kind of evolved. A channel is part of a much broader marketing mix.
Evan Kaeding:
Understood. So it's not only the marketing, the set of marketing activities around getting your products into the hands of consumers, but it's also part of the definition of how you get those products to consumers, the actual distribution channels as well.
Tyson Drake:
Yeah, the easiest way to think about it is DTC Direct Consumer is a distribution channel and is one of several distribution channels.
Evan Kaeding:
Understood. And for many that might be an online experience, but for a growing number or perhaps a large number of established brands, it might actually be an existing channel as well based on physical retail stores.
Tyson Drake:
Yeah, exactly. Maybe you have a dominant physical retail presence of your own stores, a dominant wholesale presence. Wholesale is very dominant channel, so you sell through your own stores and you sell through suppliers via wholesale and maybe online retail or direct to consumer is a very small percentage of your overall distribution mix. And so one can benefit the other. You have the halo effect with online if you have wholesale or retail. And for a lot of direct to consumer brands or digitally native brands, wholesale and retail can be completely new channels and they brands have never invested in wholesale before. They need to learn how to do that or they go out and experiment how to run their own retail. So yeah, high level overview.
Evan Kaeding:
Great, thank you for the categorization. So it sounds like it might be one of the ways that brands get off the ground by selling directly to consumers or potentially a distribution channel that mature brands with an existing channel strategy might add on for scale. With that being the case, what does a typical media mix look like for supporting a seven or eight figure D two C brand? The kinds that you typically support today,
Tyson Drake:
I would say let's bucket this as a brand who is starting out, and let's put aside for a second, like a multi-billion dollar brand who wants to invest more into direct to consumer? So let's ask a brand who was starting out. So typically the two most ad platforms are Facebook and Google. So one represents demand capture, you could say. So Google ads in this instant with the exclusion of YouTube for now is a demand capture channel. So someone needs some new shoes or some new pants, maybe they have a specific, maybe they want a water bottle, so they're like, I need a water bottle. So they can go directly to Amazon if they want and type that in Amazon being search engine. Or what a lot of people do is they just go to Google and they type it in and they go, yep, these brands look interesting, we'll buy from.
So I want to refer to that as demand capture channel. Pretty straightforward, direct to consumer 1 0 1 there I'd say. But the problem with that is there's only a limited number of search queries. There's any limited number of monthly search queries that happen for all these different categories. And so keeping with a water bottle trend, you could say that let's just make a keep numbers simple. Maybe there's like a hundred thousand water bottle queries a month in Finland for example. And actually that's probably really generous. I'd be surprised if there was like 50,000 queries something might look at afterwards. So let's just highlight a hundred thousand queries and of that maybe there's like 20 drinkware brands that advertise to Finland, and so they're not going to get an even distribution of demand capture. And so all these different brands have to compete against each other and then can the cost per click goes up and some brands can afford to pay for customers more than other brands.
Maybe they have complimentary products that increase their LTV and so they can afford to hire CPA to acquire new customers versus someone who might just be starting out. And so there's all those competitive dynamics within the auction. And so you have the demand capture, but then you have or meta Facebook and Instagram mostly, and they have messaging and all that, but Facebook and Instagram, and I would argue it's kind of a demand capture channel, but it's not a demand capture channel in the way that we think of Google ads where someone's typing in a C searcher query. And so I would argue advertising on Facebook and Instagram is advertising to InMarket audiences. And what I refer to an in-market audience is someone who has demonstrated via whatever buying behavior signals that meta has on them that they're in the market for a specific product or category.
An example might be you Google Drinkware and all of a sudden you get all these drinkware ads on meta or you might be looking for a new desk or something and all of a sudden, boom, you haven't been been to Facebook, but all of a sudden you're in the desk category, maybe you read a desk review or something. So Meta's got all these data points. It's why it's the most sophisticated advertising platform in the world is because they have all these data points and then via creative, you can then target people who are in the market of these products. So I would argue that if someone is in the market, that meta has deemed in the market of this productal service that it's kind of like a demand capture channel anyway, because the demand is there, you're just interrupting people via creative to say, Hey, we think you're in the market for this product or service.
Here's our offer, click on add and come buy or at least come and have a look at our website. But once again, the percentage of people who are in market for a given category is limited and it's bucketed by geography as well. And so while Google and Meta are the primary channels for a new brand starting out, there's kind of a framework that I think through and that is, so think of a chart. So on the X axis of the chart, that's the bottom. You have time and then on the Y axis of the chart you have percentage. And so you could say that we start off in a brand's lifecycle where they're at the bottom, the intersection of the X and the Y down the bottom there was 0%. And then over time they launched their product and then over time they spend some money on Google targeting ads captured responses, and they do some Facebook ads.
And within their category, let's say drinkware, they're competing against all these other advertisers. And so they're competing against all this in market demand. And then so maybe revenue, it goes up a bit and then over a few months it hits a bit of an equilibrium or a few years, maybe two or three years, there's maybe some equilibrium where you have captured as much in market demand as you reasonably can. And now anything else is just the growth of the category and you're just writing the category growth or you are growing at all because you are not innovating or not launching new products or you're not internationalizing, whatever that may be. And then that's like demand capture channel and then that might be five or 10 of the category. And then there's always this other part of the category, whereas people who aren't looking for your product and service yet yet.
And so maybe you're not in the market for a drinkware, a drink bottle. And so in the future you might be though. And so this is an expansion of media channels. And so Facebook and Google absolutely dominate this demand capture channels and then maybe you could bolt on a Pinterest or SNAP or other channels that might add some incremental value. But the big transition happens when you go from targeting in market demand to targeting people who are not yet in the market but will be in the future. And that's where the majority of people are is they're like, I'm not looking for a desk right now, but I'm probably going to need a, I want to stand up desk in the future. I don't want one yet. I'm definitely not in the market, but I probably will in the future. And so what you want to do as when you reach this diminishing returns of channel of market penetration, you could call it, you want to transition to, I guess you'd call it, I don't like the term brand channels because every channel is a brand channel.
I mean what does that even mean? But I call it non in-market audiences. And you need a transition from targeting in-market audiences to targeting non in-market audiences. And so the simple framework is if there is behavioral targeting involved where you can target search terms or target in market, then it's probably more of a demand capture channel. However, if it was something like broad reach, if it's a broad reach channel where you're targeting the number of people who maybe a demographic or specific air times within a popular TV shoes like Olympic ads for example, that's broad reach, it may be targeting sport watches or something like that. And so that would be like we want to target these people to create what's called a positive, what's called mental availability, which is a fancy word for when they are in the market. They then consider your brand as part of their consideration set. And so that's a long way of saying in the beginning for A DTC brand's lifecycle, you start with meta and Google. Over time it gets saturated, you kind of max out in market demand and then you transition to more broad reach channels. But that's not the only way. You can also bolt on distribution mechanism. There are lots of ways to grow. We're talking about scaling now. Interrupt me, I'm just going to keep going.
That's just one way. So within let's let's say your existing market, Finland, what's the population of Finland? What's it like four mil? Five,
Evan Kaeding:
Five and a half million? Yeah,
Tyson Drake:
Five and a half bit of trivia there, five and a half million. And eventually you're just going to saturate the amount of people in Finland who want to buy drinkware. And so even with TV ads, and maybe it's not even economically viable to do TV ads because maybe you have a low IV product or something like that or low LTV products, even if you have a lot of purchases, people just don't buy drinkware very often. I don't know. But that's one way to scale is you just go through the EU countries and you do the in-market playbook again. So you start with Facebook and Google do the in-market playbook, launch, launch challenge with translations there as well. Currencies, how do you deal with ops and shipping and taxes and all that stuff, all the fun stuff. Or you can just advertise to us or UK as bigger populations. So that's one way to isolate. The other way is adding new channels. So Amazon, retail, wholesale, another way you can grow new products as well. And new products is one of my favorite ways to grow because new products works across new markets and new channels as well. They get the benefit, everything gets the benefit of new product if it's successful as well. So yeah, that's broadly pretty broad answer to what channels do people usually start with and some phases of how that plays out.
Evan Kaeding:
I like how you broke down the categorization between demand capture and then the ability to target consumers who may not necessarily be in market. It sounds orthogonal to the dichotomy between brand and performance. Maybe first part of a two-part question is do you agree with the categorization of media as brand and performance media? And if so, how does that jive with the framework that you just laid out as well?
Tyson Drake:
Yeah, so I think it's more of a semantic issue rather than what I think it is. So I think if I understand it correctly, the way most people classify brand when they talk about brand is broad reach. And so people would say brand advertising or I shouldn't say people because I know a lot of people who have quite different definitions of brand. And so I'll say the advertising industry would classify brand advertising as tv, radio, billboard. That's what they would probably classify brand ads and how I would classify that as broad reach. And I like to distinguish between broad reach and this demand capture and brand versus performance because everything is a performance channel. If you can, let's look at TV for example. Well, for decades the CPG brands have been using marketing mix modeling to measure tv. And so does that mean TV isn't a performance channel?
No, TV is absolutely a performance channel. If you can optimize any channel, in my view, any marketing medium, it is by default a performance channel. Let's look at billboards as another example. And so I worked for an internet service provider that iPod in Australia called Unity Wireless from 2017 maybe to 2019. And yeah, we did billboard ads and due to our billboard ads, we could tell due to when people would enter in their, you go to internet service provider in Australia, what's the first thing you do? You check if you can get connected, you check if the internet service provider can service your internet pretty straightforward. So the first thing you do is enter your address and then it comes up with a yes, we can service you for what you want or no, we can't. And so we would have all these billboard ads everywhere.
We had different cost per query, you could call it a query being a search on the website. We had different cost, cost per query prices that we could see with rollout of billboards in different suburbs of different places in South Australia where I live in Adelaide. And then we'd use that to model out, well what is this from a CPA basis and what suburbs look attractive to continue or roll out a deeper penetration of billboard ads versus suburbs that don't. And so that's a performance channel. You've got an active CPA performance query target and then you can model it out into percent of people who convert and you get a cost per acquisition, incremental cost per acquisition per these billboards. And so that's classified as a brand channel, but I will classify that as a performance channel. And so for me, every channel is a performance channel.
You approach how you approach measurement is different for every channel. With online channels like Meta and Google, you typically at a very early level look at in platform metrics and then maybe get a bit more sophisticated and start to maybe implement some third party attribution software. It's not perfect, but it helps for directionality and that might be your next evolution and maybe you get a bit more sophisticated and you start doing some incrementality testing and market mix modeling and using incrementality testing to recalibrate your model. And so there's like an evolution of sophistication and the seven figure brand who's just starting out doesn't need incrementality testing, market mix modeling, just looking at contribution dollars or am I profitable is the best indicator in platform metrics. So there's only two channels, it's probably a pretty good start. And so I just don't like the brand versus performance marketing because in my view everything is performance and it's more about what stories can we tell across different channels, what brand stories can we tell across different channels and what is unique about this channel that allows us to use it to tell a story a certain way or is there a certain type of creative that this platform is unique that we can utilize with this platform?
Like TikTok is a classic one where how you tell stories on TikTok is different to how you tell stories on tv. And so I would say what is unique about the channel that allows for your brand message to be told and then I think that's a bit more important. And then measurement completely different per channel as well.
Evan Kaeding:
Yeah, and it sounds like to me based on your definition, any channel is a performance channel if you can establish somewhat of a tight feedback loop on Yeah, exactly of that channel. With that, probably the next best question would be what are some of the most common approaches that you see in performance measurement for D two C brands?
Tyson Drake:
So the evolution goes in platform metrics. So what is Facebook telling us? What is Google telling us? And then we all know the challenges with Post OS 14, we all know challenges with that environment now and we all know that channels over attribute conversions and we all know the challenges associated with that. But if you're a new brand, that's what you're looking at, you're not paying some marketing mix modeling provider for MMM, you don't even probably know what that is. So the evolution is looking in platform metrics and right now direct to consumers come a long way in my view. Previously what used to happen was people would look at ROAS return on ad spend and they'll look at that in platform and they'd use that to make kind of decisions. That model is long dead for I'd say the majority of brands. And what's happened now is the industry has really been pushing forward to contribution margin or contribution dollars.
And what that means is basically gross revenue minus discounts, minus returns, and then you get minus your variable costs. So variable cost being cost of delivery, which is basically how much does the product cost to go from a factory to a warehouse that they own warehouse at their party warehouse and what is the cost then to get that product to the customer. So cost of delivery, fully baked in cost of delivery, and then minus marketing spend and that marketing spend across all different channels. Some people include creative in that as well. And then you're left with what's called contribution dollars being the dollar amount or contribution margin being the percentage leftover divided by net sales. And so a lot of now are tracking this kind of metric on a daily basis. And so they know, okay, well if my contribution margin percentage wise is like 35% weekly or daily and if dollar wise let's just make up a number and say dollar wise that's like $50,000 as long as my contribution profit dollars at the end of the month is greater than the cost of my opex, so that's like people rent, et cetera, then I am making money.
That's great. And so the good thing about direct to consumer is that, and that's a simple financial breakdown. I'm definitely not a CFO, so I'm not going to go into cashflow or all the fancy terms that the accounting world loves, but what I'd say that is the sophistication of direct to consumer reporting has gotten a lot better and people are tracking basically contribution dollars or net profit every single day. And what that allows is for really fast feedback loops in marketing spend and decisions about where to allocate dollars are happening real time or close to real time, as tight as the feedback loop that you can where people are going, why aren't we profitable today? Why is that contribution margin negative when taking into consideration our opex costs, what's going on? What changes do we make? Whereas previously people would have to wait two weeks even sometimes a month after the p and l was released to be like, oh shit, we didn't actually make any money that month.
And now that could still be the case if you have other challenges with your business that you don't directly see as part of the contribution or calculation I mentioned. But in general, the industry is getting a lot more sophisticated and this is with the help of Shopify helping standardizing data infrastructure, things like platforms like Supermetrics where you can just export marketing dollars across all of these platforms and it into a simple Google sheet and then implant some formulas like take Shopify minus some cost of delivery variables that you could calculate pretty basic and then minus your marketing spend. And then as long as that amount is greater than your opex, you're pretty happy. And so yeah, the industry is getting a lot more sophisticated and for the better. And so back to the question about this evolution of measurement. So in the very beginning, new brands that are launched, I know a bit more sophisticated are tracking that from day one.
Some are mid seven figures are tracking that, but the infrastructure is there, it's all accessible via these APIs. And if you're a bit more sophisticated, you probably have some third party attribution in place like a North Beam, which is a pretty popular, the thing about directing Schumer is it has all its own MarTech marketing infrastructure. It has all this own and it's largely built on Shopify. And the Shopify app ecosystem is very robust and allows for plug and play of these apps that you can. And there's a lot of ones out there that people know best practice or whatever or industry recommended or I don't even know what that means, but just ones that people are good, I've used some good software, I used some bad software, I recommend some good software, I don't recommend bad software. So you'd be more sophisticated. You're looking at this multi-touch attribution. Yes, it's not perfect, but from experience it's more so about the directionality versus is it perfect, right? You can do without MTA, if you're just tracking contribution dollars on a daily basis, you're like, oh, we increased spend on Facebook. Oh shit, we lost money the past two days. Okay, well that probably wasn't a good idea. Let's ladder it back down and see what happens. I would argue that's probably a better than this MTA
Evan Kaeding:
Two takeaways for me from that, which is one is if you're tracking revenue as opposed to contribution margin, you might be shooting yourself in the foot and dramatically overspending to acquire customers. And the second is that D two C as a distribution channel is going to give you far faster feedback loops that are maybe far easier to optimize than what you'd probably otherwise get selling through traditional retail channels. Is that a fair interpretation?
Tyson Drake:
Yeah, absolutely. Traditional retail, I've seen brands have a one month feedback loop for how their products are selling. I think two weeks is maybe even weekly now, but I think two weeks is a bit more standardized now, but monthly before. So imagine not knowing how you've done for a month. I mean you can get some updates about how things are going, but you don't really get the reporting until the end. And so then the third part is you're a bit more sophisticated, you're doing some marketing mix modeling. The problem with that is correlation and maybe it doesn't give you the best read. You always have to update the model with these priors, which are just a fancy word for saying these new inputs, all these new results from experiments, you update the prizes and then you use incrementality testing to recalibrate this marketing mixed model.
And then it's just feedback loop. So you've got, you're running marketing experiments by incrementality testing, you're looking at how this is affecting your contribution dollars, are you going either increasing or decreasing is really the source of truth. And then from there you're like, okay, we ran this experiments, this incrementality test proved to be positively, incrementally was incremental. Great, let's update the MMM with the results, update the prior run the correlation analysis, and then that spits out your budget allocation suggestions and you just keep that feedback loop. That's probably a bit more sophisticated system. I wouldn't recommend that for anyone, only probably mid eight figure brands should start thinking about testing like that. But I know brands that really haven't done that until they reach nine figures as well. In revenue, yearly revenue, you can get away with just contribution dollars and like some MTA for like a while.
Evan Kaeding:
Sure. Well, and with all of the data that you get access to in D two C channels, you have quite a bit of personal information on the customers, like what they're ordering, where they're ordering from average order value, things like that. But there are a growing variety of media tools that you can use to personalize communication and send relevant offers to your audience Based on what you know about them, what are some of the ways that you're seeing brands use their first party data to personalize the experience of their most loyal customers?
Tyson Drake:
I think probably the best one that I like is this survey platform. So the survey platforms call themselves zero party data. I don't really like that. I just think it's first party data if you collect they're customers. And so the way I think that the heuristic for me is if it's a customer and you are collecting that data from the customer, it's first party data. So that's like products that they're purchasing, it's behavior from their behavior on the website after they purchase, if they fill out a survey for example, when did you first hear about us or what other products would you like to see? Tell us about your shopping experience today. I just classify that as first party data. It's customer data, it can be tied back to an order and if the information can be tied back to an order in Shopify and they classify that as first pay data, the survey is first party data if it is an existing customer.
And so that's how I think about it. So I really think the value is marrying the quantum qual, the quantitative data with the CPAs, how much we're spending the contribution margin, the behavioral insights of did this person click here, heat maps of you that as the quantitative data and how many people signing up for your survey, et cetera. And then marrying that with qualitative data. And I think the best insights coming from marrying the two. And so I would say the, but you have to know what questions to ask. And so you have these survey platforms, no commerce and faring are two probably most popular ones. And if it's a Shopify plugin, you put it on the website and after the purchase they ask you, there's some best practice questions that you can ask, when did you first hear about us? You can target new customers or existing customers with different questions.
You can understand what channels you can ask questions, when did you first hear about us? And then you can use that insight touch upon measurement again to help inform your mm m. So if you are seeing that, a good example is you won't see TV in MTA, right? You won't see tv, the effects of TV in MTA as a channel. You will see the downward effect on that as like, well direct increased page search queries increased, therefore a spend increase on page search are retargeting increased on meta or in market audiences might've even increased as well. So you all see these demand capture metrics kind of do really well during a period of tv. For example, a big investment in TV where you can meaningfully look at, see the downward effect of it,
But you won't see that in, you won't see TV as a channel in mt a, right? And so what you have, there's a few ways to then, so you have a gap, you have a measurement gap. And so one way to do that is MMM, yep, put TV as a channel, put the spend correlation though you get some insight from this. Demand capture channels we're like, okay, this was the effect on demand capture TV probably did that you can look at do a simple before and after analysis, but in other ways is survey concept. So where'd you first hear about us? And if all of a sudden TV goes from if you haven't done TV before, and so you get a benchmark of you get a baseline before doing tv. And we all know the problems associated with survey questions like false positives, false negatives or whatever.
So you launch TVs a channel before you do tv, you get a baseline of wrong answers and maybe you look at percentage over time or total numbers over time and maybe you get one or two or three a day. And the good thing about these surveys is they're randomizing the answers all the time. So for a question, when did you first hear about us? Which you only ask to new customers, it just randomizes the channel results. It's not perfect, but it's better than nothing. And so you do a TV ad, all of a sudden your response rate goes from one to two a day to 50 60 a day. You go, okay, well this probably has an effect on tv. And so then you can just do like, well, what's the CPA for survey filling out? So total survey is divided by TV span. So you get a post-purchase survey, a cost per post-purchase survey.
So what's that? It's not a data point. If you've only gotten three or four, probably wasn't very effective and your mm m saying probably wasn't very effective, and maybe your demand capture channels are saying, well, nothing really happened wasn't too effective. Well it probably wasn't that effective. But if you have lots of channels telling you similar things in different ways, it is probably a good channel. And that's a very, very simplified version of that. Of course, there's ways to optimize tv, time placement, channel placement, length of creative, et cetera. But I think surveys in that regard have really, really helping marry the quantum, the qual of understanding, okay, well where could our measuring gaps be? What products could help best help us serve our customers? And what experience could we think about providing our customers? It's always good to have customer insights in a world where your customers are all over the world, it makes it a bit harder to do in-person focus groups.
Evan Kaeding:
Yeah, there's only so much you can tell from clicks and taps on a screen, on a page and from checkout. I know we recently had a few guests from Manscaped on the marketing intelligence show and they were also talking to us about how they use quite a bit of the data from their post-purchase surveys to involve, to understand the effect of where their marketing runs, their product placements, product bundling strategies, all of which would be very challenging to get out of traditional impressions and clicks measurement.
Tyson Drake:
Yeah, absolutely. I mean, if super managers could find a way to integrate with these survey providers, that would be awesome. At the moment, I've got to either export a CSV and then manually put it in some reporting like my reporting that I use or I've got to manually get someone to manually enter the survey results in. Sometimes it's quicker to just punch in a two rather than filter for the report, download the CSV upload. Sometimes it's easier to enter the number. So that'd be my feature request integrate with these next, these surveys
Evan Kaeding:
Noted, noted. Shoot me a list and I'll give the feedback to the product team and we'll make sure to get it on the roadmap. Well, one more thing I want to make sure we touch on the D two C front is maybe the new frontier with ai. Of course, marketers are looking at a variety of different AI based applications, and I'm curious to hear from your perspective, are there any specific use cases that you've seen for AI and D two C marketing that have you excited?
Tyson Drake:
I think the biggest one, well actually two that I've seen, I don't think they're amazing yet, but I think I'm looking at the directionality. So the first one is data analysis using ai. So there's another attribution platform or tool called Purple Whale, and they've just recently launched an ai, I think they call it, or maybe not, where they basically query, you can query your data within the platform, you can ask questions about your data, it's an analytics platform as well, and you can speak to their ai. So that's an interesting way that you have this talking to your data, your own data in a way that maybe it helps without having to do analysis yourself. Or maybe if your direct report doesn't have to ask you a question, they can just go in there and ask the question or someone gets asked the question, I don't know how to do this analysis.
Dunno what data I need to pull. They're like manually do is just go in there, ask the question. So I think that's one way I haven't played around with it, but I've seen videos follow the founders there and I think that's quite an impressive product of what it will do in the future as well. That's like one angle, query your data. And then the other angle would be using AI for things like creatives or creative strategy in general. So creative strategy, think of it as all the functions that are required to make an ad. And so what is that? That's script writing, that is whiteboarding, that is making video ads and static image ads have script writing if it's a video ad. And so I've seen, there's lots of agencies out there in the direct to consumer space who are using these AI script writing.
I've seen some scripts where I can't tell if it's written by AI or by human. You can train the models and you can tell 'em not to use these words or use these words. So script writing for direct response creative is one area that I've seen. You can still kind of tell when something looks like an AI image, it is the hands or it looks too polished, they're too beautiful, an 11 out of 10. And so I think people are pretty good at spotting that now. However, I think in the next year or two, I think it'll probably be really hard to tell the difference. And so it's important to look at the directionality of where these technologies are going. And I think within a year, maybe two years, it'll be really hard to distinguish humans from AI for creative strategy that's going to have a lot of implications, not just for direct to consumer. I think that's probably the least people's problems wander at that stage. But I think we are going to get there and you can look at it from a, okay, well from a cost efficiency, cost reduction perspective that's like maybe that is the price of creative goes down. Great. What does that mean for budget allocation? Well, Facebook and Google is getting more spend. That might be one unintended consequence.
And I think people, you have people in-house creative strategies in-house using these tools, a lot more script writing and being more of a prompt engineer just learning how to use a tool. It's just like when Adobe first came out, you have to learn the tool and then all of a sudden it's just like the default creative tool now. So I think AI is just a tool. There's a million ways to use it. I think two interesting paths are like this, talking to your data, putting, giving an analyst power, giving everyone the power of their own analyst, but then it comes down to the sophistication of your prompting and the questions you ask and of course the model and then it is creative strategy and that is like script writing, the creatives, the art, the image, and maybe even video as well, editing as well in the future, once they figure out how to stop the people look like an AI image, the video will come and then all of a sudden we have bigger problems rather than being able to tell if something is selling you like a fake person is selling a drink bottle or not.
You've got bigger problems there I think.
Evan Kaeding:
Yeah, yeah. Well I can tell you Supermetrics is definitely watching the AI enhanced data analysis piece very closely. So certainly something we're keen on. And if the advances do come for the creative strategy, I think that's something that'll largely benefit the ecosystem generally, more creative variation, more opportunities to build authentic touch points with your customers at the end of the day is the way I see it.
Tyson Drake:
Tyson and tell stories tell really good brand stories and maybe everyone, everyone just gets a world-class brand storyteller, access to world-class brand storyteller and that's great. I'd love everyone to have access to that. I'd love to have access to
Evan Kaeding:
That Sounds like a powerful tool in the right hands. Tyson, if you will, I think something many of our listeners will be interested in is learning a bit about your role as a fractional CMO. So maybe if you could you tell me a bit about what is a fractional CMO, how do you define it and maybe how did you get into this kind of work?
Tyson Drake:
Yeah, so as a fractional CMOI perform the same functions as an in-house CMO, but I help two to three brands at a time. And so the way I do that is I bucket my days into different brands. So any given time I might have working for helping one brand for two days and helping another brand for three days or very rarely might be a brand for one day and then another brand for two days, another brand for two days. So maximum I only have three clients. One, I don't like the switching costs of having multiple clients at a time. I've worked at a few companies that were in the past. Before I was a fractional seamen, I worked for a few companies that were a rollup you could say, or a holding company, and they had multiple brands underneath them. I probably worked for four of those in my time.
You work across four or five brands a time, sometimes up to 10. And I don't like the mental switching costs of going like, all right, we're doing this or we're doing this. So brand two hours a day you work on this brand another two hours a day, you work on another brand, another brand, another brand. It's not the way I like to work. And so I bucket my days into one single brand. I just found it allows me to go deeper as well. And so main benefits are cost efficiency. So maybe not in a position where you want to hire a full-time, CMO, maybe not big enough or you don't want to invest in a C-level talent, you're just not ready to invest in full-time, that level. Maybe you don't have the work load for them either. Maybe it's like we just want someone to come in and help us out for a one to two days a week and help manage your team and do one-on-one meetings or even just help with strategy or whatever that may be.
And so that's one benefit. The other benefits are like I, I've worked across a lot of brands, a lot of different industries, everything from MRO, so it's maintenance repair online to have a hundred thousand SKUs to single hero products. You just have one product and eight figure brands, like nine figure brands, everything in between across different markets. And over time you just build up these playbooks, right? You're like, okay, well this brand over here I worked with that had these problems that I helped solve, it turns out that the solutions to them are also applicable to this other brand over here, but we just have to mold them a certain way and fit it through the brand. And so there's repeatable playbooks across different brands that I've worked with. And so I can't do three things. One thing is this measurement, measurement and analytics. So that for me, that would be helping brands with all forms of measurement.
That's like multi attribution, market mix, bowling, infantile testing, custom dashboards, customer scorecards I can build for them as well. Let's say bucket one like measurement. And bucket two is this thing I like to call professionalizing the marketing org, which is just team leadership, how to build a team. Usually brands in this phase are either using agencies for everything and they want to start to transition in-house. They just don't really know how to do that. Or maybe they've just transitioned to in-house and they're like, well, we don't really know how to manage these people because we just used agencies before and the agency used to manage their people. It's now how do we get people to work together? How do we do one-on-one meetings? What are the SOPs a standard operating procedures for how they work together and how do we run a marketing team, an in-house marketing team now?
And so that's something else, but all things around team and leadership, and sometimes I have a team in place that's functioning pretty well and they just need maybe a founder's being responsible for that team and the founder wants to go do product development or whatever they need to do. And then I kind of just replace the founder and be responsible for the marketing team. And so there's some common paths. And then the third bucket I call is growth strategy, and maybe they've hit a plateau and the brand doesn't really not do next. And so there's a bit of an exploratory processes like, okay, well there's a new markets. Is it new channels? Is it new products? What is the best path? And what do we think is a bit of an exploratory process and some recommendations and sometimes even after recommending a lot of the time actually helping roll out those initiatives.
And that can take some time depending if it's new product launches or whatever. And so they're like the three buckets that I play to. And I first came across the concept of a fractional CMO when I was working in Berlin. I was working for a venture capital firm called Project A Ventures at the time, it was 2014 that I started there and I was 27 at the time, didn't really thought I knew a lot about e-commerce in Australia and traveled Europe and backpack all throughout Europe and eventually reasons I decided to stay in Europe. My dad's British, so I had a British passport and it was useful then not so useful now because they're not in EU anymore and you don't get the same benefits. So I was eligible to work and live in Berlin with my EU passport and worked for this venture capital firm as an intern.
And I remember doing intern there and I was just like, oh my God, I don't know anything. I remember just having this moment where I'm like, I've been working in e-commerce five years in Australia, and I just dunno anything compared to this Berlin VC firm who are a breakaway from rocket internet. And the managing directors there helped build Lando and some other large European brands. And I'm just like, I don't know anything about internationalization. I don't know anything about, they're doing marketing mix modeling and all this stuff, and I'm just like bi and I'm like, I don't know any of this. I'm like, I've got to learn everything. I have to be a sponge and just absorb everything. And so I eventually became a junior. My specialty is in my performance marketing, and so I was a paid search, was my background there. Then, so the model was they did, the VC firm did three things.
We incubated our own brands, consumer specializing in consumer and marketplace. So Inc. Incubator owned brands, they put a seed ran in and would do proof of concept and then kind of roll that out. And I help the marketing side for paid search. The other part was part two would invest series A with other VC firms and then we would go into to the firms. So what I forgot to mention is we were a full stack VC firm. So I had a hundred person operational team, performance marketing, business intelligence, hr, recruitment, front end dev, backend of dev, full stack VC firm. And so I was in the performance marketing team. So the second thing we did is help our series A investments. Either we'd go into the brand and be their performance marketing team, or if they had a performance marketing team in place, we would come in help maybe upskill 'em, teach 'em about some of these concepts about ality testing or market mix modeling or maybe implement frameworks of like, okay, you've been doing seven figures now or eight figures now I want to get your nine figures, so here's how your frameworks have to change a little bit.
And the third thing is they would work with private equity firms who would either acquire businesses like EET from Sweden who would buy, just buy e-commerce businesses, like nine figure old school e-commerce businesses. And we would come in and kind of modernize them, digital transformation and then work with their team doing very similar. And so got this unique view across 7, 8, 9 figure brands at the time and for different frameworks, worked with different brands. And the head of the performance marketing team was this lady named Dorothy, amazing marketer, one of my mentors when I was at Project A, lots of respect for her. And she was a interim CMO, so she would have the role of being the CMO of the performance marketing team project day. And she would go into these when we made new investment, she would then bring select team members into this new investment that one of our portfolio companies, and then we'd report to her, we'd work underneath her. And so she was an interim CMO. And the main difference between interim CMO and a fractional cmo, I'd say is an interim CMO usually has an end date, and there may be an interim CMO for three months or for a year, but there's usually a duration to interim, hence the word interim where there's a fractional CMO, there is not usually an end date. However, apart from that, you basically perform the same function. You like this part-time. CMO one usually has an agreed end date, one usually doesn't.
So, but with me, sometimes I have end dates agreed upon end dates for clients, and that might be due to the scope of work we need to do X, Y, Z, and I estimate that to be three months, and therefore the end date is three months upon completion on the scope of work. And so I first came across the fractional CO concept at this VC firm, and I was so it's interim CO concept, and I was just like, that is the coolest job for me, just going into these companies as the marketing leader and helping them with these frameworks of these playbooks that you build up over time and just helping implement best practices, getting them up and running, whether that be SOPs to the team or measurement infrastructure or growth frameworks, teaching people, handing it over and either continuing on the execution of that or fading out and then working on another brand. So for me, that was the coolest model. And so I've been doing that for about over a year and a half now, I think. And yeah, I love it. I love the breadth of brands I get to work with and the problems I get to solve and yeah, I couldn't imagine doing anything else.
Evan Kaeding:
Sounds like an awesome role and certainly a lot of great exposure that you got early on in your career, and I'm sure that the exposure to the number of brands that you worked with and probably also the quality of brands was a huge impact on how you work with your clients today.
Tyson Drake:
Yeah, yeah, absolutely. I'd probably say maybe 50 to 60% of the things I learned was in the three year period I was in Berlin.
Evan Kaeding:
Yeah, sounds like an exciting experience. Well, Tyson, last question for you before we wrap up today is I know that you're an active champion of supermetrics on Twitter or X as many know it today. We'd love to hear maybe just a bit about how you use Supermetrics in your day-to-day, maybe with some of your clients as well.
Tyson Drake:
Yeah, so for me for is all about getting the marketing spend variable for contribution margin. So for me, I love the ability to, I have a Google sheet as my data warehouse, very sophisticated setup, and I love the ability that as a non-data scientist, as someone who doesn't even know how to use an Azure or whatever data warehouse, I use Google Sheets as my data warehouse. I just plug the plugin in the sheet and then I just connect to the different marketing platforms I need to use and I just download the data, right? It's great. And then I just use simple Google Sheets formulas to take Shopify net sales data, like gross profit data minus variable costs and platforms you can use, or you just use an estimate and then you just minus marketing spend. Then you get your daily contribution dollar tracking. Very, very simple setup that you can do the, I've been personally using this kind of setup for like five years now, and it hasn't really changed too much in the sophistication at all.
So I mean, you can go out there and pay for a data warehouse or you can just use a Google sheet and use super metrics and export your marketing spend to do some really basic math, Google key calculations to figure out what your daily profit is and then just manage it quarterly. So I love it. I've used others and I don't like them as much. I think Supermetrics has the best feature set. One of the feature sets that I really do like about Supermetrics is the currency conversion. When you export marketing spend as well, that's really valuable because the problem is a lot of advertisers export. So they might just have one Facebook account and everything's in USD and you can't export the currency for SD. Or they've had agency one set up your Facebook ads for one country and then Facebook account two set up targeting another country, but the currencies might be USD still. So you've got these currency conversion issues that you have to account for if you're managing p and ls for different businesses. Two metrics just makes it really easy to, that's just one feature that I use a lot is just the easy currency export in whatever the currency is available there. Super useful.
Evan Kaeding:
Love to hear it. Thank you, Tyson. I'm sure many of our listeners will take some inspiration from today's podcast. Anything you'd like to leave our audience with before we wrap up?
Tyson Drake:
If you'd like to learn more about me or what I do, I'm very active on Twitter. You can find me or x Tyson Drake, check out my website, Tyson drake.com. If you want to learn more about what I do, if you just want to chat, I'm pretty open to talking about anything direct to consumer.
Evan Kaeding:
Excellent. I hope our audience members will take you up on that. Thanks a bunch, Tyson, you've been an excellent guest today. Much appreciated.
Tyson Drake:
Awesome. Thanks for having me, Evan. Thanks to the Supermetrics team.
Evan Kaeding:
Of course. You bet.
Edward Ford:
Thank you for listening to the Marketing Intelligence Show brought to you by Supermetrics. If you're enjoying the podcast, then we'd love for you to tap that subscribe button, leave a review, and share with your colleagues and peers. We'll see you in the next episode.