Tuesday morning, Squawk on the Street. Harley Finkelstein walked in, clean white shirt, no hat , where is Harley’s hat? , and announced that Shopify had crossed $101 billion in GMV. In one quarter. Revenue up 34%. Strongest growth in four years.

The stock fell 15%.

We are still , somehow , sponsored by Sumoblue. An Irish marketing agency whose pitch this season is simple: the basics are back. No new shiny thing. No agentic anything. Just PPC, SEO, conversion, the work. Side effects include calmer Mondays. Tell them the V Spot sent you. They will deny it. Tell them

anyway.

The Stories

Amazon Logistics: AWS Walks Into A FedEx Bar

Monday. Andy Jassy launched Amazon Supply Chain Services , opening Amazon’s full freight, fulfillment, and parcel network to any business, anywhere, regardless of whether they sell on Amazon. FedEx fell 7%. UPS fell 9%.

Shopify Did $101 Billion And Got Punished. $101 billion GMV in 90 days. AI-driven traffic to Shopify stores up eight times year-over-year. Q2 guidance: high twenties percent growth. Wall Street voted. Stock dropped 15%.

Dyslexic Spotify shareholders, briefly delighted by the dip, then realised. Again.

Whatnot Plugs Into Shopify, Brings Live Commerce With It

Whatnot launched its direct Shopify integration. Whatnot did $8 billion in seller GMV last year. More than double 2024.

Live commerce in the US is a $20 billion category projected to double again in 2026.

Germany Sends Shein And Temu A €2.4 Billion Invoice. The German Retail Federation released a study claiming Shein and Temu deliver 460,000 parcels into Germany every single day. The kicker , half of Shein and Temu shoppers said they’d have bought identical products from German retailers at the same price.

Then why didn’t they.

Klaviyo Lets Brands Build Skills In English. This one is small but matters. Type the rule, ship it. No code. Every CX manager in your network is currently figuring out whether this means less work or more work.

Amazon Decides To Win UK Grocery The Old-Fashioned Way

The Grocer reported on Tuesday that Amazon Fresh in the UK is now the cheapest grocer on 64.9% of 2,313 tracked SKUs. The bold new growth strategy. Be cheaper than the competition. Mind-blowing stuff. Take notes, Harvard Business Review.

Big thanks to Omnisend for closing us out , the email and SMS platform pulling $68 in revenue for every $1 spent. May your 4th, damn it too late....

So this week , Amazon became a logistics company, Shopify printed $101 billion and got punched in the face for it, Whatnot dragged Shopify into the livestream era, Germany discovered Temu owes them a tax bill, Klaviyo handed agentic commerce a plain-English keyboard, and Amazon learned how to use a calculator in the UK.

Retail’s not dead. It’s just getting better at being strange.

That’s the 9 minutes you won’t get back this week.

There is a number in the Mapp Fashion Intelligence Report that should be in your head more than it does.

Next’s Google exit rate: 84%.

Five out of every six visitors who arrive at Next via Google leave without buying. In the Mapp dataset, 397 retailers, 7 markets, 53.7 billion visits, that is the highest exit rate in the sector. River Island at 79%. Primark at 79%. Next at 84%. The report I wrote last month cited this as one of the “clearest signals of site-side discovery failure.” I stand by that characterisation. The discovery model that built UK online retail is breaking, and these numbers are part of the evidence.

Next reported Q1 full-price sales up 6.2% this morning, beating its own 4.0% forecast by £28 million. Online grew 10.1% in the UK. The LABEL platform, third-party brands hosted on Next’s infrastructure, grew 15.7%. Full-year pre-tax profit guidance raised to £1,218 million. Shares up.

So: highest Google exit rate in UK fashion. Best results. Profit guidance above £1.2 billion. How do those two things exist simultaneously?

The answer to that question is probably the most important thing happening in UK retail right now, and almost nobody is saying it clearly.

The rabbit hold revisited

The context matters. In The Rabbit Hole Got Deeper, I laid out the structural case for why UK online retail is in genuine difficulty. The IMRG data covering March 2021 to early 2026: conversion baseline drifting relentlessly downward, from around 4% to sub-1.6%, with March 2026 session conversion rates down year-on-year from 1.72% to 1.59%. CPA rising 22.53% in the same period.

The Mapp Fashion Report, 397 retailers, Sarah McVittie’s work, the most rigorous dataset available on UK fashion discovery economics, delivered the verdict underneath the chart. Paid search nearly tripled its share. Organic contracted 4.1 percentage points. The industry looked at a declining organic position and instead of asking why, reached for the credit card. Paid search added 3.2 billion visits, a 264% increase, while organic barely moved.

The financial consequence is the danger zone: brands running paid at 4-8% of their traffic mix produce average operating margins of 3.7%, improving in only 48% of cases. They are paying full auction prices for discovery they used to earn. The maths doesn’t work.

And in that same dataset, Next’s Google exit rate is 84%.

Huh?

Here’s why the 84% exit rate and the £1.2 billion profit are not contradictory. They are, to me, the same story.

Next’s Google exit rate being the highest in the dataset doesn’t mean Next has the worst site. It means Google is the wrong entry point for Next’s customer. A Next customer who arrives via Google is typically arriving with a specific, narrow intent, a particular search term, a competitor comparison, a price check. They didn’t start at Google because they wanted to browse Next. They started at Google because they were shopping the category. For that visitor, Next’s site is not where they want to end up. They bounce.

The Next customer who converts doesn’t arrive via Google. They arrive direct. They arrive via the app. They arrive because they have a Next account and a relationship with the Directory. They arrive because a friend mentioned it. They arrive because they just got a Next Finance statement and clicked through. They have, in the language of the Mapp report, brand gravity, and brand gravity doesn’t show up in the Google exit rate. It shows up in the direct traffic, the repeat purchase rate, the LABEL platform numbers.

This is the point about owned audience I made in the Rabbit Hole piece, and it is perfectly illustrated by the Next paradox. The brands that will be fine are the ones that built genuine direct relationships before the market changed. Next is the case study.

The 84% Google exit rate is a symptom of a brand that has successfully moved most of its customer relationship to channels Google cannot see.

Read more

You are blinded by the lights. Every channel is an emergency. Every founder thread on X becomes a strategic priority before the coffee is finished.

You have Squirrel Syndrome. You know who you are. TikTok Shop Monday. Retail media Tuesday. Full rebrand by Thursday. The previous initiative , the one from three weeks ago that was going to change everything , is in a Notion board. Unopened. Judging you silently.

Objectives shift every other week. Your team has stopped asking why. They know the answer will be different by Thursday.

You don’t sleep well.

You self-medicate with caffeine and conferences. You post pictures of the same industry dinner every week as though being in the room is the same as having a plan.

And then Monday comes.

And the basics still aren’t done.

We know what works. We have always known what works.

Clean data. Clear positioning. Retention that actually runs. Paid that makes mathematical sense. A product page that converts.

Basics.

In 2026, we are bringing basics back.

We say this every year.

Every year, the market doesn’t listen.

Every year, we’re still here.

Sumo Blue has been bringing basics back for over ten years.

We should be clear , Sumoblue did not bring sexy back. That was Justin Timberlake. 2006. Different brief. We bring basics back. Which, in this market, is honestly the sexier move.

Sumo Blue. The basics are back. Again. They never left.

⚠️ Disclaimer: Sumo Blue accepts no responsibility for the mild embarrassment of realising the thing that fixes your business is the thing you already knew. Side effects include calmer Mondays and fewer panic Slack messages.

Bonus Material - if you are still reading.

🗳️ Hot or not , Are you an opportunity junky?

🐿️ Full Squirrel , new strategy every week, thriving

📋 Recovering , learning to stick to the plan

🧱 Basics first , boring and winning

😶 The Notion board is a graveyard and I live there

This is part of the Sumo Blue PSA series. Brought to you by common sense and mild professional despair.

Next week: …...

We tried to sit down with Ryan Cohen, founder of Chewy, chairman of GameStop, and the man currently attempting to buy a company four times his size. to discuss ambition, audacity, and what exactly he said to his bankers. He said no, we were bonkers. This is what we think would have happened.

Let’s start at the beginning. You built Chewy from nothing into a $3.35 billion exit. That’s a legitimate origin story. Why isn’t that enough?

It’s never enough. I see undervalued things and I want to fix them. That’s just who I am.

Right. And GameStop,you saw something undervalued there too.

Absolutely. The market had written them off. I saw the community, the stores, the potential in collectibles and trading cards. We repositioned the business.

You did. And the stock went from about $4 to $500 in a matter of weeks in 2021. That must have been something.

It was extraordinary. The retail investor community believed in what we were building.

They did. And then it came back down to earth. Currently trading around $28. Are those retail investors,the ones who bought at $400,part of the community you’re still building for?

We’re focused on long-term value creation.

Of course. Let’s talk about eBay. You started accumulating a 5% stake in February,the same week eBay announced the Depop acquisition and posted its strongest quarterly numbers in years. Most people buy undervalued assets. You bought one mid-momentum. What did you see that the market didn’t?

The market is pricing eBay at its current trajectory. I’m pricing it at its ceiling. There’s a significant gap between those two numbers.

There is. And to close that gap, you’ve proposed $2 billion in annualised cost cuts within twelve months. $1.2 billion of that comes from sales and marketing. eBay spent $2.4 billion on S&M in 2025 and added one million net active buyers. You want to cut half of that budget. What happens to the buyers?

The marginal dollar of marketing spend at eBay is producing almost nothing. The model is inefficient.

The marginal dollar, yes. Rick Watson,one of the sharper analysts writing about this,makes exactly that point. But he also notes you’re making a much bigger claim about the average dollar. Cut half the budget and you don’t just lose the marginal spend. You potentially lose the brand. Anyone who ran a DTC P&L through 2023 watched that happen in real time. Why is eBay different?

eBay has 134 million active buyers. The brand doesn’t need $2.4 billion to hold that base.

It might not need it to hold the base. It definitely needs something to grow it. eBay just signed a headline sponsorship of Saturday Night Live UK specifically to introduce eBay Live to a new generation of buyers. That’s brand spend doing exactly what brand spend is supposed to do. You’d cut the budget that funds moves like that.

We’d be more targeted. More efficient.

More targeted than Saturday Night Live UK?

[pause]

We’d find the right vehicles.

Let’s talk about the vehicles. 1,600 GameStop stores as authentication centres, fulfilment hubs, live commerce studios. That’s the part of your proposal that survives the most scrutiny. It’s a genuinely interesting idea. But GameStop’s mainstay business buys goods wholesale and resells them through physical stores. eBay earns fees without touching inventory. These are not just different companies. They’re structurally opposite business models. How do you integrate them without breaking both?

The stores are the story. The physical network is something eBay could never build organically. We’re not merging the business models,we’re adding infrastructure.

You’re adding infrastructure that employs tens of thousands of people across a retail footprint that’s been contracting for years, onto a marketplace business that has succeeded precisely because it doesn’t carry the costs of physical retail. Morgan Stanley called the business models “fundamentally different.” Michael Burry,who once compared you to Warren Buffett,sold all his GameStop shares and called the strategy pedestrian. These are not obscure critics.

Michael is entitled to his view.

He is. He also said your honest intent is to dominate collectibles and used goods, not compete with Amazon,that the Amazon framing is cover for a narrower play. Is he wrong?

I think the opportunity is bigger than collectibles.

You said on CNBC that “we have the ability to issue stock to get the deal done.” That’s the financing answer you gave on national television for a $56 billion acquisition. eBay shareholders receiving GameStop stock as part of their consideration are being asked to swap equity in a cash-generative, debt-free marketplace for equity in a $12 billion retailer whose valuation is partly a function of the deal itself. Can you see why that’s a hard sell?

We have $9.4 billion in cash. We have a $20 billion highly confident letter from TD Securities. The financing works.

The financing works if the stock leg is accepted by institutional shareholders who own the majority of eBay’s float. Those shareholders watched eBay trade at $110 against your $125 offer,a fifteen dollar gap that represents the market giving your deal roughly a coin flip’s chance of closing. What do you say to the institutional eBay shareholder who has watched the stock rise 20% this year, seen Depop acquired, seen the SNL UK partnership, seen Q1 GMV up 18%, and is now being asked to take GameStop paper as part of the exit?

We’re offering a 46% premium to where eBay was trading in February.

A 46% premium to February’s price. A 20% premium to Friday’s close. And the premium is being paid partly in the currency of a company whose market cap depends significantly on retail investor sentiment and the success of this very deal. That’s not a criticism,it’s a structural observation. Has any institutional investor told you privately they’re comfortable with the stock leg?

We’re in active discussions.

You went on CNBC and the market went to $110. You needed it to go to $125. That’s the gap between a press tour and a deal.

We’re just getting started.

Let’s end where we began. You looked at eBay,a company with the right strategy, the right infrastructure, genuine momentum, a 30-year brand, 134 million buyers, $1.5 billion in free cash flow,and decided it needed you to rescue it. eBay’s position is that the turnaround is working and disruption is the wrong call. Make the case, in one sentence, for why you’re right and they’re wrong.

They’re executing at the speed of a company that doesn’t want to get fired. I’m offering the speed of a company that has nothing to lose.

That’s a very honest answer.

I know.

It’s also, for an eBay shareholder, the most frightening thing you’ve said today.

[smiles]

Ryan Cohen, thank you for your time.

The Ostrich Report covers ecommerce strategy, platform dynamics, and the business of retail for senior operators and industry insiders. Sometimes we interview people who haven’t technically agreed to be interviewed.

Ryan Cohen declined to comment. The quotes above represent what he would have said if he had.

It’s May Day. International Workers’ Day. Across the world the working class is meant to rise up and demand fair wages, dignified labour, and an end to exploitation. eBay celebrated by going completely offline. I spent the week in Belfast at the heart of Irish and UK ecommerce. We learned that the Mack in Noel Mack is short because, he didn’t eat his greens. No, in fact, like a famous Wrexham owner, his name is Irish - McElhinney. He wowed our audience by Sesame Streeting us - with some carefully placed swear words and the numbers - the standout one being $1BN - their turnover. He also shared some photos of a Young Ben Francis but colour me skeptical.. It seems young Ben has been aged somewhere between 21 and 24 for almost a decade now. What age are you Ben?

Right. Let’s do the news.

eBay Goes Dark Right Before It Has To Speak To Investors

eBay’s marketplace fell over on Sunday and stayed over, intermittently, for the better part of 48 hours. A hacktivist crew calling itself 313 Team claimed responsibility for a DDoS attack. eBay refused to confirm anything. Buyers couldn’t search. Sellers couldn’t list. Auctions ended in a state of metaphysical uncertainty.

eBay turns over about $230 million in gross merchandise volume per day. That’s somewhere between zero and Iceland’s GDP gone, depending on which hour you ask. The bit that’s almost too perfect, earnings call was Wednesday. So executives spent the weekend rehearsing the line that says “we are investing heavily in platform reliability.”

And remember, this is the same eBay that recently announced it was passing on agentic shopping. Won’t let the AI agents in. Couldn’t let the humans in either, as it turned out.

Walmart Wishes You A Pivotal Moment

Walmart published its annual report. John Furner’s first one as CEO. The headline numbers are strong — $715.9 billion in revenue, global ecommerce up 24% to $150.4 billion. Furner’s shareholder letter said, and I quote: “We are at a pivotal moment.”

We are always at a pivotal moment. Every annual report ever written has been published from inside a pivotal moment. If we were not at a pivotal moment, the annual report would say “nothing is happening, please continue buying cheese.”

But what was interesting,cheese aside, was Sparky, Walmart’s AI shopping agent, is driving baskets 35% higher than the non-AI experience. So if you’re still wondering whether agentic commerce is a fad, Walmart’s basket size has answered.

Amazon Quietly Reroutes Its Sellers’ Cash Flow

Amazon’s overhauling how seller ad fees are paid. The new policy (yes another one): ad costs deducted automatically from your sales proceeds. No more credit cards. No more floating the spend to next month. Sellers complained loudly enough that Amazon pushed the deadline back to August 1st. But the direction is set in concrete.

In plain English, Amazon’s millions of SMBs just lost two things they’d built their cash flow around. Float. And credit card rewards. For a third-party seller, those points weren’t points. They were the difference between making payroll and not. And now they’re going to a payment rail Amazon controls end to end.

If you’re a small Amazon seller and you’re not yet in your feelings about this, give it 90 days. You will be.

Spotify Buys A Bicycle

Spotify announced a global partnership with Peloton on Monday. 1,400 on-demand workout classes integrated into Spotify Premium — yoga, pilates, strength, the lot. Dyslexic Shopify shareholders briefly thought Tobi had bought a stationary bike company. He had not.

What this actually is, Peloton finally admitting their hardware-first model has plateaued, and Spotify finally admitting that music alone doesn’t justify the subscription anymore. Two companies in their thirties, both very gym, both very content-tired, decided to start a wellness journey together. May they grow stronger. May we never have to install an app called Strava-but-Spotify. May Daniel Ek find his core.

BigCommerce Turns Into A PayPal Tracksuit

BigCommerce announced Wednesday that it’s plugging PayPal’s Store Sync into its app marketplace. The pitch: BigCommerce merchants can now connect their catalog directly to — quote — “AI surfaces.”

AI surfaces. That’s the new word for “places where commerce happens that aren’t your website.” Sharon Gee, BigCommerce’s SVP of AI, said merchants need to “meet shoppers in those moments.”

And for a change, If you’re Irish or European watching this from across the water — this applies to you too. Just in three more languages and with a GDPR notice taped to the front of it.

Domino’s CEO Predicts Misery, Generously, On Behalf Of His Peers

Final story. Domino’s stock dropped Monday on weak Q1 results. Same-store sales soft. The CEO went on the call and said paraphrasing only slightly more chains will follow.

Generous of him. Big “the iceberg got us, but don’t worry, it’s coming for the rest of you” energy. Domino’s, to be fair, sells pizza in a tariff economy where every box, every bit of cheese, every tin of sauce costs more than it did last year. And the consumer — sitting on a $1,300 hidden tariff tax and a delivery surcharge is not springing for extra pepperoni.

So when the Domino’s CEO says the rest of you are next, believe him. He’s not threatening you. He’s hugging you.

We somehow continue to be sponsored by SumoBlue We say “somehow” because, frankly, the longer this goes on, the more concerned we get for their judgment. SumoBlue are an Irish Marketing agency who clearly have given up and let me sing for their supper. Tell them the lad with the accent sent you.

That’s the 6 minutes you won’t get back this week.

Some Other Stories Catching our eyes

UBS: Tariffs and Immigration Policies Could Drive More Retail Closures in 2026 URL: https://www.retaildive.com/topic/e-commerce/ Subject: Retail Macro / Tariffs Region: US Summary: UBS analysts published research arguing that the combination of sustained tariff pressure and immigration enforcement is weighing on the industry and likely to accelerate the pace of store closures across the year. Specialty retail flagged as the most exposed. POV: “Tariffs already had the kitchen on fire. UBS just noted that immigration enforcement is now also locking the back door.”

eBay Marketplace Goes Dark Following Suspected DDoS Attack URL: https://www.retaildive.com/news/ebay-scrambles-fix-outages-marketplace-down/818601/ Subject: Marketplace / Cyber Region: US (global impact) Summary: eBay’s marketplace site experienced widespread outages from Sunday afternoon into Monday and Tuesday, with API and search failures affecting buyers and sellers globally. Hacktivist group 313 Team claimed responsibility for a denial-of-service attack. eBay declined to confirm the cause. The platform turns over $230M in GMV per day. POV: “The company that wouldn’t let AI agents into its checkout couldn’t keep humans in either.”

Walmart Releases 2026 Annual Report — $715.9B Revenue, 24% Global Ecommerce Growth URL: https://www.retaildive.com/news/walmarts-annual-report-ecommerce-store-investments-AI/818542/ Subject: Earnings / AI / Retail Strategy Region: US Summary: Walmart’s first annual report under CEO John Furner shows total revenue up 5.1% to $715.9B, with global ecommerce up 24% to $150.4B. Sparky, Walmart’s AI shopping agent, is driving 35% higher basket sizes vs the non-AI experience. Capital allocation to new stores up 212% YoY. POV: “Walmart’s annual report is at a pivotal moment. So is every other annual report. Has been since 1962.”

Spotify Launches Fitness Hub With 1,400+ Peloton Classes for Premium Subscribers URL: https://www.cnbc.com/2026/04/27/spotify-peloton-fitness-content-hub.html Subject: Subscriptions / Wellness / Content Region: ROW (US, UK, EU, Australia, Mexico, Canada) Summary: Spotify and Peloton announced a global partnership making 1,400+ on-demand classes (yoga, pilates, strength, meditation, outdoor) available within Spotify Premium. Marks Spotify’s first major fitness category and Peloton’s furthest move yet from the hardware-first model.

Domino’s CEO Warns More Restaurant Chains Will Follow As Q1 Sales Slip URL: https://www.cnbc.com/retail/ Subject: Retail / Macro / Consumer Region: US Summary: Domino’s reported weak Q1 same-store sales, and the CEO said on the earnings call that other chains would likely report similar pressure as the quarter unfolds. Stock fell on the day.

BigCommerce Integrates PayPal Store Sync, Targets “AI Surfaces” URL: https://www.pymnts.com/news/retail/2026/bigcommerce-bets-on-agentic-commerce-with-expanded-paypal-pact/ Subject: Agentic Commerce / Payments Region: US (global rollout) Summary: BigCommerce announced an integration with PayPal’s Store Sync product, allowing merchants to syndicate catalog and inventory data into AI shopping channels. Positions BigCommerce alongside Shopify (Agentic Storefronts) and against eBay (which refused agent checkout) in the agentic commerce platform divide.

eBay Q1 2026 Earnings (delivered amid the outage tail) URL: https://www.valueaddedresource.net/ebay-server-errors-search-descriptions/ Subject: Earnings Region: US Summary: eBay reported Q1 2026 earnings two days after a major site outage. GMV trends and ad revenue both under analyst scrutiny.

Caterpillar Q1 Earnings Beat — But Tariffs Cost $710M in Quarter URL: https://www.cnbc.com/2026/04/30/caterpillar-cat-q1-earnings.html Subject: Industrial / Tariffs Region: US Summary: Caterpillar Q1 EPS came in at $5.54, 30% YoY, beating consensus by 20%. Manufacturing costs related to tariffs hit $710M in the quarter. Backlog of $62.7B provides visibility, but tariff drag is now a permanent line in the model.

Amazon Tightens Grip on Seller Cash With Card Payment Overhaul URL: https://www.pymnts.com/amazon/2026/amazon-tightens-grip-on-seller-cash-with-card-payment-overhaul/ Subject: Marketplaces / Seller Operations Region: US Summary: Amazon’s overhaul of seller ad-fee billing — moving from card-based payment to automatic deduction from sales proceeds — has been delayed from April 15 to August 1 after seller pushback. The directional shift removes float and credit card rewards from millions of SMB sellers.

Woolworths (Australia) Freezes Prices on 300 Own-Brand Products for 3 Months URL:

Retail Slop

Retail Around The World: 30 April 2026

EUROPE Aldi Nord enjoys big 2025 gains across European markets…

Read more

13 days ago · 1 like · Howard Lake

Subject: Grocery / Pricing / Macro Region: ROW (Australia) Summary: Woolworths Group will freeze shelf prices on 300 own-brand or exclusive products from May 1 for three months, absorbing supplier cost increases on fuel, fertiliser and packaging. Group ecommerce sales rose 20.2% to AUD 2.7B in the prior period.

He says tht at at Nike says we’re nowhere near digital maturity. And we never will be. That’s the strategy. Such a massive contrast compared to listening to Noel Mack yesterday from Gymshark - and Noel said it himself. Just because you can box, does not mean you get in the ring with Ali. (or Mayweather).

Mike is a study in calm intelligence, and the first word he reached for, in the first minute of our conversation, was excited. 25 episodes into this show, no other guest has done that.

His operating frame, which I’m still chewing on:

→ Nimbleness is the strategy. Not a personality trait, a discipline.

→ Culture is the determining factor in whether your nimble thinking actually ships. It is not the HR poster. Anyone knows me, knows my feelings about culture and its place.

→ Diversity of perspective is operational, not decorative. If your team doesn’t reflect society, you cannot build for the consumer. This is a consistent message that i heard all over the last few days. Be close to culture - it is not a throw away comment or reserved for the biggest companies. It is something we can all do.

And the quiet detonation of the episode: at Nike, “anyone with a body is an athlete.” Total addressable market = everyone. Now try designing for that.

Massive thanks to Omnisend and Sumoblue for backing the show this season. We’re open for further sponsor conversations drop me a line if your product belongs in front of operators who think for a living.

Episode here ↓

I have been sitting with this conversation since January, trying to find the right words for it. The first piece I wrote, The Soul of the Sidewalk, came close. But it was written quickly, in the aftermath of feeling, and it didn’t fully do what I wanted it to do, which is to make sure that anyone who reads it understands that meeting Tren’ness Woods-Black is not a networking event. It is not a content opportunity. It is an encounter with someone who carries, in her person, something the commerce industry has been trying to manufacture for thirty years and consistently failing to, the real thing.

I want to write this properly. I want to get it right. Because she deserves that.

“We must forever conduct our struggle on the high plane of dignity and discipline.”

Those are Martin Luther King’s words. They are also, without her ever having said them, the operating philosophy of every day of Tren’ness Woods-Black’s life. She doesn’t describe it that way. She says things like “solve and serve” and “find the commonality” and “don’t label, go to the heart of the story.” But the structure underneath all of it is the same: dignity, discipline, and an absolute refusal to let the hardship become the whole story.

This episode is about her. But it’s also a quiet ask, to the industry, to the brands, to anyone reading this who has the capacity to help, to show up for people like Tren’ness with the same commitment she shows up for everyone else.

1962, A Fifteen-Stool Luncheonette, and a Woman Called Sylvia

In 1962, a woman named Sylvia Woods opened a small restaurant on 126th Street in Harlem. Fifteen stools. Soul food. The kind of cooking that carries memory in every mouthful, collard greens, black-eyed peas, cornbread, smothered pork chops in a savory gravy with onions and green peppers. Not a cuisine. A language.

Harlem in 1962 was the destination for Black Southerners who had been escaping, for decades, the atrocities of Jim Crow, the violence, the degradation, the daily terror of living in a country that had not yet decided whether they were fully human. They came north looking for something. Sylvia’s, whatever else it was, gave them something specific and irreplaceable: a place where they didn’t have to explain themselves. Where you could sit next to a stranger and immediately be understood, because you’d come from the same town, or your families had, or because the food tasted like home in a way that nothing else in New York did.

Sylvia called them guests. After the first meal, she called them family.

Over sixty years, the restaurant expanded. Presidents ate there. Celebrities. Tourists from every country in the world. The New York Times wrote about it. Barack Obama came. Bill Clinton came. Denzel Washington came. But the neighborhood locals kept coming too, because Sylvia’s was never for the famous. It was for whoever walked through the door.

When Sylvia died in 2012, her granddaughter Tren’ness was already a grown woman who had inherited not just a family business but a way of being in the world that Sylvia had made non-negotiable from birth: every person who comes through the door chose to spend their money with you. You never forget that. You never treat it casually. You receive it as the gift it is.

Tren’ness has been carrying that gift, and that responsibility, her entire adult life.

Who Is Tren’ness Woods-Black?

She is a strategic partnerships consultant, a community builder, a marketer, and a woman who has raised $1.4 million in sponsorship for community events in Harlem. She has worked with DoorDash, Thrillist, and partners across the hospitality and tech ecosystem. She has spent years trying to bring digital tools, the kind that NRF attendees take entirely for granted, to the small businesses and restaurants of Harlem that form the bedrock of a neighbourhood most of the industry only passes through on the way to something else.

She is also, as of the last twelve months, someone who hit a wall.

Read more

I thought I was writing satire. It turns out I was mostly writing a roadmap.

Hot take coming through

This week PYMNTS ran Karen Webster’s Monday Conversation with Libor Michalek, president of Affirm, under the headline Agentic Credit Rewrites the Rules of Consumer Lending. The thesis, on its own terms, is entirely persuasive. The 60-year-old revolving credit line is a workaround that won because underwriting was hard and data was thin. Now it isn’t. Lenders can assess the transaction itself, this purchase, this price, today, against this person’s actual cash flow this month, and price the risk in real time. Affirm calls it agentic credit. They are, to be fair to them, very good at it.

Buried in the framing is the sentence the rest of us should be reading three times.

Affirm sees millions of consumers every year shown the full cost of a purchase and walking away from it. Michalek’s framing: a credit business that earns money when you decide not to spend is, structurally, on your side. He’s right. It’s a better model than the old revolving line, where the lender’s incentive runs in the opposite direction.

Now picture that machinery sitting inside an agentic flow.

Collision

The agent has been delegated a basket. The user has gone for a run, or fallen asleep, or is in another meeting. The agent gets to the lender’s API. The lender does its real-time assessment. The lender comes back and says: no, not this one. Or: yes, but only on these terms, and only for half of it. The user, being absent, doesn’t see the prompt. The agent has to make a call. Drop the basket, downgrade it, switch BNPL provider, kick the whole thing back to a manual checkout, or just abandon. Many will feel embarrassed and does this footprint stop you from a purchase elsewhere? We have been free to manage these sorts of daily struggles ourselves, forever, this might just limit that.

This isn’t a failure of the technology. It’s the technology working exactly as designed. Honest finance and frictionless commerce have, on a long enough timeline, structurally opposite goals. Honest finance puts the cost in front of the consumer at the moment of purchase. Frictionless commerce removes the consumer from the moment of purchase entirely. Affirm’s proof point, the thing they are quietly proud of, becomes a conversion problem the moment you take the consumer out of the loop.

And that’s just the American half of the story.

THE ATLANTIC FLIP

If you’re Irish or European watching this from across the water, there’s a second layer the US-led press isn’t quite reading carefully enough.

From 15 July 2026, the UK’s Financial Conduct Authority brings BNPL, formally re-named Deferred Payment Credit, into full consumer credit regulation. Affordability checks on every transaction. The Consumer Duty’s “good outcomes” test. Access to the Financial Ombudsman Service. And, crucially, Section 75 joint liability between lender and merchant for any purchase over £100. The UK BNPL market grew from £60 million in 2017 to over £13 billion in 2024. The regulators have decided that’s enough running room.

Then on 20 November 2026, the EU’s Consumer Credit Directive II (CCD2) lands across all 27 member states under full harmonisation. Every BNPL provider in the bloc has to do proper credit checks, supply standardised pre-contract info, give consumers a 14-day withdrawal right, and stay inside national APR caps. The Netherlands has gone further and banned BNPL for minors outright. Germany is folding the whole category into the Bürgerliches Gesetzbuch like it’s a building society loan from 1972.

Karen Hao would recognise the shape of this. When a category grows fast inside a regulatory vacuum, the regulators arrive eventually, they always arrive late, and they always arrive with everything they wished they’d done five years ago. UK and EU legislators watched BNPL go from a curiosity to 20% of UK adults, 10.9 million people, using it inside a year, and decided that no protections for those who use it repeatedly and may not be able to afford it was no longer a policy. It was a problem.

The flip matters. In the US, agentic credit is being framed as a product innovation, Affirm telling us it’s the next evolution of consumer lending. In the UK and EU, the same machinery is arriving alongside a regulatory regime that demands affordability checks, robust documentation, customer signposting to debt advice, and a complaint route through an ombudsman. Same technology stack. Wildly different political economy.

Which means the agentic checkout in 2027 won’t look the same in San Francisco as it does in Birmingham, never mind in Berlin or Tralee. The agent that one-clicks a Foundrae piece in Manhattan is operating in a different credit reality from the agent trying to buy a £400 set of trainers in Manchester. The merchant on the EU side will have a regulated lender between the agent and the basket, declining transactions for reasons the agent has no jurisdiction to challenge.

Turn

The honest read is not that agentic commerce is dead, or that BNPL is dead, or that Affirm has misread the room. The read is that frictionless commerce was always a marketing phrase, and the friction we’re now adding back, affordability checks, real-time underwriting, Section 75 liability, ombudsman recourse, is the friction that should have been there all along. The merchant on the ground, the operator who wants to model agentic into 2027 GMV with any seriousness, has to assume a higher decline rate at credit, a different conversion ceiling, and a regulatory landscape tightening faster than the technology is shipping.

Landing

The Klarna piece I wrote in March was meant to land as a joke. The warehouse full of hot tubs was supposed to be the absurd reductio. Six weeks later the FCA has confirmed its July date and the EU has its November one, and the joke is becoming policy. The bots, increasingly, will need a co-signer. The co-signer will be a regulated entity in a Consumer Duty regime, watched by an ombudsman, declining transactions for reasons the consumer will eventually have a right to ask about.

That isn’t dystopia. And, It’s also, from a 2027 conversion standpoint, going to be very expensive.

I wrote that piece with ASOS in my peripheral vision the entire time. I didn’t name them directly because it wasn’t only about them. But the data describes them precisely. The online pioneer has been in a period of structural recalibration for three years now, and this morning’s H1 FY26 interim results are the first moment where I think the honest answer to “is this working?” might actually be yes. Provisionally. With caveats. But yes.

Let me show my working.

The Numbers

Losses before tax for the half-year period ended 1st March 2026 were £137.9 million, down from £241.5 million in the same period last year. Revenue under £1.2 billion, a 14% drop from £1.3 billion in H1 FY25. On the surface, that looks like a business still going backwards. Revenue down. Still loss-making. Still burning cash in H1. Maybe better revenue - relative to the delinquent customers they had 3 years ago - causing all kinds of crazy.

But the surface is the wrong place to look right now. What matters is the structure underneath it. The turnaround has been about insourcing, better inventory management and to Jose’s huge credit, huge credit - get it? Institutional finance has given them 3 refinance opportunities - bought them runway. And this was not without risk - once bad quarter and covenants kick in - ask many football clubs about this. Debt instruments can be punishing.

Gross margin grew to 48.6% from 45.1%. Adjusted EBITDA rose 51% to £64 million, driven by gross margin gains and cost efficiencies. Profit per order increased 30% year-over-year, supported by higher average order value and reduced return rates. The underlying returns rate reduced by approximately 160 basis points year-on-year.

Eight consecutive quarters of gross margin improvement. Read that again. Not one good quarter. Not a one-off clearance benefit. Eight in a row. That is a structural shift in how the business operates, not a lucky set of conditions.

Inventory fell 10% in H1, with management saying stock remains tightly controlled and health is improving. Sell-through rate on autumn/winter stock improved 60 basis points year-on-year, with the CEO describing it as the highest sell-through ever for the company.

For a business that was drowning in dead stock two years ago , fake fulfilling, sitting on inventory it couldn’t move, running promotions that were hollowing out its margin architecture , those are not small numbers. Those are the numbers of a business that has genuinely changed how it buys and trades.

The Other Number

Buried in the results, and not getting nearly enough attention, is this: for the first time since September 2021, ASOS recorded positive growth in new customers. In March 2026, new customer acquisition rose by 9% for the group. In the UK specifically, new customers increased by approximately 10% year-on-year. The customer database has also shown signs of stabilisation, with churn reducing by 150 basis points during the first half.

This is the one. Everything else in this results pack , the margin improvement, the EBITDA growth, the cost discipline , is the foundation. But a fashion brand with a shrinking customer base is a fashion brand with a structural ceiling. The moment new customer acquisition turns positive is the moment the turnaround stops being a cost story and starts being a growth story.

App downloads moved from a negative year-over-year trend to more than 30% growth. ASOS implemented 50 new app features in H1 , including virtual try-on, ways to style, and a more immersive trending section. Customers engaging with the updated app are buying 9% more.

Building on the success of ASOS.World in the UK, which has reached approximately 3.5 million members , up from 1 million at the end of FY25 , the loyalty programme has recently been rolled out to customers in the US, Germany and Austria.

3.5 million loyalty members in under a year from a standing start. That’s not a vanity metric. That’s the beginning of a owned channel that doesn’t depend on Google’s auction.

Was I Right? Was I Wrong? Let’s Be Honest.

I have been a long standing skeptic and I offered where I was wrong all the way too, but In The Rabbit Hole, I argued that the discovery model that built UK online retail is breaking , that brands substituting earned traffic with bought traffic are on a countdown timer. I cited the Mapp Fashion Intelligence data showing the “danger zone”: brands running paid at 4-8% of their traffic mix producing average operating margins of 3.7%, improving in only 48% of cases.

ASOS has been living in that danger zone. And what this results pack tells me is that they’ve been trying to get out of it , just more quietly than the headline numbers suggest.

Management said ASOS is shifting to more top-of-funnel activity and will continue weighing marketing alongside other investment options such as product quality and customer experience. Marketing spend increased by 50 basis points while the company reduced the cost of acquired visits. Less spend per visit. More top-of-funnel. More owned channel via the loyalty programme. More in-app engagement as a retention mechanism rather than discount dependency.

That’s the right direction. It’s early. The Mapp data suggests UK conversion is still declining industry-wide , the group reported a 9% year-over-year decline in GMV, though underlying data suggests a recovery is underway. You cannot fully insulate yourself from a structural market problem by changing your own behaviour, no matter how well you execute. But the direction of the response is correct. And against the backdrop of that market profile, it makes this story all the more noteworthy. I would love to interview the leadership and offer them credit where it is due.

Where I was right: the inventory-to-cash argument. I’ve been saying for eighteen months that day trading , the real-time read of what’s selling, what’s stalling, and how fast you can turn stock back into cash , is the most important skill in ecommerce right now. The ASOS results validate that thesis from the inside. The CEO explicitly described moving away from “a lot of stock, a lot of promotion, and a lot of performance marketing” toward a model “based on newness, excitement, and full-price sales.” That’s the trading model. That’s the discipline.

The Adidas collaboration sold out in three days, with 80% sell-through and best sellers north of 90% over that period, even after doubling the buy. Overall Adidas sales on ASOS rose 43% year-over-year. That’s what good trading looks like. Limited buy, curated product, emotional urgency, full-price exit. The antithesis of the promotional model that nearly destroyed them.

Where I’m still watching: the revenue line. The group maintained its FY26 guidance, expecting adjusted EBITDA between £150 million and £180 million, with gross margin projected to improve by at least 100 basis points to a range of 48% to 50%. But GMV is still negative , down 9% in H1 , and the market’s patience for “profitable decline” has limits. The question isn’t whether the turnaround is happening. It demonstrably is. The question is whether the revenue inflection comes fast enough to keep the market’s confidence before the balance sheet becomes the story again. But again is this better GMV - can they sustain this against more new customers and less churn - manage this and this is a truly awesome turnaround story.

Free cash flow was a £93 million outflow in H1, described as consistent with seasonal working capital patterns, with an H2 inflow expected to deliver broadly neutral free cash flow for the full year. After the period end, ASOS also repaid £74 million of the remaining 2026 convertible bond. The balance sheet is being managed. But it’s being managed carefully, not comfortably.

The Jeffers Variable

Which brings me back to Natasha Jeffers.

I wrote about her appointment earlier this month , the CV that teaches rather than impresses, the career that spans the full arc from ASOS’s golden era through Arcadia’s decline, Forever 21’s fragility, and Amazon’s operational discipline. I said then that this looked like a bet on the right skill for the right moment.

Today’s results confirm the bet and raise the stakes simultaneously.

ASOS has fixed the foundations. The commercial model is cleaner. The inventory is healthier. The margin is structurally improved. Now the problem is growth , and growth in a market where, as the Mapp data shows, organic discovery is contracting, paid costs are rising, and the old playbook of “buy traffic, push promotions, move volume” no longer works.

What you need in that environment isn’t a brand marketer or a performance marketing specialist. You need someone who can read the trading signal every morning, connect it to the assortment, adjust the pricing and promo architecture in real time, and make the storefront earn its own traffic rather than pay for it. You need someone who understands that womenswear, the priority category, outperformed Group GMV in H1, delivering approximately 10 percentage points of improvement in its growth rate versus H2 FY25 , and can build on that signal rather than wait for the next quarterly report to confirm it.

The rate of sale for Heart items , ASOS’s monthly curated edit , doubled versus other items. That’s curation working. That’s the ASOS of 2008 visible in the data of 2026. The instinct is still there. What it needed was the infrastructure and the discipline to scale it.

Jeffers has built both. At every stop on her career.

What Numbers Would Give The Market Confidence

GMV turning positive. The sequential improvement is real , GMV improved 4 percentage points from Q4 FY25 to Q1 FY26, followed by a further 2 percentage points in Q2 FY26. That trajectory needs to continue into positive territory. The guidance frames GMV running 3-4 points ahead of revenue as flexible fulfilment scales. The market needs to see that inflection before it commits.

New customer retention, not just acquisition. Getting new customers in March at 9% growth is the green shoot. Keeping them through Q3 and Q4 , getting them to their second and third purchase , is the proof. Churn reducing 150 basis points is the early signal. Current trading in Q3 shows further sequential improvement, with womenswear entering positive year-on-year growth. That needs to hold.

UK as the proof of concept. The UK is ASOS’s largest market and the one receiving the most investment , loyalty rollout, out-of-home advertising, returns improvements. The UK performed 4 points better than the rest of the company early in Q3. If the UK can demonstrate what a recovered ASOS looks like , positive GMV, positive new customer growth, improving retention , it becomes the template for Germany, France, and the US.

Gross margin holding above 48%. The structural case for ASOS’s recovery rests on the argument that the new commercial model , full price, curated, lower markdown , is permanent and not a function of current conditions. If gross margin holds or improves while volume recovers, that’s the proof. If volume recovery requires reverting to promotional depth, the story falls apart.

The US. The group faced £7 million in costs related to IEEPA tariffs in the US and has initiated a process to pursue refund claims. The US is the wildcard , tariff volatility, de minimis changes, and a market where ASOS has never quite cracked the cultural code. A stabilised US is enough. A recovering US would change the conversation entirely. Although, they would amke it to “the list”. A list worth taking for cash in the bank. US expansion can and should wait - the US consumer has options.

The Floor

I titled this piece The Rabbit Hole Has a Floor because I think that’s what today’s results represent. Not a full recovery. Not a share price catalyst. But evidence that the descent has stopped , that someone has put their hands out and caught the walls.

The discovery model that built UK online retail is still breaking. The structural pressures I described yesterday haven’t changed because ASOS posted better margins. UK conversion is still falling. CPA is still rising. Vinted and Temu and TikTok Shop are still eating at the edges of a market ASOS used to own.

But ASOS, specifically, is doing the things that give it a chance. Cleaner inventory. Better curation. A loyalty programme with genuine early traction. A new customer trend that’s turned positive. And now, a trading operator in the most important commercial role in the business.

The rabbit hole has a floor. Whether ASOS can climb back out of it is the question that makes this one worth watching for the next twelve months.

I’ll be here.

Sources: ASOS H1 FY26 Interim Results (April 23, 2026); ASOS Pre-close Trading Update (March 25, 2026); ASOS FY25 Full Year Results (November 2026); ASOS H1 FY25 Interim Results (April 24, 2025); ASOS H1 FY24 Interim Results (April 17, 2024); Daily Political , ASOS H1 Earnings Call Highlights (April 23, 2026); FashionUnited , ASOS results: revenue declines, adjusted EBITDA up 51% (April 23, 2026); BusinessCloud , ASOS halves losses as turnaround plan enters final stage (April 23, 2026); Mapp Fashion Intelligence Report 2026 (Sarah McVittie); IMRG Online Retail Index (Andy Mulcahy).

The conversation we recorded for the season opener of The Ostrich Report wasn’t a forecast. It was a post-mortem with the corpse still warm. Malte had called the biggest marketplace stories of the last five months already, in a quiet chat we’d had some time back. Hendrik and I sat there and basically held a light while he described how we got here.

Let me walk you through the receipts.

The Amazon Door Charge

Malte’s framing for last year’s robots.txt crackdown was elegant and a little brutal: Amazon didn’t panic-close the doors on Perplexity and every other agentic bot because it was scared. It closed them because it needed time to build doorways you’d have to pay to walk through.

Six months later, here we are. The $1,400 annual Selling Partner API subscription kicked off on January 31. The monthly per-call usage fees start this month, April 30, roughly one week from now. Every Perplexity bot, every Copilot assistant, every ChatGPT Go shopping agent that wants to window-shop inside the Everything Store is about to become a line item. Amazon has also, for good measure, filed a computer-fraud complaint against Perplexity.

That isn’t Amazon defending itself. That’s Amazon monetising a new traffic category before the category has a Gartner quadrant. The quote that stuck: “Amazon was simply buying time.” Time, it turns out, is expensive.

The PayPal Move Malte Saw Coming

When Malte and I first kicked this around months ago, he told me the companies who win this cycle are the ones making merchant catalogues fluent in the language of agentic surfaces. Not storefronts. Not marketing. Catalogue hygiene, product data, inventory sync, boring plumbing.

PayPal announced on January 22 that it was buying Cymbio for an estimated nine figures. The pitch on the tin: sync product data, inventory, and orders across Microsoft Copilot and Perplexity, with OpenAI and Gemini next. “Merchant-of-record intact,” the press release insists. That’s Malte’s thesis with a logo and a term sheet.

As we covered in the de minimis and TikTok long-game pieces earlier this year, the pattern now is consistent: the money moves toward whoever controls the layer just underneath the consumer interface. In 2018 it was the checkout. In 2022 it was the ad network. In 2026 it’s whoever can make your product legible to a bot in a way the bot trusts enough to surface it.

The John Lewis Footnote

We spent a few minutes on John Lewis too, and here the critical reader earns their keep, because that one was already public when we hit record. On March 9, John Lewis announced a TikTok Shop launch and agentic commerce integration via commercetools, with ChatGPT and Gemini named explicitly as discovery surfaces. £800m transformation budget behind it.

If you’re Irish or European watching this from across the water, don’t read that as John Lewis finally catching up. Read it as John Lewis quietly re-architecting the shop window before anyone else on the UK high street notices the glass has moved. That’s the Kevin Kelly point about technological shifts that were never really invented — they were latent, waiting for the right trading conditions. Agentic retail didn’t arrive last month. The plumbing for it was already live. John Lewis just agreed to connect.

The Continental Divide

Hendrik pulled the camera back to the global map, and this is where he earned his seat. His take: this cycle has Cold War geography. China’s adoption of AI commerce is already common sense, Alibaba’s Qwen app hit ten million downloads in a week, silently. The US makes enough noise for three continents. Europe is sitting in regulatory dread. Africa and Latin America come last, as they always do, and most likely via government-run rails rather than platform rails.

Ireland sits in both worlds, as usual. Most of the big US platforms host their EU headquarters a cab ride from my office. Every one of them is writing an answer to the DSA at the same time they’re writing an answer to Rufus. The merchant on the ground is the one carrying both bills.

Hendrik’s line of the episode, which I think I’ll be stealing for the next six months: “Kiss the ring. The ring’s name is Rufus.”

The Turn

Here’s the complication, and it’s worth sitting with. Malte’s predictions held. That doesn’t mean his prescriptions will.

The scariest thing about Amazon’s playbook right now isn’t the paywall on the API. It’s that the paywall feels reasonable. The scariest thing about the PayPal/Cymbio deal isn’t what PayPal bought, it’s that merchants now have three or four more doors they have to pay to appear behind. Malte also dropped a stat worth your attention: in some categories, baby, pantry staples, beverage, non-branded search has already crossed 90%. Every brand manager you know will tell you those are the categories where trust matters most. The review layer already ate the brand layer. The brand team was at an offsite and missed the memo.

The Landing

At the end of the recording I pitched Malte and Hendrik an idea I want to build into this season: a Cluetrain Manifesto for agentic commerce. The original Cluetrain Manifesto tried to write down what the internet was actually for, before it calcified into banner ads and cookie consent pop-ups. Agentic commerce is at the same inflection point now. Someone should write down what it’s actually meant to do, before the toll booths become the only thing it’s ever really been.

That’s what this season is for.

A sponsor joins us from next week, more on that shortly. If you’re moving the needle on marketplaces, agentic AI, cross-border trade, logistics, or any of the messy stuff in between and you’d like to guest or back the show, come talk. We’re building The Ostrich Report toward the number one seat for marketplace news globally this year, and we’ve got fifteen episodes and an Atlantic to cross.

Hendrik, thanks for the company. Malte, thanks for the receipts.

See you next Thursday.

Since this episode was recorded, Chris has become a father for the second time. I want to flag that up top, because everything in the twenty-odd minutes that followed makes more sense once you know it. A man who talks with the quiet conviction of someone who has thought a lot about what home should feel like, and is now in the business of building it again, in real time, for a small person who will grow up inside it.

The first time I met Chris was 2024, in a room in Dublin that was struggling to breathe. Irish brands pitching to US buyers, nobody quite sure what to say. The buyers were giving nothing away. The brands were second-guessing themselves. The temperature dropping by the minute. Chris jumped in with feedback. Unsolicited, generous, sharp. He provided the oxygen in a room that needed it.

That’s the starting point for this conversation. Empathy. Where does it come from, how do you sustain it in a professional context, and is it the thing that actually separates the people who last from the people who don’t.

Chris’s answer is grounded in loss. Small family, members going too soon, and the way those absences shaped his instinct to bridge. Through loss, empathy materialises. It’s a sentence that does a lot of work. What he’s describing is the process of turning grief into a professional operating system without ever naming it as such. The curiosity that comes from it, the reflex to bring people into a conversation rather than hold them outside of it, is not a trick he learned at sales school.

He pushed back when I suggested the constant curiosity must be exhausting. A self-proclaimed introvert with some extrovert quality, he framed it as quality over quantity. Depth over reach. A wallflower at the corner of the bar until you put him in a room with a clear purpose, at which point he turns on and goes. I recognised the description immediately. Half the best operators I know work that way. The ones who don’t are usually selling something.

Then he dropped the line I’ve been thinking about ever since. He name-checked Mary Beech from Kate Spade, the marketer who took a customer persona and turned her into someone called the Kate Girl. A twenty-something who would spend a whole paycheck on a lamp, eat cake for breakfast, and aspired to travel the world. No cohort language. No demographic slicing. A person.

Read more

Over the last 2 weeks I have been consider some comments made when Sundar Pichai sat down with John Collison and Elad Gil on The Cheeky Pint (the most Irish of Podcasts) and the conversation wandered into territory that, on the surface, has nothing to do with ecommerce. Except it does. It has everything to do with ecommerce. It has everything to do with how any of us running businesses right now should think about the decisions we make, or more to the point, the decisions we keep deciding to defer. At least that is my take away - talks of compute and supply and demand were equally interesting but I was particularly smitten by this flippant comment.

Pichai was talking about Google’s internal planning philosophy. The Five-Year Horizon. The idea that when you’re operating at the scale of something like Google, you can afford the luxury of patience. You seed a project, you nurture it, you don’t expect returns until the timeline has breathed out fully. And to be fair, when you’re building something that billions of people use, there’s a version of that logic that makes sense. You don’t want the infrastructure holding up search for half the world’s internet users being redesigned on a whim by someone who had a good idea on a Tuesday. Let’s not forget, they are not building or maintaining just one thing. They have many many businesses to consider.

But then he said something that cracked it open. Because the Five-Year Horizon is now a comfort blanket. A way of making the absence of urgency feel like strategic discipline. And Pichai, to his credit, said that out loud.

We’ve been doing versions of this in ecommerce for years. I know because I’ve sat in the rooms. The platform migration that’s been on the roadmap since 2022. The pricing strategy review that needs three more quarters of data before anyone commits. The international expansion plan that is perpetually almost ready. The Five-Year Horizon is seductive because it gives every delay a rational frame. We’re being considered. We’re being thorough. We’re not rushing. And meanwhile the market is not doing any of those things. But it is in the fog of indecision that up starts and new brands are making their arbitrage moments turn into money.

“The gap between research breakthrough and product utility has shrunk from years to months”

The line that Pichai threw out almost as an aside is the one that has stayed with me longest. The gap between breakthrough and application used to be measured in years. A model architecture paper would land and the industry would spend half a decade working out what to actually do with it. Now the gap is months. Sometimes weeks. The transformer paper dropped in 2017 and in the room with Pichai and the Collisons someone pointed out that if the Five-Year window applied, the Gemini moment should have arrived in 2022. Pichai’s answer was essentially: yes. And we nearly missed it because we were still operating on the old clock.

Read more

Not the kind that follows a long weekend. (for me caused through excessive red wine). It is The kind that shows up on your P&L.

App bloat.

Agency bloat.

Headcount bloat.

It builds slowly. Quietly. One completely reasonable decision at a time.

And then one Tuesday morning, someone in a glass-walled room says the word efficiency and the scissors come out.

The popular opinion and it is very popular right now:

Cut fast. Blame the macro. (meta thoughts for your own consumption here, I am not getting cancelled). Most people post something warm about AI unlocking new ways of working.

Call it transformation.

Move on.

The unpopular opinion. The one nobody puts in the press release.

Your bloat happened for a reason. The apps were solving real problems. Badly, expensively, and in triplicate but real problems. The agency roster grew because someone internally didn’t have the skill, the time, or the brief to do it themselves. The headcount expanded because you were building for a version of the business that didn’t show up. That is a planning problem. Not a people problem. Cutting without understanding is a painkiller.

Six months later, the same symptoms return. Different packaging. Same bill.

One of the most expensive things in ecommerce isn’t bloat.

It’s fixing the same problem twice. This has been a public service announcement on behalf of operators everywhere who have been handed a set of problems and told to sort it out by Friday.

If that’s you talk to Sumoblue before you reach for the scissors.

We diagnose before we prescribe. Helpful when you been asked to get the snip.

The V Spot Nearly News. Seen through the eyes of a madman. Sponsored, this week, by common sense and SUmo Blue.

HOT OR NOT — WEEKEND POLL

Agency bloat. What’s your move?

🔥 HOT: Cut fast, explain later

❄️ NOT: Diagnose first, cut with precision

😐 Pretend it’s not happening until Q3

🙋 I am the bloat

— So, Dave.

, So, Vinny.

— You’ve been sayin’ Ireland’s at its cultural peak.

— I have, yeah.

— Peak like Croagh Patrick or peak like the queue for a chicken fillet roll in London?

— Higher. Spiritual, even.

— Jesus.

— Exactly.

— Go on so. What’s the evidence?

— Music.

— Music?

— Fontaines, Kneecap, CMAT, Bicep

—, Ah stop., I won’t., You’re tellin’ me Ireland’s takin’ over the world because some lad in the Phoenix Park owns a modular synth?

, Pretty much, yeah.

— And cinema.

— Sure that’s just Cillian Murphy bein’ handsome for a living., Handsome men can be cultural too, Vinny.

— Go ‘way outta that.

, And Guinness.

— Guinness isn’t culture.

— It f*ckin’ is when it’s the best-selling drink in the UK.

— In fairness, that is funny.

— They’ve queues in London for spice bags.

— Jaysus.

— Chicken fillet rolls stretchin’ to Hackney.

— We’ve exported obesity, so.

— Exactly.

— What about Irish business culture?

— Oh yeah. We’re everywhere.

— Like Japanese knotweed.

— More like bamboo. Softer.

— Softer me hole.

— And why is this all happenin’, Dave?

— We’re the most socially fluid people in the world.

— Fluid?

— Like Guinness, yeah.

— Well I’ll drink to that.

— Tell us this: you’re always talkin’ about creativity an’ content.

— Go on.

, How come you’re so good at it?

— Culture.

— Culture?!

— I’m seeped in it.

— Like a teabag left too long?

— Exactly.

— Jaysus, that’s grim.

— But is ecommerce meant to be experiential or entertainin’?

— Yes.

— Yes to which?

— Both.

— Dave, don’t be annoyin’ me now.

— I’m sayin’ brands need to solve, not sell.

— Solve what?

— Problems.

— Like the way Gymshark solved joggin’ in the dark by givin’ women a run club?

— Precisely.

— Fair play to them. They’d want to give the rest of us a lift home too.

— You’re writin’ a second book, aren’t yeh?

— I am., What’s it about?

, Irishness thrivin’ globally.

— So a book about you, basically.

— Shut up.

— I’m only sayin’ what the people are thinkin’.

— It’s culture, business, tech, the whole lot.

— Will there be a chapter about chicken fillet rolls?

, Possibly.
— Well then it’ll sell.

Before We Meet Dave: The Bus That Didn’t Come

Half-five in the morning. Limerick. 2003. A bus to Punchestown that may or may not materialise. A teenager standing in the dark wondering whether the whole thing is more hardship than it’s worth.

The bus didn’t come. Or it was late. The details are unimportant. What matters is what happened next: they figured it out. They pushed through. They got there. And when Witness opened up, the stages, the sound, the gathering, the specific sensation of thirty thousand people deciding, collectively, to be somewhere together, something shifted.

Dave Morrissey got a tattoo of it. On his arm. Witness, with a pint. Which tells you almost everything you need to know about the man.

That morning, the hardship, the push, the arrival at something better on the other side, became a private philosophy he has carried through fifteen years in tech, two career-defining platform jobs, a book, a second book in progress, and a consulting practice named, in Irish, after stories. Scéal. He benchmarks his life in milestones: before Witness, after Witness. Before Facebook, after Facebook. Everything that came before and after the moments that cracked something open.

He is, in the best possible sense, a man who is still running towards the festival.

Who Is Dave Morrissey?

Dave is a Growth and Culture Consultant operating under the name SCEAL, having spent over a decade across startups to Big Tech, at Meta helping grow brands like HUEL, Alibaba, LEGO, Very and Gymshark, and at TikTok leading the Retail and eCommerce Lifestyle team, overseeing the growth of athleisure, beauty and home brands. He is also the author of Grow Like Tech, a playbook for applying the growth lessons of Big Tech, culture, scalability, product development at speed, to any business in any industry.

He is from Limerick. He lives in London. He is writing a second book about Irish culture thriving globally, which is the most on-brand thing a person can do when their entire worldview is built on the idea that where you’re from is an asset, not a footnote.

He believes it comes down to culture and the why, it’s not the code that drives these companies, it’s the people making decisions and how they work together. He has been saying this in rooms from Manchester to the Shard, and anyone who’s heard him say it in person will tell you it lands differently when Dave says it. Because he means it in a way that isn’t strategic. It’s just how he’s wired.

He also worked with the Rubberbandits early in his career. Which is either a biographical footnote or the most important sentence in this review. I’m going with the latter.

Read more

A year later, a relief valve blew on a 500-gallon stainless steel tank during a Yankees game. The loudest explosion he’d ever heard, immediately followed by someone shouting the two most dangerous words in the English language: “Free beer!” Beer knocked a customer off their chair and flowed through the restaurant like a very specific kind of flood. That was Bryan House’s early twenties.

Now he’s CEO of Elastic Path. He’s swapped IPAs for APIs. And he might be the most interesting product mind in composable commerce right now.

Bryan House has a degree in brewing from the Siebel Institute of Technology, the oldest brewing school in America, founded in the 1870s, which means it predates prohibition. He brewed beer professionally for eight years. He has an MBA from Harvard. He was a founding team member at Acquia, where he helped build the company to over $170 million in revenue. He ran product at Neural Magic, a deep learning startup. He loves GWAR, and if you have not looked up GWAR, please do, because your Wednesday needs it. His song pick was “Left of the Dial” by The Replacements, which tells you everything you need to know about the man in about three minutes and forty seconds.

What got me was how directly the brewery skills translated. Every one of those 37,000 people walking across from Fenway would come in and ask for a Bud Light. Bryan didn’t have Bud Light. What he had was a craft beer story and about thirty seconds to tell it. Product marketing on the fly, in rubber boots and a jumpsuit, face-down with the crowd on top of you. He told me that’s the foundation of everything he’s done since, the ability to communicate what your product actually is to someone who didn’t know they wanted it. That’s not a keynote skill. That’s a bar-fight skill. And it shows.

Bryan is what I’d call a product CEO, and that matters. The commerce technology space is thick with financial operators, people who run companies like spreadsheets and talk about TAM the way other people talk about weather. Bryan talks about products the way a brewer talks about beer. He believes in what he’s making. He has a self-described bias for the new, and he’s honest enough to call it both a strength and a weakness. The punk rock underpinning isn’t an aesthetic. It’s a disposition. He’s comfortable being different without having to shout about it. That’s rare in a CEO chair,. His shorthand for Elastic Path’s positioning is “commerce for weirdos,” and I am completely here for it.

Read more

The average number that actually get implemented?

Three. On a good year.

Somewhere out there, eleven agency decks are sitting in a folder called “To Review FINAL (2)” gathering digital dust and emotional baggage.You are not alone. There are support groups. I’ve seen the folding chairs.

Here’s what nobody says out loud about RFP season:

The process is broken before the first deck lands. Most brands go to market without a clear internal brief. They know something isn’t working, CAC is climbing, retention is flat, the site feels stale, but the RFP goes out as “we need a new agency” instead of “here’s the specific problem we need solved.”

So fourteen agencies respond to a vague brief with a vague pitch. And the brand picks the one with the best fonts.

Then three months in, everyone’s frustrated because the deliverables don’t match the expectation that was never written down in the first place.

The RFP isn’t the problem. The absence of a proper scope is. The brands that get this right do the hard work before they open the process. They define what’s broken. They set measurable outcomes. They ask agencies to respond to a real problem, not a PDF beauty contest.

RFP season doesn’t have to feel like speed-dating at a conference where everyone’s wearing the same lanyard and saying “omnichannel” like it’s a personality trait.

This is a PSA from myself and the team at Sumoblue because fourteen decks is a symptom, not a strategy. If your RFP process needs fewer decks and more decisions, give them a shout.

Right. Let’s do the news.

SPONSORED BY

We somehow manage to still be sponsored by Omnisend — a company brave enough to associate with us publicly. Omnisend the email platform not trying to be your CRM.

THE STORIES

Story 1 — A flip-flop story just in time for the summer. Amazon and USPS:

Amazon and the US Postal Service finally agreed to keep seeing each other. USPS retains about 80% of Amazon’s delivery volume, that’s still over a billion packages a year. Now, Amazon had threatened to cut two-thirds. The USPS said it would literally run out of cash by October. So what we got was a compromise or, as it’s known in logistics, the thing you do when you realise you can’t actually deliver to 160 million addresses by yourself. [deadpan] Amazon is spending four billion dollars building out rural delivery. USPS has accumulated 118 billion in losses since 2007. It is still less than the 128BN AWS generates annually. If this were a marriage, you’d call it co-dependent. But in logistics, we call it strategic alignment. Same thing.

Story 2 — Google’s Universal Commerce Protocol Gets an Upgrade

Google updated its Universal Commerce Protocol this week. For those unfamiliar, UCP is the open standard that lets AI shopping agents talk to retailers, check inventory, build carts and check out all inside Google Search and Gemini. New features include multi-item carts and loyalty linking. Which means an AI agent can now buy three things at once from your store and apply your loyalty discount. And you, the retailer, don’t even get to say hello. Google’s framing this as “frictionless.” Retailers are framing this as “we just became a warehouse with a checkout API.” If you’re European watching this, UCP is US-only for now. But don’t get too comfortable. It’s coming. And it’ll bring GDPR questions the size of a French customs declaration.

Story 3 — Saks Global Rises from the Ashes (Sort Of)

Saks Global, the luxury retailer that went Chapter 11 in January owing Chanel 136 million dollars yes, Chanel,has secured 500 million in exit financing and plans to emerge this summer. Over 650 brands are shipping again. Inventory in March was up 18%. Online conversion jumped 11%. They even reversed three planned store closures. If this were a rom-com, we’d be at the part where the couple gets back together. But in luxury retail, trust is rebuilt one paid invoice at a time. Good luck to them. And send Geoffroy van Raemdonck a coffee, the man’s done this bankruptcy thing before. At Neiman Marcus. In 2020. Some people collect stamps. He collects restructurings.

Story 4 — Amazon Buys a Robot Dog

Amazon acquired RIVR, a Swiss robotics company formerly called Swiss-Mile, that makes a four-legged wheeled robot designed to carry parcels from the van to your doorstep. It’s a dog on roller skates. I am not making this up. Their CEO literally called it that. It can climb stairs, carry 60 kilos and navigate uneven terrain. So basically it does everything your current delivery driver does, except it doesn’t ring your bell and run away. Amazon shut down its own Scout delivery robot programme in 2022. Now they’ve bought a Swiss one. Because if there’s one thing the Swiss know how to do, it’s be neutral while delivering things precisely on time.

Story 5 — The WTO Drops the Ecommerce Ball

Here’s one that flew under the radar. The World Trade Organisation’s moratorium on customs duties for digital goods, the one that’s been in place since 1998, just collapsed. For the first time in nearly three decades, they couldn’t reach consensus. What does that mean? It means countries can now start taxing digital commerce at the border. Streaming services, SaaS, digital downloads, all fair game. If you’re running a cross-border digital business, you may want to Google “tariffs on code.” Actually, they might tariff the Google too.

Story 6 — The Agentic Commerce Cold War

The divide is widening. Google rolls out UCP and says “let the agents shop.” eBay says “absolutely not” and bans AI buy-for-me agents from its platform. Amazon is suing Perplexity while simultaneously building its own Buy for Me feature. So to recap: Google wants AI agents to buy things. eBay says no. Amazon says no to yours but yes to mine. And the merchant in the middle is standing there like a bartender at a bar fight going “does anyone actually want to buy something?” The agentic commerce future is coming. It’s just arriving via three different doors, with three different bouncers, and none of them are talking to each other.

Big thanks again to Omnisend for backing the show. We offered them a four-legged robot to deliver the invoice. They said they’d prefer email. Fair enough. Head to Omnisend.com and tell them the lad with the accent sent you.

Thanks for watching the V Spot. That’s the 6 minutes you won’t get back this week.

That’s Finn. If you know him, you already understand. If you don’t, this episode is about to fix that.

Dave Finnegan, Finn to his friends, and now to you, is an operating partner at BlackFinn, an advisory board member for NRF’s CIO and CMO Councils, a member of the Explorers Club, a man currently pursuing a master’s in anthropology at Harvard, and the kind of person who makes you feel like a better version of yourself just by being in the room. He helped launch Build-A-Bear. He ran the show at Orvis during their best years. In 2022, he took a year out to visit every continent and interview indigenous leaders, scientists, and, his favourite story, the librarian of the world’s oldest library at the Monastery of St. Catherine. He’s the Ross Geller of our industry, and I mean that as the highest compliment.

We started with ancestry. His dad, a quiet cowboy poet from a mining town in Western Montana. His mum, a driven Scot. A handful of aunties who made sure he could get away with nothing. That combination, quiet resilience and entrepreneurial steel, runs through everything he does. When I asked him about the strong women who shaped him, he traced a direct line from those aunties to Maxine Clark at Build-A-Bear, to Vicky Cantrell, to the women throughout his career who matched the energy he was raised with. Not many people draw that line so clearly, or so generously.

As always when speaking with him, came a moment I didn’t expect. Finn told me about sitting fireside at the Explorers Club with some young engineers who were working on the first anthropomorphic robot for the International Space Station. Two arms, no torso. And the thing that had consumed their research wasn’t power or processing. It was likeability. They’d learned that if a robot moves too fast, gets too close, acts too mechanically, humans reject it. The unwritten rules of what makes us human, our comfort with proximity, our response to rhythm and pace, had to be programmed in. I asked him if the tools are changing the humans. His answer was better than my question: the tools are teaching us about ourselves.

This is where the conversation turned into something I’ll be thinking about for weeks. I’d sent Finn a story about Canvass White, a 21-year-old engineer who, in 1817, walked over 2,000 miles across England on his own dollar, studying how the British built their canals. He came back with drawings, surveying instruments, and, crucially, knowledge of hydraulic cement that would waterproof the locks of the Erie Canal and save what was being called Clinton’s Folly from becoming a hundred-year money pit. Bill Bryson wrote that Canvass White didn’t just make New York rich, he helped make America. And nobody paid him for the cement. Over 500,000 bushels used, not a penny honoured.

Finn’s read on it was characteristically sharp: real progress happens when someone takes responsibility for a problem that no one owns yet. Everyone saw the water leaking out of the canal. Only one person went to do something about it. That’s the line between observation and action, and it’s the line most of our industry is still standing on the wrong side of.

And then he said the thing that I think defines this episode: “They learned how to build the canal by building the canal.” No playbook. No case study. No five-year-old proof point to benchmark against. They figured it out in the doing. His argument, and I think he’s right, is that being early with AI is less about being right than it is about being willing. Jefferson called the Erie Canal project “near madness.” They built it in eight years. Willingness beats certainty. It always has.

What I love about Finn is that this isn’t just philosophy for him. Through BlackFinn, he evaluates companies for investment, and he told me that at least half of what they assess is the person, not the product. Can they attract others? Will they own problems nobody else will touch? Do they build rockstar teams? In a market where every investment target is months old, not years, the human part of the equation isn’t soft skills. It’s the whole game.

We talked about how anthropology gets miscast as old stuff. It’s not. It’s the study of human meaning, and human meaning didn’t stop being relevant when we started building large language models. The people who understand why humans do what they do are going to be the ones who build the things humans actually want. That’s not a prediction. That’s a pattern.

Finn’s song pick was Temple of the Dog, “Hunger Strike.” About remembering where you’re from as you become successful. About the people who got you there. It couldn’t have fit more perfectly if he’d written it for this episode.

And yes, I did ask about what goes on under the kilt. A Scottish gentleman never tells.

This is Season 3 of the Struggle Bus. We’ve changed the look but not the soul. The conversations are still about the beautiful friction of building, the humans behind the hype, and the belief that shared empathy isn’t a soft skill, it’s the fuel that drives innovation. Finn was the perfect guest to open with because he embodies all of it. Pour a large one. We’re just getting started.

Season 3 of The Struggle Bus is brought to you by Sumo Blue and Omnisend. There’s still room for one more at the table, if you’re interested, you know where to find me.

Not slowing. Over. She presented research showing that the volume of fashion goods sold in the UK in 2028 will be approximately the same as what was sold in 2019. The rising tide has stopped rising. Every boat must now paddle for itself.

And it gets more specific than that. One in four items sold in the UK is now second-hand. Vinted, which Sarah describes with the mix of admiration and slight personal complicity of someone whose kids are very much on it, is generating revenue of around £900 million with EBIT margins around £100 million. It is, in her words, an extraordinary business, because it is a wealth creator: it turns dormant wardrobe assets into spending power, and about 36% of that money cycles back through the platform.

The Text Message That Built Everything

In 2003, before anyone in fashion retail had heard the phrase “product data enrichment,” before Vinted existed, before Shein’s founder had pivoted from SEO to fast fashion, a woman named Sarah McVittie had an idea about SMS.

She co-founded Texperts, the world’s first Q&A SMS service, inspired by her first job working as an analyst at an investment bank. In 2008, it was sold in a multi-million-pound deal to its largest competitor. Sarah was, at that point, already on her second chapter before most people in the industry had finished writing their first.

The leap from investment banking to SMS to fashion tech is not an obvious trajectory. But it makes complete sense once you understand what connects all three: Sarah is interested in the gap between what information people have and what information they need. At a bank, that gap is the whole game. At Texperts, it was the entire product. At Dressipi, which she co-founded in 2010 and built into one of the UK’s most respected fashion AI platforms, it became a mission, to decompose the product into data points as a customer sees it, and put a customer and fashion lens onto those products, to truly understand the customer’s intent in real time.

Mapp acquired Dressipi in early 2025, and Sarah moved into the role of VP of Marketing, carrying with her fifteen years of data that most fashion retailers would trade significant percentages of their margin to access.

She also brings, around Christmas week in London near St. Paul’s, the kind of energy that makes you want to immediately go and find a sandwich shop called Fuzzy’s Kitchen. Which tells you something about her too.

Who Is Sarah McVittie?

Sarah is the person in the room who has already done the maths while everyone else is still debating the question. She is warm, specific, and completely unsentimental about the state of an industry she clearly loves, which is the most useful combination of qualities a commentator can possess.

She was named by Management Today as one of the top 35 female entrepreneurs under the age of 35, and by The Times as one of the UK’s top rising female entrepreneurs. She has been writing about fashion retail’s structural problems for years, with the patience and precision of someone who built their career on understanding what data actually says versus what people want it to say.

She is also the kind of person who, when asked what song her career would be, says We Are Family by Sister Sledge, and means it completely, because she has spent two startups working with people ten times better than her at everything, and she knows exactly why that worked. That’s not false modesty. That’s the founder’s instinct for complementary capability, which is rarer than it sounds.

She is passionate about using data to drive more efficient retail experiences and operational processes, and has spent the last several years building the index that proves how few businesses are actually doing it.

The Origin: Two Startups, One Through-Line

The through-line from Texperts to Dressipi to Mapp is a single obsession: closing the gap between what a product is and what a customer needs. At Texperts, that gap was informational, people wanted answers, quickly, on a device they already carried. At Dressipi, it was personal, people wanted to find clothes that worked for them specifically, not just clothes that existed. At Mapp, it’s operational, helping retailers show each visitor the items they’re most likely to buy and keep, with products automatically tagged with three times more detail, delivering incremental improvements to profit, revenue, returns, and sell-through rate.

What she built at Dressipi was a translation layer between how merchants describe products and how customers actually think about them. Merchant language says “midi dress, polyester blend, floral print.” Customer language says “I need something for a rooftop wedding that covers my arms.” The gap between those two descriptions is, in Sarah’s framing, where most of the industry’s problems live, and increasingly, where its future is being decided as LLMs take over discovery.

She saw that problem clearly in 2010. The industry is only now catching up to why it matters.

The Share-Game Market Nobody Wants to Admit

Here is Sarah’s central argument, stated plainly, because it deserves to be: the volume growth era for UK fashion retail is over.

Not slowing. Over. She presented research showing that the volume of fashion goods sold in the UK in 2028 will be approximately the same as what was sold in 2019. The rising tide has stopped rising. Every boat must now paddle for itself.

And it gets more specific than that. One in four items sold in the UK is now second-hand. Vinted, which Sarah describes with the mix of admiration and slight personal complicity of someone whose kids are very much on it, is generating revenue of around £900 million with EBIT margins around £100 million. It is, in her words, an extraordinary business, because it is a wealth creator: it turns dormant wardrobe assets into spending power, and about 36% of that money cycles back through the platform.

Read more