Apply for a Free Strategy Call

Ecom Growth Insider · Podcast episode

Carl Weische: Fix AOV First, Then Conversion Rate

With Carl Weische · Published 2026-04-24 · 48 minutes · Hosted by Andrej Tumachowitsch

Carl Weische is co-founder of Accelerated Agency, a CRO agency for DTC brands. His team works with AG1, Huel, Vessi, The Oodie, and IM8. In this episode he explains what to fix when ROAS or MER drops as you scale spend. He assumes the creative and media buying are already sound. Inside the funnel he starts with the offer. Conversion rate comes second, and presell pages come last.

Carl's bundle SOP is three tiers on the product page. The first tier is the product on its own. The second costs 20 to 30 percent more. The third costs 50 to 80 percent more. He says that lifts AOV by 10 percent at minimum, and usually 15 to 25 percent. Before any funnel work, he asks one question. Can ads to the product page reach 100k a month in revenue at 10 percent EBITDA?

This one is for DTC founders and operators who are spending real money on ads. It fits brands stuck at a scale ceiling, where more spend just lowers MER. You will get Carl's audit process, his checkout leak list, and his sample size rule. He also names the stage where an in-house CRO team makes sense.

What you will take away

  • Clear the creative first, then fix the offer Carl has a media buying expert check the creative and the platform numbers first. If that side is sound, he calls it a funnel problem. Then he fixes AOV before conversion rate, starting with the offer.
  • The three-tier bundle SOP Carl replaces the single hero product with three tiers. The first tier is the product itself. The second is a bundle 20 to 30 percent above it. The third sits 50 to 80 percent above. He says this lifts AOV by 10 percent at minimum, usually 15 to 25 percent.
  • Presell pages come after a ceiling Carl adds presell pages once a brand stagnates on ads straight to the product page. He uses an advertorial or listicle format. He moves fastest with brands that rely on one, two, or three hero products.
  • The 100k test before any funnel Carl's first check is simple. Can ads to the product page scale to at least 100k in monthly revenue at 10 percent EBITDA? If not, he says the offer needs fixing before anyone builds a funnel.
  • Revenue per user is the north star Carl judges tests on revenue per user rather than conversion rate or AOV alone. His example: a 6 percent conversion rate at a $50 AOV returns $3 per user. A 10 percent rate at a $10 AOV returns $1.
  • Four reasons carts get abandoned Carl names missing trust, no urgency, an unhandled objection, and open questions about shipping, returns, or costs. He says trust and urgency start in the ad, long before the cart.
  • Research is 80 percent of the offer Carl says 80 percent of the offer is research. Perceived value comes from understanding the buyer, not from discounting. In one test his team took a phrase straight out of customer reviews into the website copy and lifted conversion by more than 10 percent.
  • The sample size rule for split tests Carl's rule of thumb is 20,000 users and 1,000 orders per variation. That means 40,000 users and 2,000 conversions for one A/B test. Below that volume he validates with ad campaign tests instead.

"If you're not able to scale in the simplest way possible, meaning ads to PDP with your offer, then probably your offer sucks."

- Carl Weische

Jump to a chapter

Full transcript

Lightly edited from the episode captions for readability. The episode is the record - where they differ, the audio wins.

00:00 - Why Your ROAS Dies When You Scale

Andrej: If your return ad spend or MER drops the moment you try to scale spend, this episode is going to be a game changer for you. Today's guest is Carl Weische, founder of Accelerated Agency, one of the top CRO agencies in the DTC space. Trusted by brands like AG1, IM8, Huel, the Oodie, Vessi and over 200 more. He works with companies where the stakes are real: high ad spend, high expectations. And now let's see whether this works.

Carl breaks down the actual levers that unlock profitable scale. How to fix AOV and conversion rate without killing each other. What exactly to look out for when high-intent customers drop off in cart or in the checkout process. And when it's time to introduce presell pages to make colder traffic convert. Let's get into it.

00:43 - The Real Job of CRO in 2026

Andrej: Carl, most founders think CRO is just making some tweaks and making some small optimizations to improve the performance. You position it more as scalability and really as profits. What would you say is the real job of CRO in 2026?

Carl: Good question. I like this as an opening question for the podcast. First of all, I really want to thank you for inviting me to the podcast and enabling this conversation. So really excited to speak to you today and also get your thoughts on things. For me, CRO - I've been in the CRO field for six or seven years. I started in 2019, now going into 2026, so almost seven years on the dot. And for me it has always been something really fundamental of understanding the way that people work. So understanding consumer psychology, specifically understanding how to sell to people. And for me, CRO was just one way of doing that.

And I think a lot of people think about CRO, as you said, as some minor tweaks: changing the color of a button, changing a little bit of text here and there on the page. Whatever. But for me, the way that I view it, it is one of the biggest levers, or the biggest lever at least, on everything funnel.

So from the first touch point after they click on the ad until the thank you page, which is a lot of ground to be covered. Especially nowadays where we have a lot of funnels that start on a presell page, then go to some form of product page, then go into a checkout, then have some post-purchase upsells and then the thank you page. So that gives us a lot of different pages where we have a lot of room to optimize for.

And then I don't view it isolated. For me it's one part of the equation where the paid acquisition team has to do a good job to drive high quality traffic at the lowest cost possible to the page. So then the online store or CRO team has a good job at converting most of their people at the highest AOV possible, meaning driving up revenue and profit per user. And then the retention team has to do a good job at making sure they have a high customer lifetime gross profit. So for me it's just one part of the equation, but it is a super important part of the equation.

I think more and more people start to realize how big of an impact it is if you have a higher conversion rate, higher AOV, because then you can spend more and everything becomes more profitable. But again, I'm not sitting here saying that it's the one magic pill of a business to make it more profitable. It's just one part of the equation that you have to dial in as a business owner, similar to how you have to dial in paid acquisition and retention.

Andrej: And what made you decide to focus on that part of the engine, bro?

Carl: For me it was something that happened naturally. So I used to run my own Amazon FBA business and Shopify stores back in 2017, 18 and 19. And then I met Lars, who is now my co-founder of the company, and he back then was a freelancer for conversion optimization. And then I hired him to work together on my page and my Amazon FBA business. And I was just really fascinated by it, because I was always fascinated by psychology, but more on a personal level, where I just read all of the different psychology books on the market. I went through seminars from Tony Robbins, I did all of this stuff trying to understand how people work.

And then Lars basically showed me how to use all of this passion in a professional context, specifically CRO. So for me it was just the perfect intersection of my personal passion and then my marketing passion coming together in one field, being CRO. So that's how I got introduced to the field. And then when I started working with Lars, I just saw so much potential, not only in CRO being a huge lever for brands, but also the market.

Because back then in 2019, 2020 when we started, it was a complete blue ocean. There was a handful of service providers internationally. Even today, obviously a lot of people started becoming CRO freelancers or whatever over the years, but even today it's still a blue ocean. So I'm really happy to have made that decision to go all in into this specific field.

Andrej: Yeah, definitely. And you work with a lot of different brands and also huge brands, eight figures and beyond.

05:04 - Carl's Decision Tree When ROAS or MER Drops

Andrej: When a brand comes to you and they basically tell you our return ad spend or our MER just drops when we scale spend and we can't grow aggressively, what does your immediate decision tree look like? How do you decide what the issue for them is?

Carl: Really good question. So on the one hand side, the only thing that's a bit outside of my scope of expertise are the specific media buying KPIs. So I would always double check that with somebody who is a media buying expert, in terms of, are their current creatives and media buying strategies any good, or do they need to fix that first? But let's just, for this example, put it as a given that the creatives are great and the media buying KPIs look good. But then it's a funnel issue, right?

And then the two things that we typically focus on to improve ROAS and MER is number one, higher conversion rate, number two, higher AOV. And then number three is a bit of an add-on, which is introducing presell pages so we can run more top Of funnel, creators, right?

06:08 - Fix AOV First: The 3-Tier Bundle SOP

Carl: But first of all, let's focus on conversion rate and AOV first, because that's the easiest. With AOV, the first thing that we do is we check the offer, right? So typically, most brands, what they do is they have a hero product. Let's just say they have a supplement. One supplement, hero product, $50. And then they typically run a discount of, say, 20, 30% off. And then that's their offer, right? And they run all of the ads to the product page.

Now the first thing that we would do is we would take that offer, rebate it, to bump up the AOV by 20 to 40%, typically repackage the offer. So let's just stick with this example of a supplement. We would typically go from a single hero product to some form of bundle, where potentially we would have two or three supplements, even adding a digital product or component to the offer, make it a nice big bundle, have different tiers.

So we typically like to work with three tiers, where the first one is just the product itself. The second one is then a bit bigger bundle that's typically 20 to 30% more expensive. And then the third tier is typically 50 to 80% more expensive. So that you have three tiers on the product page as different bundle options.

And we literally always follow this SOP, and it just works across the board, because then a lot of people go for the middle tier, some people go for the low tier, and some people go for the high tier. So typically, just by doing that and moving from selling one singular product to giving them different tiers and different bundling Options with free gifts, you're able to increase the AOV by 10% minimum.

07:45 - Fix Conversion Rate: Images and Copy That Resonate

Carl: In most cases we see 15 to 25% higher AOV just by doing that.

Secondly is then the conversion rate that we need to improve. So for conversion rate, the biggest thing we see are product images and copy, because that's typically where you either resonate with the traffic or you don't resonate with the traffic. So what I always like to say is do your research, understand your target audience, and then based on this understanding, make product images and copy that really resonate with the target audience, with their pain points, motives, beliefs, desires.

So basically with the situation that they're in right now, the situation they want to be in, and how your product helps them go from where they are to where they want to be. If you're able to translate that into creative assets, for example product images, GIFs and videos on the product page, and copy, you're going to have a better conversion rate. So those two are the biggest leverages typically.

08:33 - Layering in Presell and Advertorial Pages

Carl: And then the third lever, which is more of an add-on strategy that we do with a lot of brands, where let's say they hit a glass ceiling in scale with bottom of funnel ads to product page. We then like to integrate a presell page, which is typically an advertorial or listicle format, so that they can run more cold traffic, top of funnel traffic, less aware traffic to a presell page, warm them up, and then send them to the product page. So that's typically an add-on which then also enables them to spend more on ads profitably, just because they are now able to convert people they haven't been before.

Because if you run unaware traffic to PDP, it's just not going to work. So I would say first of all fix AOV, fix conversion rate, and secondly build more funnels.

Even now, you're the expert for that. But with the Andromeda update on Facebook, everything is going into the creative doing the targeting, and then it's the same with the funnels. So for some of our biggest brands that spend multiple millions per month, we literally have over 100 different funnels live to match the ads. Influencer A, funnel A. Influencer B, funnel B. Offer 1, funnel 1. Offer 2, funnel 2. And then we just have hundreds of different ad creatives, and then almost for every single ad creative a distinct funnel with one hundred percent congruency.

Andrej: Yeah, that's crazy. So many funnels is crazy. But it makes sense. At a certain scale you need to segment the audiences and who you're speaking with, and have the congruency between the ad, who you're targeting, the ad, the funnel, the offer. That all has to work together.

And I think it's really interesting that you said the offer is the first thing, because even from a paid ads side of things, that's also the first thing that we look at, or that we basically try to get the brand to optimize, or help them optimize it. Because nowadays you can have the best ads in the world. You can do everything on the creative, on the setup side of things. But if the offer is not great, it will not work.

11:00 - When Presell Pages Actually Make Sense

Andrej: I'm sure you remember, four or five years ago you could have the most basic commodity offer, and if you had some decent ads and a decent website, you'd be able to scale and get sales. But nowadays, with all the competition, it's way more difficult.

Carl: Yeah.

Andrej: You mentioned also the presell pages. How do you usually decide whether it makes sense to test the presell page, or at which stage do you decide to test the presell page? Because we have some brands that from the beginning are basically using listicles and advertorials, and that is their main funnel and their main selling strategy. And then other brands that manage to scale to eight figures without using any presell pages or any advertorials. How do you decide that it makes sense to try it out?

Carl: Good question. It's really different brand per brand, so it's really difficult to give one answer that everyone listening can follow. Typically we like to start implementing presell pages as soon as we see a certain stagnation, or them hitting a glass ceiling with their ads. Especially if brands have a small product catalog, meaning they are reliant on 1, 2, 3 products that they sell, then that's typically when we really quickly start implementing different presell pages.

Because the way that I think about it, similar to what you said earlier, is that nowadays it's almost always some form of equation of the person who runs the most volume wins. Meaning the most volume in terms of creative output, the most volume in terms of funnel or split test, the most volume in terms of different alphas and different products. So the fewer products you have in the catalog, the even more you have to focus on creative output and funnel output. But the bigger the catalog, in theory you could just continue running ads to product page.

So it's really difficult, again, to just give one answer to this. But we see this a lot with brands where they just have 1, 2, 3 hero products, that we just go super hardcore on funnels. Because then we need to find a lot of different ways to sell them to a lot of different target audiences, with different creative types, different funnel types. But if we have, let's say, a supplement brand with 100 products, then it's way more that every single product has a certain scale, and we just need to optimize that in terms of the creators and then in terms of the specific product page, for example.

Andrej: That makes sense. And is there a specific metric that you look at to decide whether the offer works and it just needs to be, for example, repackaged properly, or repackaged in a different funnel, or whether the offer is still not the right one and the offer needs tweaking?

Carl: So first, before I would start building any funnel, the first principle is: are you able to scale this profitably, just ads to product page, to at least 100k in monthly revenue? And with profitably, I mean minimum 10% EBITDA. So if that's not a given, if you're doing the bottom of funnel to PDP break even, then you shouldn't even worry about funnels, because then you have a fundamental thing in your business that's not working. So that's what I would always do first.

Meaning that if you're not able to scale your offer with just ads to your PDP to at least 100k monthly revenue profitably, then you seem to have some form of fundamental issue in the creative strategy or in the offer. Which means you need to fix your offer first. If you're already doing a couple hundred thousand per month with that setup of ads to PDP, and it's all super profitable, then please start integrating funnels to scale further.

So with everything in marketing, whether that's paid acquisition, conversion or retention, people always jump into the nitty gritty stuff and they try to analyze: okay, I need to have 2.2 ROAS, I need to have this marketing efficiency ratio, this conversion rate. And I'm always like, bro, get yourself out of the business and start thinking about first principles. First principle thinking. If you're not able to scale in the simplest way possible, meaning ads to PDP with your offer, then probably your offer sucks.

I don't even need to know the specific KPIs of your business. I just know that if that's not the case, then your offer sucks. And that's what I would always do as the founder or entrepreneur of a business. Try to look at everything on a first principle basis, without directly going into specific KPIs or metrics to make decisions.

Andrej: Yeah, that makes sense. And if a brand is not even able to convert the traffic that they send to the product page at a low scale, or at least at a decent scale, then, as you mentioned, either the offer is wrong or the product is just bad.

15:51 - Revenue per User: The Real CRO North Star

Andrej: And it's not even worth spending the time and effort into trying to scale something that doesn't work. What is a misconception that you see founders have about AOV in comparison to the conversion rate? Because oftentimes, obviously, when you try to increase AOV, the conversion rate drops. If you decrease the AOV, the conversion rate can increase. How do you balance that?

Carl: Good question. For us, the North Star metric is always revenue per user or profit per user, meaning just the combination of conversion rate and AOV. So the way that I think about it is more of a sweet spot of how high can we push conversion rate and AOV to have the overall highest revenue per user. Because let's say 6% conversion rate, $50 AOV, is a $3 revenue per user. And then, whatever, a 10% conversion rate and $10 AOV is $1 revenue per user.

So you just need to figure out what the sweet spots are, without looking again at the metrics isolated, but rather at the overall health of the funnel, which for me is the revenue per user.

Andrej: Yeah. And then if possible, probably even go deeper into profit per user, if you're able to pull that data, right?

Carl: For sure. For us, that's one of the most fundamental things. I don't know if you noticed, but we have our own data intelligence and CRO software, which is closed beta now. So we're just releasing it to market later this year. But for us, one of the main things was integrating the Shopify contribution margins and COGS that you have in the Shopify backend, together with all of the split testing data, to then analyze all of the funnel performance and the split test not only on the revenue per user, but also the profit and profit per user.

So for any entrepreneur running a business, I would always, whether you use a certain software or not, figure out a way on how to combine all of the different data sources to look at everything on a profit level.

Andrej: Yeah. So you created a tool specifically to be able to pull out all of that data, because there was no native way to do that before.

Carl: Yeah, we've used our own software for almost four years now. A bit over four years. So we started developing that in 2022. It was something where I was like, I have a service based business, I use a lot of external softwares. But then all of these external software providers have some issue. They don't have a certain feature, whatever. So at some point we just decided to build our own software. So literally just for the use case of making our agency better and reducing headache for our team, yada, yada, yada.

And there was one huge thing to figure out, like how to combine all of the funnel data with then also all of the contribution margin data, to understand not only the analytics, but then also specifically the funnels we run or the split tests we run on the profit per user basis.

Andrej: Yeah, I think for a lot of brands, having that data is a huge, huge unlock, or even something that we -

Carl: And then there's thousands of use cases that you can build. As I said to you earlier, I'm so deep with my team and all of the different AI use cases and whatever. So for example, one thing that we built is, I don't know if you know Microsoft Clarity or whatever, but those tools where you can just get the heat maps, the click maps, the session recordings. So what we figured out is how to connect that data then to the profit per user data.

So you not only see where people click in terms of general clicks, but you can essentially match that with people who made a purchase and then also look at what was the profit for that purchase. So you don't have a general overview of data, which is already great, like heat maps and click maps are great. But you take it one step further of identifying what are specifically the clicks that made me the most profit, what are specifically the customer journeys that made me the most profit, and what are also, on the other end, specifically the customer journeys where I lost the most. Or the products where I lost the most.

And yeah, so I'm just super excited with all of the possibilities now, where you can just take a lot of different data sources, merge them, consolidate them, and that gives you 10 times better insights than what we used five years ago, in terms of the KPIs to then come up with better funnels or better split tests. Yeah.

20:19 - The First Things Carl Audits on Any Store

Andrej: So exciting, with everything that is possible right now. I'm curious: when you audit the store for the first time, what are the things that you usually look at first?

Carl: Good question. For me, what I always like to do is put myself into the shoes of a regular consumer. I will literally just go on the Facebook ads, look at the ad creative, click on it, and then go through the whole customer journey like a regular user. And then identify the steps in the journey where I am confused or missing something. Let's say, just for example, I click on the ad creative and in the ad creative they talk about a certain pain point. Then they push me to an advertorial and it's a completely different angle. Then I would just write down: okay, missing congruency from ad to advertorial.

Then I would start reading the advertorial and potentially see that this doesn't resonate with me at all. So they need to do better customer research to identify pain points, or whatever. So I literally just like to go through the whole customer journey end to end, buy the product, take all of the upsells, to understand where the areas are where I stopped, or what was confusing for me, or where something didn't make sense, and just write all of that down.

Andrej: And what are the most common leak points that you see? Specifically in, let's say, eight-figure brands.

Carl: Congruency is a huge one, from the ad to the page. So typically they have a certain angle, customer persona, or offer in the ad, and then they have a completely different thing on the page they push the traffic to. So that's number one. Number two, there's oftentimes something that's confusing, whether that's the flow of the customer journey, certain buttons, whatever. So there's always a certain part in the customer journey where consumers are confused, where they don't understand what's the next step, where they have to click, how they can check out now, whatever. So that's number two.

And number three is then just more generally your UX/UI: things like, is it easy to understand the product benefits? Is it easy to understand the value proposition? Is it easy to understand what this product does? Is it easy to understand the offer or the bundle? So I would say those are typically the main aspects where brands fuck it up.

Andrej: Yeah. And I assume you nowadays focus primarily on the mobile version, right? Do you even still look at the desktop version of the websites?

Carl: We do have some brands that Due to the demographics, we have a huge part on desktop.

23:00 - Why Customers Abandon the Checkout

Carl: But for the longest time we've just focused on the mobile views.

Andrej: And let's say you go through an analysis and you see that a website has a huge drop-off rate at the checkout. What are the top five reasons that usually happens?

Carl: Typically they lack some form of trust. So they don't trust you enough. They don't have enough urgency to continue. There's some form of objection that you didn't handle. There's some form of question around shipping policy, return policy, specific costs. Let's see, I cannot come up with a fifth one. But those are four big things, specifically in the checkout stage, where they will then abandon the customer journey.

Andrej: And do you think those things should be addressed in the cart and in the checkout process, or is that oftentimes also an issue that was from way before, from the ad, the landing page?

Carl: Way before. Everything starts with the ad, especially trust and urgency starts in the ad. And then everything more specifically like objection handling is happening on the sales page. And then everything around frequently asked questions or anxieties, then cart and checkout process, like communicating the delivery times, communicating the shipping policy, communicating the returns policy. And everything like that you need to do in the cart and then also the checkout.

Andrej: I think it really depends obviously on the brand and what they are already doing. But if you would have to give an e-commerce founder some quick wins and some quick things that everyone needs to implement in their funnel, and specifically also in the cart and checkout process, what would you recommend?

Carl: I would say the number one biggest leverage is the offer. So jump back to the beginning of the podcast where we talked about offering the different tiers. So number one, fix the offer on the product page with the different tiers. Number two, especially in the cart and at checkout, make sure that you incentivize if they went for a lower tier like tier one, to then take the upsell into tier two or tier three.

And in the cart you can do that with progress bars where you say, hey, one click to get a different tier to then get free shipping or free gift. And then you can do the same in a checkout. And then the second big thing in cart and checkout would then be the urgency. Everything like countdown timers, cart protection and everything like that, just to really push them through to purchase. So I would say that's by far number one, to fix the offer.

Andrej: If we go a little bit deeper into the offer, because I think that is, as I mentioned, the most important thing that a lot of founders Still get wrong.

26:07 - How to Raise Perceived Value Without Discounts

Andrej: Even though there are a bunch of great books, a bunch of resources on how to come up with a great offer, they still think that it doesn't apply to e-commerce, or doesn't apply to the same degree as for every other business that exists in the world. From all the experience that you have, what are the best ways to increase the perceived value of the offer, and ideally without killing the margins of the business with very aggressive discounts?

Carl: Great question. I would say 80% of the offer is research. I would say 80% of any function in the e-commerce business is research. So meaning customer understanding, because perceived value is just something that you derive from a great customer understanding, and being able to identify the specific pain points that you need to solve with the offer, and then also the specific desired situation that your target audience wants to achieve with the offer.

So it's literally just 80% understanding the target audience, and then the other 20% is figuring out what are the products you need to give to them, what is the unique mechanism that you say to them, and what are the price points that you can request based on the perceived value. So I would say it's just a function of spending a lot of time with the target audience and then translating that into - I always like to say current situation, desired situation, offer is a vehicle to get them from where they are to where they want to be. So you need to do the research to figure that out.

And then the offer packaging is just combining all of that. Right? Combining the offer with all of the pain points that it solves of the current situation, and all of the bad emotions that they have. And then also combining the offer ingredients or features with all of the dreams and desires they have of their desired situation.

And then making it a super logical way of how the offer, or more specifically the unique mechanism, is something completely new - as the word says, new unique mechanism - which is now able to help them, and also doing a great job at explaining why this is different to anything that they've seen before.

And then the offer is directly going to have a high perceived value, because it's an all-in-one solution to a huge problem that they are facing. And then they are able or willing to spend a lot of money for it.

Andrej: Yeah, I definitely agree that the research is super important, whether it's for the ads that we come up with, for the offer, for the customer acquisition. In every single aspect of the business you need to dial in your research, and it becomes easier and easier. Nowadays with AI you can analyze a shitload of data very, very fast, and what used to take us days and weeks we can now do in minutes or hours. But I still think that a lot of founders think that they don't have to do any research at all anymore, because either they were Not doing it before and they're still not doing it, or they were investing a little bit of time before and now they invest even less time, because they think with AI they can just ask ChatGPT what are the three pain points, and that's it.

29:00 - Why AI Is Splitting Great Marketers From Average Ones

Carl: I'm so glad that we have AI, because the way that I view AI is it's just going to be an even bigger split between the best marketers and the average marketer. All of the average marketers are going to do that with ChatGPT, and they're just going to do even worse creatives, funnels and everything, and the best marketers are going to become even better. So for me, AI is almost going to differentiate between somebody great and somebody average even more, and make it more clear.

And I think the people who are great and are already in the top 1% are just going to have an exponential advantage over the average marketers, or the people who are now starting to learn marketing with AI. Because I feel like a lot of people just externalize all of the ownership and all of the creativity and strategy to AI, which is the worst you can do. And I think a lot of people are going to be really surprised when they figure that out in the next coming months and years.

Andrej: Yeah, 100% agree. Can you give us a brief overview of how your research process looks?

Carl: Sure. For us, we do have full-time data specialists who just do research. In the current team we have five or six full-time data analysts that just do research. They work with a multitude of quantitative and qualitative data analysis methods, client specific and then market specific.

Client specific, what we look at is typically everything: Shopify data, then all of those heat maps that I was talking about earlier, all of the customer reviews, all of the - we generally call it message mining - meaning anywhere where customers have spoken to you, whether that's in the reviews they give, and emails they send to your customer support team, in comments on the social media, in DMs. So that's all client specific, surveys, yada yada yada.

And then market specific is everything like competitor analysis through Amazon reviews, through their website, scraping everything. We find niche specific areas where their target audience is spending time, like Reddit, Quora, certain online forums. And then we just have a lot of those data sources, client specific, market specific, and that's all of the databases on what we then use to come up with the specific customer persona or specific copies, split tests and all of that stuff.

Andrej: Yeah, I really love analyzing or using what the customers say about the product in the messaging and in the ads and the funnels. And to be honest, I think I've heard that from you, or from one of your LinkedIn posts a while ago. But basically take all of the reviews that the website has, that the brand has, take all of the post-purchase answers, every single touch point or every single communication with the customer.

And then analyze that through AI to find out what are the terms, what are the sentences, the phrases that your customers are using to describe the product, and what kind of problems they are talking about. Basically really digging deep into that data, because you get a lot of insights from that. And obviously, ideally, if you use the exact messaging, the exact language that your customers use to describe the product, the conversion rate will increase.

Carl: Yeah, it's crazy. We had this one experience like four years ago, and I think that's the LinkedIn post you were referring to. We had a client in the fashion niche where they were selling these pillow slides, which is a super crazy product since then. And we just went through the testimonials, meaning all of the client reviews. And then it was super interesting, because the biggest thing that all of the clients would say is, it's like walking on a cloud. And we were like, we never heard that before, it's nowhere on the website.

Why don't we run a split test and put it as a copy on the website? And then we just said, this product fits like walking on a cloud. And just that copy, adding that to the website, increased the conversion above 10%.

And that just opened my eyes on how it's so easy to literally just go ask your customers, or just go through all of the reviews, and then understand what is the language that they're using, what are the specific words they're using, what are the metaphors, analogies, how do they speak and how do they speak about the product. Because as a founder you're so blindsided - you have a deep understanding of the specific product, all of the features and benefits, that you oftentimes oversee this.

And then that has literally become the main value proposition, the biggest angle in their creative strategy and everything. And it was overlooked before because nobody cared about what the customer said, right? And that just really opened my eyes in terms of what's possible, or why it's so important to look at what the customer is saying.

Andrej: Definitely.

35:00 - When a Brand Is Actually Ready for Serious CRO

Andrej: I think there are two different kinds of brands.

Some are way too early in their journey and they think they need to spend a lot of time and effort tweaking every single thing on the website, making those small improvements, small optimizations to get better results, even though they don't have enough traffic to make any meaningful tests or get any good data.

And then on the other side, there are those kinds of brands that are already generating a lot of revenue and they think they don't need CRO testing because they're already getting results, getting sales, and they just want to keep focusing on their paid ads, keep ramping up the ad budget, and doing all of that. What would you say is the perfect position, where a brand should be, where it makes sense to focus on CRO?

Carl: For me, CRO are different things, right? We have two fundamentally different ways of working with seven figure brands than we have with eight, nine, ten figure brands, right? So when we work with seven figure brands, meaning anybody that's doing from 50,000 to a million a month, we work with them on their general funnel setup, their offers, copy, product images, all of the big things, right? Fundamentally what makes the business work.

When you work with 8, 9, 10 figure brands, 90% of the work is split testing, meaning they come to us and they say, hey Carl, we have this one advertorial funnel. We spend a million a month on it. Do anything you can to get the last percentage points of performance out of this funnel. That's the only use case why they hire us. They did everything they can do internally. Now they want us to just increase the conversion by the last 5% and the AOV by the last 5%, because that's huge for them at that scale.

So then for those brands, the biggest thing that we do is we set up a high velocity split testing program where every single month you run a lot of split tests on all of the important pages.

So yeah, seven figure brands typically struggle with the fundamentals of finding a functioning offer, a functioning funnel and all of that. Whereas 8, 9, 10 figure brands struggle with completely different things. They have a lot of funnels that perform amazing and now they need to iterate it, fine tune it without fucking it up or without breaking anything, right? So I would say everybody needs to do some form of CRO. It's just, what is the specific form of CRO that makes sense for the growth stage of the business?

37:09 - Testing Cadence at Scale: 5 to 30+ Tests per Month

Andrej: And for those eight, nine and ten figure brands, what would you say is a realistic testing cadence for those brands? I think you mentioned monthly. Is that how often you usually do it?

Carl: It really depends. For the smallest amount it would be 5 tests per month. Right? 5 split tests per month that we run. For other brands, we test 20, 30 funnels plus then 10, 20 split tests. Right. So it's really, really based on how much volume they are pushing, meaning how much ad spend, and then how much data we have, meaning how quickly can we iterate and test new stuff.

So it's just really, really depending on the volume. That's the number one denominator of how many tests we run or how many funnels we test. It's how much ad spend and how much data do we have from the brand. And then obviously, a brand that's doing a million a month is going to be completely different to a brand that does 10 million per month, or a brand that is doing 50, 100 million per month.

38:12 - The Sample Size Rule: 20k Users + 1k Orders per Variant

Andrej: Do you usually decide it based on the traffic that they're getting, or the number of conversions? Because obviously there are some brands that do a lot of revenue at lower sessions, lower traffic. And there are some with very low AOVs, very high AOVs. How do you make sure that the tests are really valuable and you don't get false positives?

Carl: It's always a mixture of sessions and conversions.

Andrej: Is there a minimum amount of sessions that you need for a test, or is that tough to say?

Carl: There are rules of thumb that you can follow. Typically, what I like to say is, absolute minimum: you should have 20,000 users, not sessions, users per variation, plus a thousand orders, so 1,000 conversions. So overall for an A/B test you should have a minimum of 40,000 users and 2,000 conversions for one single A/B test. That's a general rule of thumb, right?

Sometimes in one A/B test we have a million users and 30,000 conversions, just because the effect is so minimum or whatever. So if you're below that volume, then we rather run an A/B test with ad campaigns, right? Meaning you take a certain budget, like 2, 5, 10 Gs, and then you rather set it up as a quote unquote Facebook A/B test rather than an on-page A/B test. If you want to validate a new funnel, then that's how I would test it.

Andrej: Yeah, that makes sense. And for all of those tests, what is the true North Star that you're shooting for? Is it always the revenue per session?

Carl: Completely different, bro. Typically it's something like revenue per user, profit per user. So that's typically the North Star. But then also we have big fashion brands that do hundreds of millions per year, and then their biggest thing is something like customer lifetime gross profit or return rate. So it's really depending on the niche. Or we have big software companies where it's sign-up rates, or time spent on page for brand awareness. So it's really, really depending on the niche.

But typically the North Star is something around revenue per user or profit per user. Just for that specific niche where their main focus is on, let's say, the retention of the backend, then it's rather customer lifetime gross profit or something like that. Or how many months did they stay on the subscription, how low is the return rate, things like that. So it's really based on niche and then business model, I would say.

Andrej: Have you ever had it happen that, for example, the conversion rate went up, the revenue per user went up, but then the retention of those customers went down, and with that the lifetime gross profit of those customers declined?

Carl: Sure. A lot. There are so many things that you can do on the front end that are then going to have a negative impact on the back end. Let's just say supplement product. We do a shady subscription in terms of the UX/UI of the buy box and the cart and the checkout process. And then the conversion rate is higher and we have a higher subscription take rate because we do it a bit shady. But then a lot of people don't understand that they opt into a subscription, then we have a lot more churn and a lot less customer lifetime gross profit.

So that would be just a super easy to understand example of how quickly something that you do in the front end has a negative impact on the return rate, customer lifetime gross profit or something like that.

Andrej: So when you do the split tests, or work with a brand, do you also look at the lifetime value? Because obviously the revenue per session or the profit per session with one test could look significantly higher. And then the founder is like, yeah, let's put everything on that one. But then after a few months they start to realize the revenue is going down or the retention is not there.

Carl: Always. Yes, especially for those niches where it makes a lot of sense, we always monitor the data while the A/B test is live, and then after we've implemented it, not only for the then new page performance but then also for all of the cohorts of the A/B test.

So I would say there's three areas: the data of the A/B test while we're running it, the data on that specific page and product after you've implemented it, and then also the cohorts from the people of the A/B test to understand subscription take rate, customer lifetime profit, how many months they've been on the subscription, all of that stuff. Yeah, to understand The overall business impact.

43:11 - When to Build a CRO Team In-House

Andrej: What do you think of brands that want to do CRO in-house?

Carl: I think it makes a lot of sense at a certain growth stage. We've had a lot of businesses where we've worked together with them for two or three years. Started with them when they were a 5 million per year business, and then they grew to a 100 million per year business. Started when there was just a founder and five people, and then they became a 100 people team. So at a certain growth stage of the organization, the business, it makes even more sense to have an in-house team than an external service provider.

And then we literally just build them the in-house team. We give them our software to run the A/B tests, and then sometimes we give them a bit of consulting on how to hire the team, train the team, manage the team, or then some check-in calls on how to use the software and make sure that they are doing everything perfect. But at a certain stage of the business and organization, the value equation is better on an in-house CRO team than on an external service provider.

Andrej: What is the minimum effective CRO team or org chart that a brand needs?

Carl: Our state of the art version of a CRO team is five people, where you have one specialist for each part of the process. So that's, in my perspective at least, how a brand should build it. You have one data specialist who understands data, one conversion specialist to understand the CRO program, split testing and all of that, one UX/UI designer who can build the variations and the split tests or pages, one developer to program the A/B test, and one quality assurance person to make sure that everything is running without bugs on the website.

And typically those five people are the team, and then you need some form of head of paid acquisition, CMO, growth manager to manage the team and the program. So that's typically the team. That's our in-house team. And that's also always the team that we built for the brands when they requested to have it in-house.

Andrej: And at what stage would you say does that make sense? Is it in the mid eight figures, nine figures?

Carl: Yeah, typically mid to high eight figures. So most of the brands started doing that at around 50 to 100 million per year, when they moved from us as an external service provider to an in-house team. Some did it at 50, some at 70, some at 100. Really just based on their overall organizational structure.

It's also a question of this team. Based on the location you hired, it costs anywhere from 20 to 50,000 per month, based on what country you hired that talent from. So overall it needs to make a lot of sense for the brand, where they have a significantly higher impact of that team running it in-house than a service provider who costs 10, 15,000 per month.

46:15 - Rapid Fire: Underrated, Overrated, Biggest Mistake

Andrej: Yeah, that makes sense. Now to end this, I have a few rapid fire questions prepared. So just answer them quickly, with what comes to mind.

Carl: Sure.

Andrej: What is the most underrated CRO lever in 2026?

Carl: Offer split testing.

Andrej: What is the most overrated CRO tactic?

Carl: Payment providers under the call to action button.

Andrej: Really? Does it not make a difference?

Carl: Sometimes, but it's, yeah, as you said, overrated.

Andrej: What's one change you'd make for a brand that is stuck at 200k per month and is trying to get to 1 million per month?

Carl: Offer or presell page.

Andrej: One change for a brand at 1 million per month trying to keep scaling profitably?

Carl: CRO program.

Andrej: And the biggest "I can't believe people still do this" CRO mistake?

Carl: Ask ChatGPT for ideas.

Andrej: What about Shopify's new copilot thing?

Carl: I would say the Shopify one is a bit better, just because it's based on your specific data. But then, the way that I think about it, if a corporate Company pushes a feature that has to apply to a huge total addressable market and be super generic, meaning you as a direct-to-consumer brand that's doing seven to nine figures are not going to have a lot of value from a feature that has to apply to you.

48:11 - Where to Find Carl

Carl: Yeah, a huge ton.

Andrej: Yeah, that makes sense. Carl, it was awesome. Thank you very much for hopping on. Where can people learn more about you?

Carl: I would say the best source is my YouTube channel. You've seen some videos yourself, so that's why. I just jump super deep into everything funnels, offer, copywriting, CRO, and also share a lot of case studies really openly.

Andrej: Awesome. That sounds great. Thank you.

Carl: For sure. Thanks so much.

Get the next episode. Ecom Growth Insider ships biweekly - long-form operator interviews and short solo lessons, on YouTube and every podcast app. Browse all episodes.

Carl's decision tree covers what happens after the click. The spend before the click is a separate question. Start with a free 30-minute call: the $5,000 Profit Clarity Audit.

Apply for a Free Strategy Call

Apply for a Free Strategy Call