Why Marine Layer Thinks Agentic Commerce Is the New SEO

Michael Natenshon · Founder, Marine Layer

Why Marine Layer Thinks Agentic Commerce Is Just the New SEO

Chord Commerce 42 subscribers

Michael Natenshon couldn’t find a soft enough t-shirt, so he spent eighteen months developing his own fabric and launched a pop-up shop just to collect email addresses. That accidental pop-up became Marine Layer’s whole strategy — 50+ stores and counting, built on a simple rule: never raise so much money that you stop having to be disciplined. Now he’s betting that agentic commerce is just the next version of SEO.

Behind the Expert

Michael Natenshon founded Marine Layer in 2009 after his girlfriend — now his wife — gave him an ultimatum over a beloved, worn-out college t-shirt he refused to throw away: it was her or the shirt. She won, and Natenshon, an econ major with a finance background who'd gone back to business school at Berkeley to figure out his next move, decided to just make himself a new favorite shirt. It took a year and a half to develop a custom, incredibly soft fabric, knocking on the doors of L.A. knitting mills with no industry connections, no real money, and no idea what a minimum order quantity or a bill of lading was. He launched Marine Layer online-only in 2009, in the same DTC wave as early Bonobos, then opened what was meant to be a modest pop-up shop to collect email addresses for the website — it worked so well that it reshaped the entire business into an omnichannel retailer years before that was fashionable. He built the company with his childhood best friend Adam, who left a finance career to co-found it with him, and the two have run it together ever since. Marine Layer now operates 50+ stores nationally.

The Quick Hits

• A pop-up meant only to capture email addresses became the accidental foundation of the brand — the tactile, in-person experience of the fabric outperformed trying to sell "incredibly soft" as an online claim.
• Going omnichannel almost from day one was operationally harder (different tech stacks, inventory distribution, POS vs. online checkout) but became a durable advantage rather than a distraction.
• Constrained capital was a feature, not a bug: never raising too much money forced discipline and protected the kind of decision-making that builds a real business instead of chasing growth for its own sake.
• A founder partnership works on complementary-plus-overlapping skills and the ability to trust each other through conflict — not a strict, non-overlapping division of labor.
• Agentic commerce is "the new SEO" — the same fundamentals (reviews, relevance, rich product content) matter, but for a machine that reads everything rather than a human skimming for the shortest possible answer.

The shirt that started it

Natenshon's origin story is the founder-with-a-personal-itch archetype taken almost literally: a ratty, beloved college t-shirt, an ultimatum from his girlfriend, and a year and a half spent teaching himself textiles and pattern-making from scratch. He had no industry relationships and no capital — knitting mills in L.A. turned him away because he didn't understand what a minimum order quantity was, and he didn't have a business credit line to open one.

Not coming from apparel or consumer brands turned out to be an advantage. Without legacy assumptions about how to run a clothing business, Natenshon and his co-founder Adam were free to just ask what actually made sense at each step.

A pop-up that became the business

Marine Layer launched online-only in 2009, following the early DTC playbook of the era. But converting a customer with the claim "this shirt is incredibly soft" turned out to be much harder online than expected, so the team ran a pop-up shop mainly to collect email addresses for the website. It outperformed every expectation, because people could actually touch and feel the fabric — and something about being in a physical space with other people responding to it the same way created its own kind of pull.

That first pop-up was successful enough that Natenshon signed a five-year lease and asked Adam — his childhood best friend, then working in finance — to quit his job and co-found the company with him. Adam said yes. The store was profitable in its first month, and every month after.

Omnichannel before it had a name

Marine Layer ended up multichannel almost by accident, years before "omnichannel" was standard advice for DTC brands — and Natenshon says he wouldn't do it any other way, even knowing how much harder it made the operational side: different inventory distribution, different point-of-sale technology, an entirely different set of requirements from online checkout.

The company has opened five to ten stores a year for roughly a decade (pausing during COVID), now past 50 locations — guided by what Natenshon calls a simple, consistent mantra: open stores in places you actually want to be and visit yourself.

Constraint as a feature

Marine Layer's early capital came from friends and family — including an investor Natenshon met waiting in an Apple Store line, who noticed the shirt he was wearing and asked about the brand. Natenshon is candid that never raising a lot of money kept the company from being pulled into a pace of growth that didn't make sense for the business.

He ties the same discipline to how a brand earns durability: the slower and more deliberately customers discover you, the stickier and longer-lasting that loyalty tends to be — a shortcut in speed is often a shortcut in staying power too.

The partnership

Natenshon started Marine Layer alone before Adam joined about a year and a half in. His read on what makes the partnership work isn't a strict division of labor — it's complementary skills with enough overlap that either of them can argue the other side of a decision, resolve conflict, and still trust each other afterward.

Agentic commerce as the new SEO

Natenshon treats agentic commerce as a new discovery channel rather than a replacement for retail — and points to something counterintuitive in his own data: online sales are leaning into AI-driven discovery even as retail sales are stronger than ever, which he reads as people holding onto real, in-person experiences precisely because so much else is changing.

On what actually drives visibility inside a tool like ChatGPT or Claude, his comparison to SEO is direct: the old instinct was to distill product copy down to the fewest possible words, because human shoppers don't read fine print. AI assistants are "voracious readers" that draw on everything available — reviews, product descriptions, even discussion on sites like Reddit — so the content strategy has to flip toward being more complete, not less.

He's more cautious about advertising directly inside AI chat interfaces, wary of disrupting what feels like a private, trusted interaction — preferring other monetization models, like an affiliate or referral structure, if the tools are going to handle transactions end-to-end. Internally, his team is already using AI as a creative accelerant: designers use visual tools to try a new pattern on an existing product photo instead of commissioning new samples, and leadership uses it to process data faster — augmenting the work rather than replacing the people doing it.

Sound Bites

• “We thought about it as, this is how we're going to acquire customers for the website. But it just happened to work out for our business — you get a feel of the brand when you come into the store.”
• “Every channel was a headache. It's a ruse to think that some other channel is going to be this green pasture where it'll be incredibly easy. But there's a synergy and a strength, because each one has positives and negatives.”
• “If it's something you like doing, you're going to keep doing it. That's why retail hasn't died in the way people predicted — going to our stores is a fun experience, and something that's enjoyable.”
• “We used to try to simplify our information — people aren't going to read this much fine print. We really need to distill things down. Now you have something that's a voracious reader, so some things stay the same and some things just change.”

The Chord take

Marine Layer's whole arc is a case study in constraint as strategy. A tactile product that couldn't sell itself online forced a pop-up; the pop-up worked better than planned and forced an early, harder-to-manage omnichannel model; a lack of capital forced discipline instead of a chase for scale. None of it was the original plan, but each constraint pushed toward a more durable business rather than a faster one. That same instinct carries into how Natenshon thinks about agentic commerce: not a reason to panic or abandon retail, but another surface to show up authentically on — one that rewards the same substance (real reviews, real content, real product quality) that built the brand in physical stores in the first place. The brands that treat AI discovery as a shortcut to game will find, the same way Natenshon found with capital, that the shortcut usually costs more than it saves.

Put it to work

1. If you're considering physical retail, test with something small and reversible — a pop-up — before committing to a lease, to learn whether the in-person experience actually builds affinity for your product.
2. Resist raising more capital than the business currently needs; treat financial constraint as a forcing function for sharper decisions, not an obstacle to route around.
3. Evaluate founder or leadership partnerships on shared values, complementary-plus-overlapping skills, and the ability to resolve conflict — not just cleanly separated job titles.
4. Treat AI assistants as a new discovery surface with its own version of SEO: make product content complete and let real reviews and third-party mentions do the relevance-building work.