Channel sequencing: the order matters more than the mix
I see a lot of newer ecommerce brands focusing on their channel mix, but underestimating how much the sequence matters.
A brand launching DTC from day one will deal with brutal economics. High CAC with no existing platform traffic, every single sale costs them a significant amount. But with time, and if they survive, that same brand owns something no marketplace gives them (especially not Amazon). They own a customer base, stronger pricing power, and brand equity. When they expand to other channels, they already have strong assets.
A brand launching on Amazon first gets the opposite. Faster revenue, lower CAC, immediate validation from customers. But the day they try to build a DTC presence or even expand to retail, they almost start from scratch. Their Amazon customers are hard to move to DTC, and the brand equity may not exist, because quite often it isn’t needed on Amazon.
The channel sequencing determines what the brand becomes, and each builds different assets. It isn’t limited to DTC vs marketplaces. TikTok builds the social side, while B&M retail gives strong credibility.
Here is an expensive lesson I learned when I owned a business last decade. It is more than figuring out which channels deliver revenues, but also which ones build the assets you need in the long term.
Quince raises $500M – will expansion dilute “affordable luxury”?
Quince just raised $500M to improve their manufacturer-to-consumer platform, expanding assortment, supply chain tech, and international reach. The company claims $1B in revenue.
This does not surprise me, I have always found their value proposition interesting at their price point.
M2C is a proven business model, and we have seen over the past few years manufacturers cutting out middlemen, whether they sell via Amazon, Temu, or more premium items via DTC channels.
Yet I worry their strategy to expand to other categories could dilute their “affordable luxury” value proposition. Is this a first step toward turning into another generic marketplace? I hope not, we have seen how this turned out for Etsy. There isn’t enough space for many marketplaces selling everything to everyone, even if slightly more premium than Amazon. Another concern is that this could be the start of a race to the bottom, lowering prices but also quality.
Maybe I am being too pessimistic here. The brand has strong assets, and I have been impressed with their first few years in business. Let’s see how their strategy will help them grow without losing their identity.
OpenAI kills Instant Checkout
OpenAI killed Instant Checkout. I am not entirely surprised. Less than 5 months after launch, Instant Checkout purchasing is gone. Despite what the company claims, the reality is simple: the product wasn’t ready.
I don’t think this is a big loss as of today. For people like you and me, immersed in the ecommerce industry most days, it can feel like a big deal. But we can’t ignore the disconnect between what AI companies sell to the market and how people actually use these tools.
Right now, the most viral AI use case is generating action figure images of yourself. That’s where we are. And while that’s happening, companies are asking users to trust ChatGPT with their credit card info. There is a massive gap between the features being announced and the actual adoption curve of current users.
Will agentic commerce eventually matter? Yes, I think it will, probably in a different form. Removing friction is a big part of ecommerce, does checkout embedded in a chatbot beat Amazon Buy Now button?
AI agents already have a lot of value, through researching, comparing, and recommending products autonomously. That’s a real use case. I am not sure what is next, but this will come when customers are ready.
For most businesses, the priority right now is making sure your product data is clean, your existing channels are where they need to be, and your checkout experience on platforms customers already trust is frictionless. That’s what moves the needle today.
You don’t need to sell on Amazon
Don’t listen to the Amazon consultants on LinkedIn: You don’t need to sell your products on Amazon.
Actually let’s tone it down a little. Some Amazon guys are great, and Amazon is a fantastic channel for the right brands. It delivers massive reach and Prime-fueled velocity for mass-market, commodities, price-sensitive, or even impulse SKUs.
But in 2026, the “Amazon is mandatory” advice is costing many premium DTC brands many margin points.
Here are two key aspects of this reality for those brands:
1 – Cannibalization hits hard: It isn’t uncommon that >50% of Amazon sales often substitute direct purchases from loyal customers. You pay 15% referral fees + rising FBA costs on lower-net transactions, and profitability does not scale as much with AOV.
2 – Premium items struggle: Listings get buried under cheaper competitors and Amazon can’t act as a discovery engine. PPC is essential, but a money pit, especially with rising CPC.
If your brand has strong direct relationships, considered purchases, low direct competition, and/or next-day shipping is not relevant, Amazon can erode margins without adding significant incremental demand.
In that case, I’d recommend treating Amazon as a controlled discovery channel at most, with selective listings, strict guardrails, and no heavy FBA reliance.
Brands in 2026 can’t be “all-in”, and must look further than GMV. This includes modeling full channel P&Ls, and considering customer ownership + profitability.
ChatGPT ads: Albertsons, Target, Williams-Sonoma pilot
Corporate buzzwords don’t magically turn experiments into strategies, and that’s exactly how I view how some executives justify buying into ChatGPT ads right now.
Early pilots include Albertsons, Target, and Williams-Sonoma. For example, Albertsons tells us about ads triggered by queries like ‘best flowers for Valentine’s Day’ or ‘how to celebrate Galentine’s,’ directing users to local landing pages with deals, recipes, and delivery options.
I see the most value in gaining experience and gathering data more than on a few conversions. But platform stability is a big question: Will today’s data still be valuable if formats, targeting, or user behaviors change quickly, following new platform updates?
On revenue generation, I’m skeptical. While Target reports 40% month-over-month traffic growth from ChatGPT overall (impressive, but mostly organic so far, and probably low in terms of absolute numbers), it’s unclear how much paid ads contribute or whether they’ll deliver sustainable ROI. The example with Albertson and Valentine’s day may work if CAC remained reasonable, but based on what I have heard on LinkedIn, these ads may be costly. And we might be overestimating intent with agentic commerce. We’ll need to see real performance data once it scales.
25% of shoppers abandon carts over account creation
“25% of online shoppers abandon their cart because the site wanted them to create an account.”
Lots of interesting statistics in this article, but this one caught my eye. Why? Because it is absolutely preventable. Asking customers to create an account worked 20 years ago, but now that Amazon and other platforms have removed as much friction as possible, it is unreasonable to require it.
“But what about gathering customer data?” Are customer emails really that valuable if your cart abandonment rate is high due to too much friction?
Also, not forcing customers to create an account doesn’t mean you can’t build a relationship with them. But you’ll now have to give them a good reason to do so. If they don’t want an account, great: you still give them the option to buy and you get the sale. But if they do create an account, why not offer them free shipping or unique perks?
eBay’s journey: not every player can be Amazon
eBay’s journey is a good reminder that not every ecommerce player can (or should) be Amazon.
While GMV peaked at $100B in 2020, it dropped in the $70B range before reaching $79B in 2025. On the other hand, the stock went from $50 in 2020 to over $90 recently.
I attribute this success to eBay understanding their environment and taking corrective action. More than just aggressive share buybacks, they refocused on core strengths like recommerce, collectibles, or auto parts. I have myself placed more orders on eBay in 2025 than during any year since the mid 2000s. Higher margins and advertising growth followed.
Unfortunately, not every story ends this way, who said Etsy?
$200k minimum to advertise on ChatGPT
How bad a gambling addiction do you need before spending $200k on ChatGPT ads?
I’ve already discussed the absurdly high initial CPM, but one detail I initially missed is the $200k minimum commitment.
My guess is that both the CPM and the minimum buy-in will drop significantly shortly after the ad rollout begins. For now, I’d recommend brands wait and see what happens once the dust settles. Is OpenAI trying to make this channel look bigger and more profitable than it really is?
That said, I don’t think the 4% transaction fee should be a deterrent. Given the nature of AI agents, this is especially true for brands and products that:
– Solve a clear, specific problem
– Are easy to explain in plain language
– Sell through logic, not impulse
– Are compared on function more than aesthetics
TikTok virality and the stockout trap
When you are done celebrating your product going viral on TikTok Shop, don’t worry about if it will go out of stock. There is a good chance it will anyway.
What you should worry about is how you handle the stockout, while customers are already looking elsewhere and competitors are pushing alternative items. We all have the attention span of a goldfish now, myself included. And the same is true for our customers.
If TikTok inventory is empty but you still have stock elsewhere, redirect to where fulfillment is still possible, or toward similar available items.
At the same time, capture the demand you just earned. You can offer restock alerts, early access, or a small incentive in exchange for email or SMS. This turns short-lived virality into a relationship you actually own.
Behind the scenes, keep the story alive and be transparent. Momentum matters in these moments, and silence kills it.
Of course competitors will try to take advantage and steal market share. That is when your differentiation matters most. Remind people why they should wait for the original, not settle for look-alike products.
Running out of inventory is not the real risk here, overstocking can be just as dangerous. Losing attention, trust, or momentum is far worse but can be avoided.