James Tyler

I get what you’re saying, but I think seasonality is more of a product engineering and pricing problem than a law of nature. To me, a business fights seasonality in two ways. One, it grows fast enough that even its off-season periods keep compounding in absolute terms. Two, it engineers enough product breadth that there is always something relevant to sell during the so-called off-season. Take a winter-heavy brand as an example. If the business is growing fast enough (in product breadth and velocity of advertising), its summer volume can still exceed what prior periods looked like simply because the company itself is compounding quickly. And if summer is still weak, that usually points to a product breadth problem. In other words, the brand has not engineered enough throughput-driving product for the off-season. That is why I see seasonality as a product engineering problem more than anything else. If you sell disproportionately more in winter than summer, then you need to figure out what product(s) can carry demand during summer rather than treating the calendar as an excuse. Then layer on pricing. If you are a winter-heavy brand, your winter goods should still be dynamically priced in summer at the willingness-to-pay level that lets them move at the pace you want, instead of just accepting lower demand as fate. And if you layer in pre-order, that becomes another lever. You can let customers buy the in-season good at a discount during the off-season for future delivery, which helps pull demand forward, drive ancillary revenue, smooth throughput, and keep the business in more of a steady state rather than letting the calendar fully dictate the shape of demand. So to me this comes down to price, product, and inventory management. If you are growing fast enough and you engineer product breadth well enough, there is no reason you cannot grow at the pace of market growth (or faster) every single month. I have a hard time imagining a situation where you cannot engineer your way out of this through dynamic pricing, better product breadth, and near-perfect inventory management. This playbook has worked 100% of the time we've instituted it at the biggest scales in DTC.

There’s actually a bigger question to answer first. Before you hire someone to “do marketing,” the first thing to understand is your TAM, meaning total addressable market. In plain English: how many people would find your product useful if it were free? You probably already have an intuition for this. You likely have some sense of what percentage of the U.S. population, or even the world, would genuinely want the product if price were not part of the equation. Then I’d think about three price points: high, medium, and low. Of all the people who would have happily taken the product for free and loved it, what percentage of them would still be willing to pay at each of those prices? Once you have that in your head, you can start to estimate the total revenue ceiling of the product. For example, let’s say you made a widget and you believed that 1 in every 1,000 Americans would buy it at a $30 price point. Using a U.S. population of roughly 340 million, that implies about 340,000 buyers. At $30 each, that is about $10.2 million in potential revenue. That does not mean you get all of that tomorrow. It means you now have a rough sense of the size of the hill you are trying to climb. Your job is basically to ride the S-curve up that hill. And depending on your own reach, you may already have enough distribution to fund the first tranche of that climb yourself. In your case, X alone might be enough to validate demand and finance the early stages. This matters even more if the product is physical and not digital, because physical products require inventory. If you sell something for $30 but it costs $12 to produce, then a meaningful chunk of that revenue is already spoken for. You can use that math to understand how much capital it would take to manufacture the total opportunity, and then how much gross profit might be left over to fund advertising over time. What remains after production cost and ad spend is what eventually makes its way into your pocket. So the real question is not just “where do I find marketers?” The real question is: what does it cost to generate enough exposures to make the relevant portion of my TAM aware of the product, at a price where enough of them are willing to pay, and does that math actually work? That is the core meta. A good example is Tesla. Tesla historically spent basically nothing on traditional marketing. Why? Because if the product is meaningfully better than the competition and you are already constrained by manufacturing capacity, spending on awareness does not create incremental profit. If they are already selling as many Model Ys as they can make, more marketing is just wasted motion. My recommendation would be to pre-sell the product first. In other words, design the product, understand how to bring it to market from a physical-goods standpoint, and then offer it for sale before fully scaling production. Kickstarter is the obvious example. My guess is that an Aella Kickstarter would probably do pretty well. And that would tell you something extremely valuable: whether there is enough real demand to justify bringing on an agency to run ads. Once you have that signal, yes, there are plenty of direct-to-consumer e-commerce agencies that do exactly what you’re describing, from paid ads to influencer management to creative strategy. I’m sure some will even reply to your post. But I think there are a few steps before that, and those steps matter a lot more than people realize.

Dynamically pricing based on available inventory is the only way to scale and manage inventory in any manner that actually maximizes total contribution. Additionally, defending against sold out variants by yielding price up is just as important as reducing prices on overstock. The last hotel room in Vegas on NYE is going to be priced a lot higher than the average daily rate. That's revenue management 101. But for some reason, this hasn't translated to retail yet. Perhaps it's because we can understand that if a room goes unsold, the hotel has lost that money but in retail a unit going unsold is still sitting in a warehouse. But actually, those two things are identical.

The best product manager at a startup is often the founder because they are the source of the vision and strategy that determine the company’s future. The same logic applies to CRO. When thinking about hiring for a conversion rate optimization (CRO) role within a smaller e-commerce business, it’s essential to understand that this function is far more than just a set of tactical tasks—it’s the very engine (or should be) that drives the business forward. The individual or team handling this function should not just be executing simple tasks like tweaking landing pages or adding trust badges on checkout pages; they should be integral to the strategic direction and growth of the company. The role of CRO within a business, especially a growing e-commerce operation, should be centered around experimentation that answers critical questions about the business’s future. It’s not just about increasing conversion rates in a vacuum; it’s about using experimentation to test hypotheses that can fundamentally alter the trajectory of the business. The best CRO programs are those that help the company understand and execute on ideas that could potentially double the business, rather than simply optimizing minor aspects like button colors or page layouts. The role, therefore, demands someone who can think strategically about the entire business, not just the website. One pitfall to avoid in hiring for this role is the temptation to offload it entirely to someone who approaches CRO from a purely tactical perspective. This is where businesses can get into trouble. Hiring someone who is focused on superficial metrics—like artificially boosting conversion rates by making some onsite funnel change that doesn’t make it back to any real profit—can distract from the more important, revenue-driving experiments that need to happen. The person you hire should have the ability to look at the big picture and prioritize experiments that have the potential to significantly impact revenue and contribution. This means they need to understand the infrastructure, people, and processes required to answer the biggest questions about the business. For these reasons, the ideal candidate for a CRO role should be someone with a deep understanding of business strategy, revenue growth, and the ability to operationalize insights across the company. They should not only be capable of running experiments but should also be able to translate the outcomes of those experiments into actionable strategies that can be scaled. This could mean finding someone with entrepreneurial experience, or at the very least, someone who has a proven track record of scaling businesses, in addition to running a sophisticated experimentation program. This person should own the revenue outcomes of their experiments, ensuring that their efforts are always aligned with the broader goals of the business. For sure it may be tempting to hire someone to simply “handle” CRO, but it’s crucial to recognize that this function is core to the business’s growth and should be treated as such. The right hire for this role will be more than just a CRO expert—they will be a strategic leader capable of driving the business forward through smart experimentation and a deep understanding of what makes the company tick. Trust badges on a checkout page might slightly increase conversion rates, but what you really need is someone who can help you double your business and operationalize the outcomes of 2x to 10x experiments. The kind of experiment that means you are SPRINTING to order more inventory. That’s the kind of CRO role you should be hiring for.

Percentages don't matter. Dollars do.

current ecomm thesis

I think what I see too often is merchant's not having a full grasp on what is actually driving the scale they are achieving. Not understanding why you are doing well is scarier than not understanding why you are doing poorly. You tend to find the root cause when you are doing poorly. Which then means we are optimizing towards an abstraction from the first principles of the business. As merchant's scale past $100mm those abstractions need to become data-driven people, process and technology that produce a regular cadence of predictable outcomes. For example, we shouldn't care if X or Y product doesn't sell because there are another 100 behind it and we have infrastructure and process in place so that doesn't matter. We should care less about the micro and more about the processes that if run, just allow us to get to the next level. We then optimize those processes to maximize revenue producing activity. Everything in service of perfect price and inventory management. I believe all that focus needs to be in understanding "how do we double this business in X period." If you don't KNOW that answer in your bones, modeling that is priority 1.

Is there really a scenario where intentionally limiting availability of a SKU leads to higher total profit? This 'scarcity strategy' is often cited to explain why brands like lululemon frequently have out-of-stock items. But let's consider a thought experiment: If we ran a test where one seller could only sell on certain days, while another could sell anytime, is there any realistic scenario where the restricted seller comes out ahead? While I could see some magic in limited-supply deals subsidizing CAC, I struggle to envision how an in-demand SKU being out of stock could ever be more profitable than consistent availability. Wouldn't managing demand through ad spend, pricing and production optimization ALWAYS yield better results than artificial scarcity?

Just as hotels optimize for RevPAR, e-commerce brands can and should optimize on 'Revenue Per Available SKU'

Hmm, in my opinion, if you have a framework for prioritization there shouldn't really be much discussion on the next initiative. It would be difficult for me to prioritize any testing that doesn't touch price + inventory over anything customer, UI/UX or brand related visual change. The big stuff is what matters. Everything else is deck chairs. The most valuable outcome for an AB test is an increase in the future annualized return on capital injected in the business. I tend to think if you can't look at the literal PnL or Balance Sheet and point to where the test affected profit or revenue, you probably didn't do anything.

High competition often means an inherent diluted value proposition. Low competition often means a strong, distinct value proposition. That's just how it works.

Let's say I have 10 SKUs in my store, all priced the same and CAC is X The fastest selling SKU goes out of stock: What happens to CAC? The slowest selling SKU goes of stock: What happens to CAC? I add an 11th SKU that now sells the fastest: What happens to CAC? I add an 11th SKU that sells the slowest: What happens to CAC?