Apply for a Free Strategy Call

Ecom Growth Insider · Podcast episode

The Pricing Lever Most DTC Brands Ignore | Drew Marconi

With Drew Marconi · Published 2026-06-24 · 49 minutes · Hosted by Andrej Tumachowitsch

Drew Marconi is the co-founder and CEO of Intelligems. His company runs price and shipping tests for ecommerce brands, and this episode is about price as a profit lever. Before Intelligems he worked on pricing at the ride sharing company Via. He says he spent four years there building a dynamic pricing engine. By the end it set prices for roughly half a million rides a day.

Drew walks through the math on a $100 product at a 10% contribution margin. Raise the price by $10 and the profit per order doubles. He says you could lose half your conversion rate and still end up with the same profit. He also shows how to pick a free shipping threshold. Look at where order values cluster, then push people to the next cluster.

This episode is for DTC founders and operators who set a price once and moved on. If you run paid ads on thin margin, price and shipping are levers you control directly. It is also useful if you sell across multiple markets. Drew explains why the same price test can land differently in each market.

What you will take away

  • Price flows straight to the bottom line Drew's math starts with a $100 product at a 10% contribution margin. Adding $10 to the price doubles the profit per order. He says conversion could fall by half and profit would land in the same place.
  • Ads get ten times more attention Drew puts the effort ratio at about ten to one. Ads and creative get that much more work than price and the on-site offer. Ad budget feels like the closest lever to CAC, so that is where teams go.
  • Pricing is downstream of your strategy Drew's first question to a brand is what they are pricing for. Buying customers at break-even to feed LTV sets one price. Getting EBITDA up before a sale sets another price for the same business.
  • Set the threshold off order distribution Drew reads the histogram of order values and finds where baskets actually cluster. He sets the threshold so the jump to the next cluster is reachable. That keeps shoppers from stalling in the cart.
  • Nines and round thresholds still hold Drew says pricing just below a threshold like 50 or 100 has worked every time his team tested it. Prices sitting at 101 or 105 are the easiest miss to spot.
  • The right price is a mirage Drew has seen the same price increase land differently in two markets. It did nothing to conversion in the US and dropped it in Australia. Season shifts the answer too, because a different buyer shows up with a different need.
  • Checkout is the underrated profit lever Drew's pick for the most underrated lever is the checkout page, both messaging and upsells. He says teams work on it but rarely test different versions. His team sees a lot of uplift there.
  • Personalized offers land better than surge pricing Drew avoids the term dynamic pricing, because buyers hear surge pricing. He says personalized offers, new-customer deals and seasonal changes are already accepted. He also says list price should not move on personally identifiable information.

"Just try stuff. It only gets harder to change and iterate the longer you wait."

- Drew Marconi

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 - The Pricing Lever Almost Everyone Ignores

Drew: Our top value at the company is customer impact. We want to do stuff that helps our customers make more money. Hearing that, it no longer feels like sales. It's just, hey, I can legitimately help you make your life easier and make more money in your business. So the surprise pretty quickly gave way to we should go build this, and it's been building ever since.

Andrej: Quick question. When did you last change your price? Not your ad creative, not your landing page, not your email subject line. Your actual price. If you're like most founders, the honest answer is you set it once a long time ago on a spreadsheet and haven't changed it since. Drew Marconi has managed over $600 million in transactions across over 1,000 A/B tests on Intelligems, and he says the single biggest lever in most e-commerce businesses is the one everyone ignores.

Here's what made this conversation click for me. If the average e-commerce brand running at 10% contribution margin raises their price by $10, they can double the profit per order. They could lose half of the conversion rate and still come out the same. Most founders have no idea the math works that way.

Before Intelligems, Drew was chief of staff to McKinsey's global managing director. Then he spent four years building a dynamic pricing engine for half a million rides a day at a ride sharing company. He's one of the most rigorous thinkers on pricing that I've come across. And he's not talking about hacks. He's talking about building the muscle. This one is packed with specific frameworks your brand can use today. Let's go.

01:43 - From McKinsey to Pricing 500K Rides a Day

Andrej: Hey, Drew.

Drew: Hey, how's it going?

Andrej: Going great. What about you?

Drew: Can't complain. It's finally getting warm here in New York.

Andrej: Nice. Nice. You were the chief of staff to McKinsey's global managing director. That's effectively the CEO of the largest consulting firm in the world. And now you're helping DTC brands figure out where their shipping threshold should be, $50 or $75. How did that career path happen?

Drew: It's a great question. It's always easier to explain looking backwards. I don't think any of it made sense along the way. This is not where I expected I'd be. So I started my career at McKinsey. I got a pretty good grounding in strategic thinking and a lot of analysis. I think people underestimate the amount of data work you have to do at the consulting firms. I did that chief of staff role for a year, which was awesome, to work for the CEO. I traveled to, I think, 36 countries that year, including Poland, and really loved that role. I did a lot of internal work and speech writing and project management.

I came back to normal consulting work. I was living in San Francisco, so I got staffed on digital projects and building apps for older companies out on the west coast. I was working with designers and product managers. And I was like, "Oh, this is awesome. This is what I want to do." And I really wanted to go to a place where I can make the decisions and have to live with them and iterate on them, because in consulting your job ultimately is to make sure the client's happy with the work and you hand it off. Sometimes you get to implement parts of it, but you're really not accountable for the decision that was made.

So I jumped into tech by happenstance. I ended up at a ride sharing company called Via in New York, which went public last fall, and only ended up there because randomly a teammate from one of my projects had ended up as a head of growth. A friend from college was there as well. So I was like, "Hey, these seem like good people. Everyone is really smart who I interviewed with. I'll go and learn stuff here."

And I joined the growth team, where our job was to get people to ride more frequently. We were trying to build commuter habits, because it was a shared ride service. It was like $5 a ride anywhere. Pretty quickly, we realized that flat pricing was not going to scale and needed to build a dynamic pricing system. Uber and Lyft had already done it. We needed to better match our costs to the price of the rides. We needed to do things like surge pricing. We needed to be smart about where we gave discounts and the total right place, right time moment.

Me and my current co-founder Adam were both on the growth team. They're like, "Hey, can you guys start to figure this out?" So we were just sitting at a whiteboard. How should we build a dynamic pricing system? And we spent four years just making it a little bit better each day, to the point where by the end we had a team of 30 around the world. And we were setting prices for probably half a million rides a day that incorporated things like weather data and event data, supply and demand, and also things like customer LTV and prediction.

It was just so much fun to build that system because it is very impactful. You make changes and they show up. We loved how data driven it was. We could always use more data to make this better. And we would constantly be running experiments and tests to figure out what was working. And we were both kind of glutton for punishment. So we really liked that pricing is never solved. There's no right answer. The market changes around you. Maybe it's right for today, but you're constantly just getting a little bit better, a little bit better. It's never done. You're building a system that can improve itself.

05:45 - The Pivot: From Mobile Games to Shopify Pricing

Drew: So, we spent four years doing that. Covid hit. A lot of different things changed, both with our company and the world, and Adam was moving. We were like, hey, we want to keep working together and we know all this stuff about pricing and offers and elasticity. Where can we make that useful? And so I don't talk about it that often, but Intelligems started as a dynamic pricing for mobile games company for about six months. We were like, yep, it's the perfect market for this: small transactions, we're logged in on the app.

And we pretty quickly realized that was not a great target market for us. It turns out that microtransaction in-app monetization for games is a pretty late stage thing. If you're a game developer, most of the early money comes from ads and platforms like AppLovin. So as a game developer, you just need to build a loop of gameplay that hooks people for 48 hours. You're literally looking at 48 hour retention. Can I retain people for two days? At that point, okay, great. You can build from that. You can sell ads.

Only once you have a really established user base are you worrying about the economy inside the game, which also means you're more likely to be acquired or rolled up into a big studio. So really the only people doing in-app monetization are these seven or eight huge studios. It wasn't a big fractured market, and they're 80% engineers and don't buy a lot of SaaS. Coming from outside the industry, there was also just a lot of suspicion. So people didn't want to talk with us. Small customers who would say yes didn't have the data that we needed.

So just in the process of trying to sell it, we were like, this doesn't quite feel right. And I'm a huge gamer. I love video games, but I hate microtransactions as someone who gets taken advantage of by those. It just didn't feel good to be thinking about every day. And someone told us, you have to love your customers, and they have to be people that you could see yourself working with and being friends with for a decade, because that's how long you want to build a company for. And yeah, we just didn't feel it.

Andrej: And then you transitioned to e-commerce right away?

Drew: Yeah. So we stumbled backwards. We had thought about e-commerce and we were like, surely people must already have tools for this. Surely people must be using Optimizely to test their prices. These are digital transactions. Of course people are doing this work. And then we had two friends, one who had started a brand on Shopify and one who was starting one, and they were like, oh no no no no. We completely put our finger in the wind and guessed that price, and I don't have any idea how I would get data about what someone is willing to pay.

And we talked to more people and talked to more people and talked to more people, and probably spoke to 100 brands in the first few months. And almost everyone was like, I know I'm not pricing well, but I have no idea how to fix it, and it seems like a painful problem to solve. So I just leave it, if it's not horribly broken. And we said, hey, we can do better.

And so we went to market thinking, even then, it was going to be this algorithmic, hey, you give us your parameters, you give us the guardrails, we will set the prices for you. And it turned out people wanted control. So they were like, no, no, I just want to know, should this mug be $10 or $15, and what happens to demand and therefore my revenue and my profit? So we just followed what the customers wanted. We built a testing tool for pricing, and that came out in 2021.

And from there it's just snowballed into this much bigger problem we are solving, of how do you extract the maximum expected profit from the traffic that goes to your site. And pricing is a big part of that. It determines what people buy, it determines how profitable those orders are. We call it profit per visitor, this metric. I can increase that by selling more deals, more conversion rate, selling bigger deals, purchases, AOV, or more profitable and margin. And pricing was just one component of that.

And we realized, oh, shoppers, they're looking at the shipping rate. They're looking at returns. They're looking at images. They're reading the copy to identify, is this high value or low value, something I need. They're building bundles and adding discounts. And we talked about how we had to go beyond the page and expand our vision of what we were looking at.

And that has led to the Intelligems today, which is like, yeah, we can help on the pricing, we can help on the shipping, we can help on the copywriting, see where people are falling off the funnel. Because it needs to be holistic. We can't solve pricing in a vacuum, just the same way that changing button colors in a vacuum doesn't generate the same results. So it's been this snowball with features and expansion and more customers for now five years.

11:14 - Why Most Ecommerce Brands Still Price on Gut Feeling

Andrej: Yeah. And I'm curious, you came from Via, where you were analyzing millions of data points and adjusting all the prices in real time. And then you stumbled into the e-commerce industry, where everyone was just deciding the prices based on gut feeling. How was your reaction to that?

Drew: Excited. Surprised. Hey, wow, okay. Turns out not everyone's doing dynamic pricing. We were further ahead than we realized. Great, I know how to do this and we can add value. That's a great feeling as a founder, when you're like, hey, this person has an important problem and I actually can help them solve it.

Our top value at the company is customer impact. We don't want to sell vaporware. We want to do what helps our customers make more money. So hearing that, it no longer feels like sales. It's just, hey, I can legitimately help you make your life easier and make more money in your business. So the surprise pretty quickly gave way to: we should go build this. And it's been building ever since.

Andrej: Yeah, that's awesome. And the gap between what, for example, Amazon can do - similar to what I imagine you were doing at Via, they have millions of data points, they're analyzing everything in real time, changing millions of prices based on signals that they're getting - the gap between that and what a $3 million direct-to-consumer e-commerce brand can do is obviously enormous. And I think Intelligems was built to basically bridge that gap and help smaller businesses be able to do testing. From your perspective, how does that look in practice?

Drew: Yeah, we use the word democratize a lot. We want to democratize smart commerce. We think that brands should be continuously learning from their storefronts. They should be adapting to each customer, and each transaction should be a little bit more efficient than the last. This is a digital medium. We can measure everything that's happening, and in our view there's no reason why every brand shouldn't be able to use a digital medium to the fullest extent.

But if you think about how many stores work, it is just a copy of brick and mortar. Everyone lands on the homepage. It looks the same. You have men's on the left and women's on the right and sale at the back. It's fixed. It's that same experience. We're not really evolving it. We think with digital you can rearrange it to produce the most profit for each visitor. So that's what we're trying to do.

And can we take these skills and expertise that have been buried in large organizations that can afford to have really large teams, and make that available down market? There is a limit, because we need some data from the store to optimize. But then also, over the last couple of years, it's been: even at companies that are doing this, they think that they're too slow. They need a data scientist, they need a marketer, they need a designer, they need a developer. They can only run one or two tests a month, it's all these internal approvals, and they don't know what to test next.

And so we also want to make life easier for those teams that have started this, but maybe aren't as high velocity or high impact as they want. It's an interesting thing to manage as a software founder. I want to build the simplest possible version of the product that delivers value for people who are like, yeah, I just want to come in and see how this works, you tell me what to do. But then also having, at the bottom of the iceberg, all of the robust technological foundation, data foundation, so that we can equip teams that are already doing this to do it faster and better.

15:34 - What Gruns Gets Right: A Storefront for Every Customer

Andrej: I find it interesting that you mentioned that a lot of e-commerce brands are just like brick and mortar businesses. They have the same homepage, the same products and the same customer journey for every single customer, and how with digital we have the opportunity to be way more personalized, way more custom. And I see that with a lot of brands that are currently winning and that are getting great results, that every single aspect of their sales funnel is really custom-tailored to the end user and to the customer.

A great example for that is Gruns, which is, I think, the fastest growing supplement brand. And they just got acquired for 1.2 billion. They're running hundreds of different landing pages for every single narrow avatar that they're targeting. They have a completely custom-built landing page and completely custom-built ads and everything, and make it as custom-tailored as possible.

And that is what allowed them to grow super fast, because when the user sees something and he feels like he resonates with every single piece of copy, creative and everything, that is obviously way higher impact. And I think a lot of e-commerce brands are just not utilizing that.

Drew: And Gruns is a customer of ours. They recently switched to us because they found us to be very fast to iterate on that stuff. Great, we've got a thousand different storefronts now. How do I get a feedback loop of data, making sure it's performing, making sure it's converting at the right balance of one time versus subscription? And so, yeah, we're pumped to work with them.

But I think of it as, rather than having one store at the mall, you have a thousand mini stores, each of which is catered to, hey, this is a parent who's looking to have their kids eat healthier. Hey, this is a young professional who is concerned about their gut health. Hey, this is someone who just wants to have more greens. And it's like, all right, this is the exact layout of the store. Here's what the salesperson is going to tell you for each of those personas. And it's a simple concept, but executing it well has historically been really difficult.

18:02 - Why Founders Optimize Ads 10x More Than Price

Andrej: And what I noticed with also the brands that we work with is that the majority of the brands are happy to spend a lot on testing a lot of different creatives, testing 50 different variations. But then they set their prices once and don't change them for two years. Why do you think that happens?

Drew: Oh, good question. I think generally you look at the amount of time teams spend optimizing their ad spend and ad creative, and it's probably a ratio of about 10 to one versus what they're optimizing on the site, be it price or the landing pages or the offer. I think there's a couple of reasons for it. Number one, ad spend is the budget. And so a lot of the resources are close to that budget. And hey, if we can operate at this CAC, we'll scale infinitely. Spending time in the ads is the closest lever of control, or feels like it, to your CAC, which is what everyone wakes up and is checking throughout the day.

And yes, we could sit here and say, well, obviously the on-site experience and the price and the offer matter a ton, because if we can convert more people, or more profitably, we can lower the CAC, or we can be more profitable and afford a higher CAC and scale more. But just that leap - it's a different platform, it's a different set of metrics - is surprisingly non-obvious and hard to prioritize.

The second reason is that pricing just is scary to play with, often for founders, where you have heard customers complain about price and it's sensitive. It's, this is my baby, my product. I want to make sure I'm delivering value. And so there's an emotional resistance and fear around pricing, and then an operational fear, because pricing impacts so many things. Are there price tags? Is this going to flow down to my warehousing system? What am I going to do with my Google feed? Will that update automatically? My customer support team needs to manage it.

So price is, yes, just one number on the website, but it flows and touches nearly every part of your business. So if you don't have a muscle of changing it and iterating on it, it's a pain in the butt to flow through, or something's going to go wrong.

And then the third reason I think why people don't change their price as often is there's not good information, good data or good playbooks on how you should think about it, right? There's so much content created - podcasts, courses, tweets, LinkedIn, ebooks - about how to build better creative, how to set up your Meta accounts to work on. There's a whole industry around optimizing that. I'm the only person I see out here talking about pricing strategy. No, no, there's some people who focus in on offers.

And so I just think a lot of founders are like, great, I don't think we're doing it well, but I literally have no idea where I would start, so I'm just going to focus on the problems that are more familiar and known to me.

Andrej: Yeah, that makes sense. I think a lot of founders just don't know that there's even the option to test different prices. They think that if they want to change the prices, it's a huge operational change that is also very difficult to switch back again. And that's why they rather don't do it at all.

21:46 - The Math: How a $10 Price Bump Can Double Profit

Drew: And I think the math for how powerful it is isn't necessarily intuitive. But any gains you can make on pricing or shipping revenue, for example, they do go straight to the bottom line. So let's say you're running at a 10% contribution margin and you're selling a $100 product. So you pocket 10 bucks from every one that you sell. If you could sell that for 105, or let's say 110, your contribution margin on that is now 20 bucks. So you've doubled the profit per product or per order, from 10 to 20.

Now we raise the price. There's going to be a conversion drop-off. Maybe CAC goes up. But you could suffer a 50% drop in conversion rate and end up with the same profit at the end of the day, because that extra $10 of price just flows straight down to the bottom line. And that math isn't intuitive. It's the same way of, great, if I can get 10% of people to pay me $5 on shipping, that's 50 cents per order straight profit. It's sunk cost and paying for shipping anyway. This is gravy.

And so I think it's slept on, how powerful the lever is. And that's a very simple example. There's lots of reasons to keep conversion rate high, and there are customers who you can get LTV from. But every time I draw those charts, people are like, "Oh, wow. That's cool." Actually, do you mind if I show it? I was just showing someone a slide on this the other day.

Andrej: Yeah, sure. Share it. But yeah, I know we've been doing the same thing with some of our clients. We basically charted out or have shown them what kind of impact it will make on the profit margins if we're able to sell the same products at $5 more, $10 more, and also how much more aggressive we could be on the ads side and how much our break-even return on ad spend changes. And it is always mind-blowing to everyone.

Drew: Yeah, this is just the example of what I just laid out with slightly different numbers. And this is from our seed deck in 2021, and still is what we work on. There's a ton of stuff on the given site. I think people underestimate how many touch points there are that involve pricing and involve how people are willing to pay. Most people just guess. I'm very proud of the fact that I raised venture money with SpongeBob in the deck. And if you get this right, it's super powerful.

Andrej: Yeah.

24:25 - The Pricing Mistakes That Quietly Cost You Money

Andrej: What is the most common thing that you see when you first look at a brand's pricing and you immediately know that that is costing them money?

Drew: The easiest one to spot - it's very different brand to brand. So it's a little hard to pull out that many global lessons. A very obvious failure mode is if someone's pricing stuff at like 105 or 101. Nines work. Pricing something just below thresholds like 50 or 100 actually does work every time we've tested it. If I see that people are - I think this is more rare now - just giving away free shipping, that's not the most profitable outcome. I guarantee you.

What I usually ask first, and it is a very telling answer, is: what are you pricing for? What is the goal for the company this year? What is the biggest rate limiter on your business? Is it CAC has got to come down? Is it we've got to make profit? Is it we have to drive more LTV? Is it that we need to introduce more product lines so that we can get beyond a first purchase? Because pricing is downstream of your strategy.

The right price may be very different depending on if we're trying to, for example, get customers in the door at break even so that we can put them into our retention and LTV engine after. Or, hey, we're getting ready to sell the business in a year or two and I want to get my EBITDA as strong as possible. Those are going to generate very different prices for the same businesses. So very few people have a clear, coherent answer on that the first time I ask. And then from there it's like, all right, well now let's go see if the prices support this strategy and this goal.

26:47 - How to Actually Find Your Free Shipping Threshold

Andrej: And you mentioned a big mistake is just doing free shipping. What is your strategy, or recommended strategy, on finding the perfect free shipping threshold? I assume the best way is to just test it and see what works. As a starting point, we usually recommend the brands that we work with just look at the AOV and then add 10 to 15%, sometimes 20%, to try to push it up. What have you seen?

Drew: Yeah, I think that direction you just shared - take the AOV and add a bit - is mostly right. I'm trying to see if I can pull up a chart here, because this is actually a chart we have. So let me share my screen again. The reason why it's not just take the AOV and add X% is we want to see that there may be particular spikes or cliffs in the order distribution.

It's not just that if you have products priced at 20, at 50, at 70, mathematically there's going to be different types of basket sizes created, and you'll see that. So we often look at this distribution of order values, and we'll notice on this histogram here's where there are particular spikes. There's a spike around 50 bucks, a lot ending right at 80, a big drop off, and there's another set that is between like 100 and 120. And we want to set a threshold that is actually achievable for people.

If my AOV is 50 and I make the free threshold 60, but my products cost $50 and $30, that's just really annoying to a customer. They're like, well, I have this $50 item in my basket, I need to spend 10 more to get free shipping, but the only thing I can buy is another 30 bucks. That creates distraction. It creates a bounce. People are searching around for the right thing. So we like to look at this distribution and roughly look at what's a bit above my AOV, but basically see, can we push people from one of these cliffs to the next?

If anyone wants to look at this data, this is in the sitewide analytics tab. You can also ask our MCP server to give you the order distribution. But we want to push people to the next reasonably frequent basket size. And then there's a whole bunch of strategy downwards from that. All right, are we merchandising this properly? Do we have the bar in the cart? Are we telling people you're only X% away from free shipping? We could add badging to products throughout the site of, hey, this one's eligible for free shipping if it's more expensive than the threshold, or if it gets you over the threshold.

A lovely little trick is we have upsells at checkout, and let's upsell a product right there that gets someone free shipping and gets them just over the line. So it's like, let's test around, let's build a hypothesis around this AOV distribution, pick a threshold or two to try, find that. But then within that, there's all of this behavioral stuff to drive even more conversion and AOV boost among the customers. Sorry, I went off the rails there for a bit, but shipping thresholds are one of my favorite topics.

Andrej: Yeah. No, you're totally good. That's so interesting to me, because it's obviously a very big lever to change the AOV and LTV of the customers, and also something that a lot of brands just set once and then never change. It's just sitting there, and for some brands it has literally no impact at all, because every customer hits the free shipping threshold no matter what they buy. And there is no incentive for them to spend more.

Drew: Yeah. And there's also personalization opportunities. It may make sense to offer free shipping to new customers at a lower threshold than returning customers. You're trying to derisk the purchase for those new customers. It may make sense to have different policies for subscribers versus not. So we do see some brands get more into the weeds there.

31:22 - Why the Right Price Is a Mirage

Andrej: I've seen you write about a brand that tested the same price increase in multiple markets. One was in the US and one was in Australia. In the US it had no impact on the conversion at all, and in Australia there was a big drop in the conversion rate. What does that test tell us about how we think about price sensitivity?

Drew: To me it's just an example of the right price being a mirage. In that case, two markets behaved differently - two geographically different markets. But I also see brands that test one price in May going into the summer, potentially their high season, and then test it again in the fall coming off the high season. They get very different results, because it's a different set of customers with different needs.

Let's say it's a swimsuit company. The spring folks are buying as part of their summer wardrobe. The fall folks are buying to get ready for a vacation to somewhere warm. Those people are coming with different willingnesses to pay, different types of needs, different ways they want to be communicated with. So we do a lot of international testing, because the answer is often different market to market. That includes things like: do I bundle that into the list price? Do I charge for shipping, or just make it free shipping above X and make the prices far more expensive? There's a lot of considerations and different preferences and markets there.

But to me it's just a good reminder that you got to not just test this once, but build a system by which you are gathering this data to be able to make good decisions and iterations over time.

Andrej: And does that mean for you that every brand should test prices individually in every single market, every single country? Or what is the best approach?

Drew: You got to figure out if it's worth your time. If you got a country and you're doing a hundred orders, just pick something and roll with it. We're building agents that will do that for you in the background, but if it's something you have to spend time thinking about, you got to focus on where the opportunity is biggest.

So in every market where you have reasonable volume that is testable, and if you find an extra 5 or 10% makes a meaningful difference to your bottom line, go do it. The answer per market will be different. But don't just go through everyone for the sake of doing it if you don't actually have traffic and orders.

34:25 - How Intelligems Uses AI to Run Your Tests

Andrej: And how exactly are you using AI at Intelligems to help brands with that process?

Drew: We have a lot of cool internal use cases that have really changed how we run the company, which I can talk about. For our brands, for our customers, I think there are elements of it throughout the product. So if you think about what it takes to test and build a good experimentation and personalization program, it's: well, I have to come up with ideas.

I have to validate those with data or debate. I need to design a new version. I need to code a new version. I need to put that in Intelligems and then QA it. And then I need to run and observe the test. And I need to analyze the data and figure out what to do after. And there are small AI elements we have deployed across that entire flywheel.

So some of the things that have been most impactful: number one, we've had this MCP server, and we have APIs that people can hook into, and they will build automations, or they'll build a Slackbot that tells them how their test is performing and how that's going. We actually put a lot of that power into a Slackbot. So any of our customers now can add the Intelligems bot to their Slack channel and just ask it stuff. It can be: hey, how do I set up this sort of test? Can you look at this for me?

But it also can be: hey, can you give me reminders every morning with a summary of my test? You can ask it, what's my order distribution? How should I think about what shipping thresholds to test? So it's well beyond a customer support bot. It is like a teammate of yours who is helping run the testing program, pull data, help you come up with ideas. And it's how people work already in Slack, just chatting with teammates.

The second thing is embedded in that, but it's our experiment summaries. We share a lot of data, right? And so I'm like, whoa, okay, what should I do? Okay, great. We can have the agent go through all the cuts so you don't have to flip through all the filters, and identify what segments are winning, what segments are losing, is this test ready to be over yet, and what should you roll out?

And then the one that is really cool: I think our test building experience has always been a bit of a friction point. How do I move things around on a page? How do I generate a new experience? We have a vibe code builder now where you can say, "Hey, I want to move my reviews section and move around sections. I want to take all of my flat images and try them as the leading images on this collection page instead of the lifestyle images."

You can describe anything to it and it will make a new version of your page that you can then tweak as you like. So it compresses the time to launching tests down so much. And then we're always keeping an eye on your data. So pointing out opportunities where you just have stuff that's wrong that you should fix is something that we're rolling out, as well as: hey, this is potentially an area with opportunity to go test.

37:46 - The Future of Personalization and Dynamic Pricing

Andrej: Where do you think that is going over the next 5 to 10 years? Do you think every customer will have a completely unique experience, completely unique pricing, everything custom-tailored to them? Or what are your thoughts?

Drew: It's a great question. I think sites will be far more one-to-one than they are today, right? Everyone's been talking about one-to-one personalization for 15 years, and everyone gets the idea and likes it. But very few brands actually do it, because in reality what that means is creating 10, 20, 500 different versions of your creative, of your page. You then need to manage those across changes.

And pretty quickly you start making these segments, and they're small, and as a human it is not worth my time to get a 10% lift on 5% of my traffic that is this segment. It's a 5% improvement to the business, but I have bigger fish to fry, and there's dozens of those, and people don't want to do it and maintain it, which makes sense. These teams are small. You have to think about the marginal return of your effort.

With agents, that effort cost goes to zero. So as long as you feel good about the work and approve it, it can go run these dozen. It can maintain a hundred or 500. It's very good at reading and tracking many more things than a human could do. Not as good at this point on the strategy and not necessarily on copywriting, but it can handle the executional side of that. And so I think what Gruns does will be far more the norm, and enabled by different tools.

I don't know if it's generative. I don't know if it is truly being generated on the fly. Some things might be. I may be able to dynamically generate you an offer and say, here's the discount if you buy this, if you do this, here's the deal. It's almost a negotiation with the customer. The content being generative, I don't think will happen, because it's not fast enough. We know that speed matters so much in shopping experiences, and we can't wait. The models will get faster, but we can't have a pause in getting the information that we want.

So I think it will be far more personalized, but a lot of rule-based and probabilistic sorting of different options, rather than generating everything from scratch. I think offers will become much more tailored and personalized, prices as well, but there's trickiness there. We don't want to be doing surveillance pricing. There's different levels of comfort. But when it comes to the offer, very greenfield, open space.

Agentic purchasing is interesting. I think some people you talk to would say, hey, it's going to be agents buying everything in 10 years. I'm skeptical there. I think especially as an American, people love to buy stuff. Shopping is a pastime. It's a hobby. It's therapy. It's fun. And the types of brands we work with, it's brands. It's things that are clothes, things that are going in my body or on my body. People want to know what that brand is about.

So the storefront will look different. The agents may help with discovery, but I don't think the actual purchasing will be done by agents, with the exception of some things. It's like, hey, I need a new outlet for my wall. Okay, just go scan through Home Depot and figure out what it's going to be. So yeah, that's what I anticipate.

And throughout the life cycle, I think that change will hit email and text too, of how much more can we know, and how can we get down to segments of one on how we get LTV out of folks.

41:56 - Is Dynamic Pricing Fair to Customers?

Andrej: Yeah. You mentioned that pricing is a little bit more sensitive, more difficult. You studied ethics, politics and economics at Yale, and there is a huge live debate whether dynamic pricing is good for the customer or just a way to extract the maximum value from people. Where do you actually land on that?

Drew: Yeah, I think there's a lot of fear and a lot of misunderstanding of what dynamic pricing actually is. I think people hear dynamic pricing, they think surge pricing. They think my burger is going to be twice as expensive at 10 p.m. versus 7 p.m. And I think that we don't like that idea. We don't like surge pricing. But if you and I were sitting next to each other and got different promo codes for Uber Eats, we'd be like, "Okay, cool. Let's just use the better one." Personalized offers are very accepted.

Ticket pricing, okay, Ticketmaster has its issues. Dynamic pricing is very accepted and normal. Of course slightly different versions of this product are going to have different rates in travel. Very, very normal and accepted. So I think the tide is coming. Things have been priced dynamically forever. If you have a guy at a market yelling out different prices and negotiating with people, that's a form of dynamic pricing.

So I think the fear is understandable if people don't want to feel like they're being ripped off for information about me. And I agree. I think we shouldn't change list price based off personally identifiable information. But for the consumer, it ends up getting you more relevant offers, helps you find products better. You're going to get discounts in some cases if they think you'll be a good customer who will buy from them again, and creates more net value. So it's complex.

I really try to avoid the word dynamic pricing because it has such a heavy connotation for people. Whereas a personalized offer, a new customer offer, seasonal price changes and clearance, those are concepts people are familiar with that are powered by data to make better decisions for the business.

And I also don't think it's a good solution to be like, hey, businesses, you can never change your price, you just have to guess once and then put it in the market. Okay, if we want a ton of businesses to go out of business, that would be a good policy. So right now it's wrapped up with the AI fear and there's a lot of noise, but I anticipate more nuance will come to the conversation over time.

Andrej: Yeah.

44:43 - Lightning Round and Where to Find Drew

Andrej: To end this episode, I have a few lightning questions. So just answer them with one sentence, as quickly as you can, whatever comes to mind. What is a brand that you admire and that you think has genuinely corrected pricing, and doesn't have to be an ecom?

Drew: I think Comfort, the hoodie brand, does really cool stuff with inventory based pricing and pre-order slow sellrough. You go on the PDP.

Andrej: Yeah, I love that.

Drew: Different colorways are priced differently, and basically the only other person I see doing that is Nike, and it's a really powerful tool.

Andrej: Yeah, you can get the products cheaper if you're willing to wait a few weeks. That is such a smart move. What is the most overrated pricing strategy in DTC?

Drew: The adding shipping insurance automatically, I hate as a customer. And we've tested it. Sometimes it works, but it doesn't always work. But the auto-add, it's icky.

Andrej: Yeah. What's the most underrated profit lever that almost no one is testing?

Drew: I think the checkout page is really underutilized, both for content and messaging you put there, as well as things like upsells. A lot of people work with it, but they're not testing it, and they're not having different versions, and we see a lot of uplift there.

Andrej: If you could tell every 5-million DTC brand owner one thing about pricing, what would it be?

Drew: Just try stuff. It only gets harder to change and iterate the longer you wait. So build the muscle of changing it and trying different offers, trying different prices.

Andrej: And lastly, in 3 years from now, what does Intelligems look like, and what problem are you solving?

Drew: Yeah. I think how SaaS looks will be tremendously different, and it will feel a lot more like a teammate and visit you where you're at. You won't need to come to app.intelligems.io to access your stuff. You want to text us, you want to email it, you want to work with us in Slack. It's to bring the software to you how you want to engage with it and use it.

I think it will look more algorithmic and dynamic at last, where it's not just AB tests, but there are things that are systems that are solving things on the fly while you sleep. And customers will trust us with that. And then lastly, I think we're going to be far beyond the world of DTC and Shopify. I mean, the problems we're solving are relevant for not just any retail commerce brand, but really any website that's trying to generate a certain action. So that's a direction that we're very excited to expand into.

Andrej: Awesome. That's exciting. Yeah, Drew, really appreciate you hopping on. It was super valuable. Where can people follow you and learn more about you and Intelligems?

Drew: Yeah, you can come to our site, intelligems.io. Or you can check out intelligems.ai, which focuses on all of our AI features. I'm on LinkedIn, Drew Marconi. You can follow me there. We put out a newsletter weekly. And Twitter under Drew Marc, or I guess X now.

So follow us. We put out a lot of content about this stuff, and guides, and ways to help people just run the business more profitably and make the most of Intelligems if you are a user. And yeah, Andrej is a partner. He has discount codes he can hook you up with. And yeah, we're always happy to chat with folks who are getting this stuff out.

Andrej: Awesome. Great chatting.

Drew: Cool. Thanks, Andrej.

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.

Price is one lever you control. The other is knowing what an order really earns after ad spend. The audit rebuilds your real CAC, contribution margin, and blended MER outside the ad platforms. 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