The Operators: Inside Underoutfit with CEO Daniel Malek
Underoutfit's CEO spent sixteen years investing in companies before crossing over to run one. A conversation about the SKU matrix, operations as the moat, and why the balance sheet decides who wins.
By Maya Bodinger, GTM Lead at Prysmic · July 2026 · 10 min read
Daniel Malek spent sixteen years on the other side of the table, investing in companies through a family office and the biotech hedge fund he co-founded. In 2021 he crossed over to run one: Underoutfit, the New York intimates brand built for real bodies instead of idealized ones. We asked him what scaling a consumer brand actually takes: the SKU matrix no planner can hold in their head, the margin that traces back to owning operations, and an AI ecosystem that now runs across five domains of the business.
This conversation has been lightly edited for clarity.
The unconventional path
Your path to consumer brands isn't conventional: hedge funds, biotech, investing. What made you take the leap?
I spent sixteen years investing in companies, public and private: first as a family office investor, alongside top-tier VCs, private equity and hedge funds, then investing directly in biotechnology companies at the hedge fund I co-founded. Across all of it you learn what companies actually need to do to succeed: capital raising, corporate strategy, business development, M&A. I then launched a corporate advisory firm helping companies with go-to-market, corporate strategy and capital raising. At some point I wanted to put that experience to work operating a company, and own the outcome, not just advise on it. When I met the Underoutfit team and saw their traction, it was the natural next step.
What surprised you most once you were inside? Was there a "this is much harder than people think" moment?
I underestimated how many things you have to be good at simultaneously to scale a DTC company. Launching is easy. Scaling is a different sport: daily performance marketing, constant data analysis, deciding how much to invest in top of funnel versus bottom of funnel, logistics, operations, demand planning, and cohort analysis, which most people outside the industry have never heard of, and which turns out to be the whole ballgame in deciding where a dollar goes. Then AI arrived on top of all of it and changed how marketing itself is conducted. But that became the opportunity: implement AI properly and you can leapfrog the competition, which is exactly what we've done at Underoutfit. The flip side is what makes it worth it.
It has never been easier to launch a brand and never been harder to scale one. The ability to scale is the moat.
And the category surprised me too. It looks crowded, but there is real whitespace: the large legacy brands no longer connect with what the customer actually wants, in values or in product. They built for an idealized customer; we build for the real one, comfort and fit for actual bodies. That disconnect is our opening, and it's wide.
What advantage did being an outsider give you?
I got lucky on one thing: my co-founders are twenty-year veterans of the disciplines that make an intimate apparel brand work, marketing, design, logistics, so I got world-class insight into the craft very quickly. My experience taught me not to try to do their jobs. The outsider's advantage was knowing my role: build the structure that lets those experts thrive as the brand scales, rather than second-guessing them on product or creative. What I brought on top is financial rigor: a clear view of the KPIs the company has to deliver in order to scale, how to balance growth with profitability, and a working obsession with the balance sheet and the cash conversion cycle, which most brands ignore until it hurts. And frankly, an investing career trains you for one thing above all: recognizing the two or three decisions that will actually set the company's trajectory, and making them.
Building the business
Give us one number that captures how dramatically the operation has evolved.
Here's the number I watch: we've increased our average order value by 20% while keeping the cost of acquiring a new customer stable. That relationship, what you pay for a new customer versus what her first basket brings in, and how the repurchase rate turns that into lifetime value, is the entire economics of the business in one line. We are a second-purchase company. In our category, once a woman tries the product, she comes back and shops more. So the whole operation is built around a single obsession: lower the overall cost of acquiring her, then bring her back as fast as possible with the largest possible basket. Everything else exists to serve that loop: planning, retention, inventory.
What does scaling a consumer brand look like today that outsiders underestimate?
The SKU matrix. Intimates is a brutal category: one style is thirty-plus size-and-color combinations, each with its own demand curve. Multiply that across a catalog and no human planner can hold it in their head. Growth doesn't just add revenue, it compounds complexity, and it eats working capital, because every incremental dollar of sales requires inventory bought months earlier.
People see the ads. They don't see that scaling is mostly a forecasting and financing problem.
When did you realize operations had become a competitive advantage, not a cost center?
From the numbers, honestly. We run gross margins above 65% with relatively low returns for the category, and when you trace why, the answer is operations. We never delegated the process. We own it. Our team ran operations at Delta Galil and other large brands, has worked together for decades, and we keep our own QA team in Asia sitting close to the factories. That's not a cost center; that's why the margin exists. The second realization came with inventory.
SKU-level inventory management is where growth quietly dies: a stockout on a high seller is growth you already paid for and didn't collect. That's why we teamed up with Prysmic on inventory management and demand planning: keep the capital in the right SKUs and never miss a sale on a best seller.
When both your margin and your growth trace back to operational decisions, ops isn't supporting the business. It is the business.
Hard-won lessons
What operational challenge kept you awake at night once revenue started scaling?
Inventory: being long the wrong stock and short the right stock at the same time. Both cost you: one ties up cash, the other forfeits revenue you already paid to generate.
When your best seller goes out of stock, you keep paying for traffic that lands on a "sold out" page. That's the nightmare scenario, and it's invisible in most dashboards.
What's one supply chain mistake customers never see but that costs millions?
Buying to a size curve built on last year's data. Get the curve a few points wrong across a big PO and you've simultaneously created a stockout in the sizes that sell and a markdown problem in the sizes that don't. The customer never sees a "mistake." She just sees her size unavailable and buys elsewhere. Also in the never-seen category: carrier invoices. Audit them line by line and you find you've been quietly overbilled for years. We automated that audit; the recovered money goes straight to margin.
AI: what's actually working
Where has AI moved from nice-to-have to something you couldn't operate without, and where did it fail to deliver?
At this point it's easier to list where it hasn't. We run a coordinated AI ecosystem across five domains: paid advertising, e-commerce content, email, customer support, and operations. The word that matters is system, not tools: each workflow feeds the next, from strategy to creative to deployment to reporting. Marketing is the deepest. An AI creative-analysis system scanned two years of our Meta creatives, it effectively watches every video, and it found things no human noticed: adding product-sizing information to the visual text doubled ROAS on static ads, and switching aspect ratio from 4:5 to 1:1 lifted spend from a younger segment by 400% with ROAS up 50%. Moving creative from intuition to data raised brand profitability by at least 25%. Today our paid media runs on AI workflows that audit every live campaign against awareness-stage and persona frameworks, find the coverage gaps, write replacement copy, and push it live through the API. When we post a Reel on Instagram, it becomes a Meta ad with zero manual steps. Our product-page content is AI-generated inside a locked design system, email campaigns are split-tested at a scale no human team could produce, and 70% of customer-support tickets are now resolved autonomously, with an AI voice system in beta. On the operations side: demand planning and inventory with Prysmic, where inventory, incoming shipments and sales are connected in one brain and replenishment is managed end to end, plus fully automated invoice and carrier auditing that finds money humans never would.
Where it failed: anywhere we pointed AI at a messy process and hoped.
AI doesn't fix bad process or dirty data. It accelerates whatever you feed it. Fix the process first, then automate it.
And returns and reverse logistics are still stubbornly manual. Nobody has cracked that yet.
If you were starting Underoutfit today with AI from day one, what would you build differently?
Honestly? Nothing. What makes Underoutfit special is that we are extremely proactive about how technology evolves. We are early adopters of anything that can meaningfully improve the business, and we adopted AI as it matured rather than retrofitting it years later. So the company I would build today with AI from day one is essentially the company we already built. The real lesson isn't a different blueprint, it's the posture: don't wait for your industry to validate a technology before you test it. The compounding advantage doesn't come from any single tool. It comes from being the organization that adopts what's next six months before the competition, because that muscle pays off again with every wave that follows.
AI and the organization
Has AI changed how you hire? Which roles shrink, and where do humans become more valuable?
It's changed the shape of hiring more than the volume. Coordination and reporting roles, people whose job was assembling spreadsheets and chasing status, we simply don't create anymore. What I hire for now is judgment: a planner who can challenge the model's forecast, an ops lead who knows when the system is wrong. Humans get more valuable exactly where AI stops: taste, product, negotiation, and relationships with factories, carriers, and partners. Nobody's ever gotten better freight terms because an agent asked nicely.
What decision do you insist on making yourself, and what would you happily hand to AI tomorrow?
Brand values, the customer journey, and product quality: those stay with me. We are obsessed with our customer: who she is, how she experiences us from the first ad to the first fitting to the tenth reorder, and whether the product deserves her trust. Those decisions are the company, and no model owns them. Everything else I will happily delegate, and largely have. If a decision can be expressed as a reconciliation, an audit, or an optimization, a machine should be making it, and at Underoutfit one already is.
The view forward
Two brands, same product, similar scale: what determines who wins? And what metric do you obsess over?
Whoever converts a dollar of inventory into a dollar of collected cash fastest, at the highest realized margin, wins, because they can reinvest faster than the competitor and compound.
Product and brand get you to the table; the balance sheet decides who's still at the table in five years.
The metric I obsess over is realized margin after returns: not the gross margin in the deck, but what's actually left after the customer has kept the product. In intimates, returns are the silent killer, and most of this industry manages to a margin number that isn't real.