The Billion-Dollar E-commerce Company You Know Nothing About

Zulily has defied the conventional wisdom—marketing to moms, sticking with flash sales, evading Amazon. Can it defy the doubting investors who think it can’t last?

The Billion-Dollar E-commerce Company You Know Nothing About

On a foggy January morning at Zulily headquarters in Seattle, a menacing little girl wearing a tiara is sitting in a little pink wagon in one of the hallways. She’s throwing a tantrum, brandishing her sippy cup like a cudgel. Our eyes lock on my way back to the waiting area from the restroom, and suddenly I’m sucked into an impromptu staring contest, which, considering the home-field advantage of the competition, I feel I’m about to lose. I avert my eyes to the paper lanterns that drip from the lofted ceilings and the carnival-sized stuffed animals, rocking horses, and playhouses that line the walls. It’s like the world’s nicest Crate & Barrel swallowed a pediatrician’s office.


It’s a little after 10 a.m., and my toddler nemesis is just one of dozens of children scattered throughout the labyrinthian premises. On the other side of a glass wall in the waiting area are at least 10 photo shoots, flickering on-and-off in chorus while busy-looking twentysomethings seemingly transposed from an Urban Outfitters catalog zip around like pinballs. Two more tiny humans dressed in bright scarves and furry boots barge through the entrance, screaming at the top of their lungs as they make a beeline for the toy wall. Bob Marley & the Wailers’ “Exodus” softly knifes through the noise from an invisible speaker system. Sitting across from me, a blonde woman in Tory Burch flats and a black pantsuit stares unblinkingly into her iPhone, waiting for her appointment, unfazed by the bedlam.

Darrell Cavens

This is Zulily’s “fifth or sixth office space” that the company has outgrown in its five years of life, according to CEO Darrell Cavens, as if office space were teensy Velcro sneakers to be thrown into the donation pile and forgotten. If you aren’t a habitué of the mommy blog circuit, you would be forgiven if you did not recognize the company by name. But when Zulily announces its fourth-quarter results and 2014 revenue on February 11, it’ll solidify its place in the still exclusive club of e-commerce companies with more than $1 billion in annual revenue. In fact, in the universe of all retail brands, only Amazon and Old Navy hit the billion-dollar revenue mark in a shorter amount of time.

Entering this rarified club would be impressive under almost any circumstances but it’s even more so when one considers the particulars of Zulily’s business. The company sells just about anything a young mother might want for her family (and herself), at sometimes steep discounts. Zulily markets its wares via the limited-time merchandising technique known as flash sales, a subspecies of e-commerce largely discredited in the wake of Gilt and Fab’s well-documented struggles. There are no stars or any product-rating system on the site, since products disappear after three days. It carries no inventory, which means that it can take weeks for it to fulfill orders, which is the polar opposite of the rest of the industry’s push toward on-demand near-instant gratification. Returns are nonexistent or extremely difficult to procure. And it operates in the literal shadow of Amazon and the metaphorical one of New York and Silicon Valley-based e-commerce startups that have commanded far more hype.


Zulily went public in November 2013, and after a surge that saw it skyrocket to a $7 billion market cap in February 2014, skeptics have driven down its stock price to its lowest point as a public-traded company. It’s easy to diminish Zulily—it does everything “wrong,” Amazon will crush it eventually—but it’s also an interesting case in the widening gulf between Wall Street and technology companies. Zulily’s business looks simple and that makes it easy to underappreciate or misunderstand how much goes into making it work. Its CEO and cofounder aren’t the types to yammer away on CNBC or do much courting of public sentiment at all.

Inside Zulily HQ

Yet it is Zulily’s refusal to adhere to the beaten path that is why it’s built the business it has and why I’m here amid the Chuck E. Cheese-style chaos to deconstruct how it has quietly been able to forge an online juggernaut with an insanely loyal customer base. A whopping 80% of its orders come from returning customers. Why? Is it merely a validation of the volatile private-sale model that caused other competitors like Fab to crash and burn? Is it a testament to the buying power of moms? Yes, and yes, but it’s more than that too. Zulily has grown its business incrementally, albeit at light speed, solving thousands of tiny daily inefficiencies along the way. It’s hired well, spent efficiently, shunned the hype game, and quietly invested in a flexible and powerful technological infrastructure that emboldens employees to move fast, break stuff, learn from their mistakes, and do just a little bit better next time.

Oh, and unlike the Amazons of the world, there is no need to go all Jeff Bezos and scream at anyone, like, ever. That’s what the ugly rubber pig is for.


Putting Out The Mommy Gazette

At 6 a.m. Pacific time every morning, 365 days a year, Zulily posts about 100 new events with 9,000 new products to its Pinterest-like front page as well as its mobile app. For the next 15 minutes or so, if all goes as planned, it then sends out millions of emails to customers with its new sales. “We’re sort of like a newspaper in that we publish something brand-new every single day,” says Cavens.

It’s not just a metaphor: Zulily is a truly massive publishing apparatus, and it’s the first step in understanding how the company hums with efficiency. From day one, Zulily has shied away from vendor-supplied stock photography, opting instead to shoot 99% of its products in-house. In Cavens’s view, it makes the website, app, and emails feel cohesive like a catalog. This stickiness appears to do its job. Although Zulily executives decline to discuss the particulars of their traffic, according to third-party service Alexa, Zulily receives about 700,000 visits a day, and shoppers spend an average of nine minutes and 50 seconds looking at the site. “One of the things is I want to make it feel like we’re a retailer, and not just [a way of] seeing other people’s products,” says Cavens. “We want to present it in a way that’s true to us.”

The process begins on the second floor of Zulily’s five-story building. There are 14 different photo sets, some as large as 20’ x 20’, shooting hundreds of images a day. Each set features elaborate lighting systems and a medley of interchangeable backdrops. Open closets are stacked full of props—from shaggy rugs to bowls of plastic fruit to bodysuits designed to make women look pregnant. These sets were designed and built in the basement of Zulily HQ by metalworkers, who will jury-rig whatever kind of set stand or camera mount the creative team needs. Meanwhile, small teams of makeup artists, stylists, photographers, and other creatives scramble around like worker ants, shooting up to 12 models a day. These subjects are usually Zulily employees from other departments (buyers, copywriters, etc.), who rotate in regularly and have their measurements, hair color, and style preferences saved into a database. “I’ve done it twice now,” a twentysomething brunette from the PR team would later tell me. “I try to schedule my shoots for Friday afternoons,” says another Zulily-ite, who lives in the trendy Ballard neighborhood just north of Queen Anne, “so I can get my makeup done and go out right after.”


And the babies! Where do all the babies come from? Stage moms and dads from the local community vie to have their children featured as models through a submission form on The young’uns come in, pose under bright lights, and often cry. They’re paid in free snacks. (Hence, my sippy-cup-wielding opponent from earlier.) Parents get a thank-you email from a Zulily community outreach manager, along with a zip file of professional-quality images. Prospective modeling career: Launched. Pennies for Zulily: Pinched.

The pure product shots happen on floor three, another photo-set village working in tandem with their second-floor model crew to shoot standalone products such as watches, perfumes, espadrilles, and other small items against white seamless backgrounds. I pull a knuckle-tattooed staff photographer to the side:

Me: “How far in advance are you shooting images for?”


Photographer: “Maybe two, three days usually?”

Me: “That’s it?”

Photographer: “Sometimes it’s a week ahead. But that’s during the slow season, like right now, after the holidays.”


Once a shoot is complete, a team rushes in to change the set, and it’s on to the next one. No move is wasted. Little momentum is lost. The employees have agency to conceptualize, style, and otherwise carry out a shoot as they see fit. The pace is not for everyone, but everyone who is there knows why they’re there.

While that’s happening, the photographer selects a few of the best shots, crops them, and sends them to the art department two floors up via ZuStudio, Zulily’s bespoke content management system (CMS) where the images are further retouched before they are published.

But that’s merely phase one.


The Love Song Of Darrell And Mark

I’m sitting inside Cavens’s cramped office on the dimly lit ground floor of Zulily offices. Up on the fourth and fifth floors, where the buyers, copywriters, designers, and marketers sit, everyone enjoys a pristine view of Puget Sound. Cavens’s office window provides a scintillating grounds-eye view of, well, a noisy street. A treadmill desk sits in the corner collecting dust, and other than photos of his young son and daughter in ski gear, the space is sparsely decorated. If not for Cavens himself, a well-padded, goofy-laughed, even-cadenced 42-year-old Canadian who always seems to have a big grin on his face, it’s actually kind of sad. “I was on a call earlier, and every time a big semi-truck came by, I was like, ‘Hold on a second,’” he says with a chuckle. “I guess I didn’t think through that part of it.”

It’s one of the few company considerations Cavens doesn’t seem to have thought through. “There’s nothing about him that’s corner office,” says Luke Friang, Zulily’s CIO and head of technology, and a veteran of the retailers, Costco, and Eddie Bauer. “He’s a big reason why I’m here.”

Cavens landed in Seattle in 1996 in his early 20s, after dropping out of the University of Victoria just outside of Vancouver. He landed a job working for Microsoft cofounder Paul Allen at Starwave as part of its Advanced Technology Team, building out the backbone of such websites as ABC News,, and, then the second largest site after Yahoo. He quickly developed a specialty in scaling web properties.


In 1999, at the height of the first dotcom boom, a recruiter poached him to go to work for a nascent online jewelry startup called Blue Nile, which helps clueless dudes find affordable, no-nonsense engagement rings for their better halves. (Cavens got engaged shortly after joining the company.) It was there Cavens linked up with Mark Vadon, Blue Nile’s founder and CEO. “When I first hired Darrell, I think a lot of people considered him one of the best software developers in the city,” says Vadon, who speaks with a laid-back, cool-history-teacher vibe.

Cavens quickly ascended Blue Nile’s ranks, eventually becoming its CTO and SVP of marketing. Blue Nile quietly rode out the downturn that vaporized innumerable e-commerce startups and by 2006, it surpassed Tiffany & Co. by selling $197 million in engagement rings and wedding bands, versus the $186 million in little blue boxes moved by the latter.

To say Cavens and Vadon became close might be underselling it. Every Zulily employee I spoke with characterized the company cofounders as unlikely stepbrothers who just so happened to run a billion-dollar business together. Their kinship is particularly evident during company-wide meetings (which don’t stretch longer than 30 minutes, typically) where the two chew the fat onstage. “When you get those two together, there’s a really nice harmony to it,” says Lori Twomey, Zulily’s chief merchant. “It’s almost as if they’re playing a song.”

Cavens and Mark Vadon

“I love him,” laughs Vadon, 44, when I ask him about his collaborator. “He’s my best friend. We’ll fight. We’ll complain to each other. At this point it’s like we’re an old married couple.” The pair aren’t so much yin and yang as they are a well-shaded Venn diagram of complementary skills, temperaments, and online retail geekery. Cavens is the smiley, bald nerd with a fierce eye for details and systems; Vadon is the hippie-ish visionary decked out in dad jeans and a plain, forest-green short sleeve, Seattle’s quietest billionaire, who daydreams in business models, and just so happens to have every Seattle-based VC on speed dial. (“Zulily was easy because I knew a lot of VCs already,” he says. “I could kind of pick who I want to get money from.”)

In 2008, Cavens left Blue Nile for Microsoft, where his wife currently works, to direct the team in charge of BizTalk Server. Meanwhile Vadon “retired” from the jewelry business. Both knew they wanted to work together again. “I think as long as we’re working, we’ll be working together in some way,” Vadon says. It wasn’t long before the two began spitballing ideas for a new kind of online retailer. The only question, naturally, was what.

They had a small inkling of what they wanted to do. At the time, flash-sale startups like Gilt were just beginning to make some noise, and Cavens and Vadon seriously considered aping the model for the home and beauty space (a la One Kings Lane) before scrapping the idea. “What we came to realize in the health and beauty space, there’s a lot of vendor concentration,” Cavens says. “Many of the top 50 brands are owned by three large companies. If you don’t have the supply there, it’s hard to go after.” They ruled out fashion as well because they deemed it too unwieldy. “What we felt like there is you couldn’t control the supplier dynamics if you’re going after high fashion,” he says, especially “if you were trying to get new freshness every day.”


As they tell it, they decided to focus on boutique products for young moms shortly after Vadon and his wife, who at the time was five months pregnant, made their first trip to Babies “R” Us. Overwhelmed by the mountain of crap that young parents never knew they needed, they made one loop through the store and headed for the exit to get lunch.

The experience would prove to not so much be a moment of clarity as a conversation starter. Vadon brought the ovum of his mom-driven business to Cavens, and they soon realized that the total addressable market for new mothers was both underserved and enormous: Some 4.5 million kids are born in the U.S. every year, and the only discount retailers in the space were, like, T.J. Maxx. If they could subvert a legacy diamond seller like Tiffany & Co., they could do something here. In mid-2009 they chose the name “Zulily” with the help of a branding agency because it was easy to say and just as important, it wouldn’t limit what they could sell. (Some of the too-cute runner’s-up included: ItsyBtsy, Tumble Up, Tip Toe, Katroo, Toodle, and Pitter-Patter.)

And so, two new dads began building an online retail store for new mothers.


Must Be The Shoes

In business terms alone, Blue Nile and what they envisioned for Zulily couldn’t be more different: The former sold high-price-point goods to confused men, once, maybe twice, in their lifetimes. The latter aimed to sell inexpensive, disposable goods to women, as long as they have an itch to hit the buy button. As a starting point, they established a few core principles to guide their decisionmaking:

1. They needed a reliable cost structure from the beginning;
2. Cavens would be the CEO, while Vadon would step back from the day-to-day to focus on big-picture stuff;
3. They needed a business model that would drastically differentiate Zulily from Amazon.

And number three really could have been numbers one, two, and three. In their purview, online retail is “this game where Amazon set up the rules, and Amazon wins,” Vadon explains. “I think in general when you’re starting a company and it’s in the retail space, you should ask yourself: ‘Does Amazon already do this well?’ And if they do, move on.” Cavens adds, “Amazon has traditionally competed in the e-commerce space by going after core products that companies sell and focusing on undercutting them on price,” he says. Think of the way Amazon targeted by selling Pampers at a deep loss until its parent company sold itself to Amazon. “We are in a unique position of changing what we sell every day, sourcing unique product that people might not otherwise find and having our customers show up each day without purchase intent—without looking for a particular item.” This maneuverability, he feels, is what will help keep Amazon from competing with Zulily directly.


Originally Cavens and Vadon wanted the selling focus to be on things like strollers, car seats, and other hard goods. Dad gear. It would have proven to be a mistake, one that was corrected by a key hire: In November 2009, they convinced Twomey, a veteran of Eddie Bauer and the catalog space, to head up merchandising. Twomey wouldn’t just prove to be Zulily’s vital tastemaker from a product standpoint, but she unlocked the secret to improving customer frequency.

Twomey knew that in order to build a long-term, sustainable business predicated on the private sale model, you had to choose products that tired parents needed but didn’t want to necessarily drop a sizable chunk of their paychecks on. And diapers had already been done. Shoes—and basics such as T-shirts, tights, and onesies that kids wear out fast—ended up being her secret lever to drive sales. “You had to think about your income, about what you’re buying and what’s disposable,” says Twomey, a voracious marathoner who rises at 4:30 a.m. every morning for an hour-long run with friends. “A stroller is something you might buy once, maybe twice throughout the life cycle of your child. But I’m going to buy new clothes every single day. And then the kids are going to grow out of them.”

Zulily went live on January 27, 2010 with an initial focus on children’s apparel. Thanks to word-of-mouth marketing, fueled in part by its embrace by mommy blogs—“Zulily is amazing!” was a common refrain on sites such as Bloggin Momma Drama—Zulily would be operating as a cash-flow-positive business by the fourth quarter of 2010, which validated Cavens and Vadon’s business model.

In December 2010, Twomey orchestrated Zulily’s first women’s sales event, adding something for mom to the mix of kids’ gear. And with that, the rocket ship reached orbit. Now a tired, overextended mom could have a daily shopping adventure of need and impulse buys: Jumpers, strollers, scented candles, Kind bars, Jessica Alba’s organic body wash, ballet flats, maybe this sassy graphic tee for hubby, and, oh! Would you look at that? These black leather Kenneth Cole ankle boots are 50% off, too. Click. Add to cart. Even better, in exchange for the exposure to such a targeted, motivated audience, Zulily pushed its merchants to discount its wares; most tend to hover in the 50% range. The excitement for this type of shopping overcame any of the annoyances that came with it. Because Zulily doesn’t hold inventory, once a flash sale would end, it would then send a purchase order to the third-party seller, who would then ship boxes of product to a Zulily fulfillment center. Within 24 hours, those items are repackaged and sent back out to the waiting customer, which is why it could take weeks for customers to receive their orders. And unless something was defective, all sales are final.

No matter. “These moms got really fanatical,” Vadon says. “Women’s apparel is the key piece of the business. They love this way of shopping. They end up buying all sorts of stuff.” Zulily’s 2011 revenue was $142.5 million, a 674% increase over 2010.

Zulily’s rapid rise barely registered a blip in the tech and business press, which turned out to be one of Cavens and Vadon’s masterstrokes. “When you’re an early-stage startup, and things are working really well, the dumbest thing is to try and get a lot of PR,” says Vadon. “If you’re digging an oil well and you hit a gusher, you want to be really quiet about it. You don’t want five other people digging right next to you. With this, we hit a gusher.” Both attribute their press shyness to being introverts by nature, though some of their employees dispute that claim.

So it wasn’t until their IPO in November 2013 that anyone noticed just what Cavens and Vadon had built. “If you look at the time the IPO, we had raised on a net basis roughly $92 million,” Cavens tells me. “We had $130 million in the bank when we went public. So we weren’t consuming capital to build a business.” When Zulily went public, it had 2.6 million active customers, 45% of whom were coming in through their phones; and $331 million in revenue, up 132% from 2012. Shares soared almost 90% on its first day.

But Fruit Flies Never Bought Any Squeezies

For all of Zulily’s deft moves as a merchandiser, they overshadow the fact that at its core the company is, in fact, a technology business. And that starts with Cavens. “He’s got a really big, hardcore technology background, which is one of the great things about Zulily,” says Vadon. “When you’re a product-driven company run by technologists, the developers respect him. He thinks about everything from the technology perspective.”

Remember all those images created by those two floors’ worth of makeshift portrait studios? Once those images hit the site, Cavens’s technological vision begins to bloom. One of Zulily’s core competencies is its content-management system, ZuStudio, which is designed to aggressively A/B test—or rather, A/B/C/D/E/F/G . . . test—everything you click, from the types of products that show up at the top of your website, to the kinds of photos that you and your family might identify with, to which photo angles might perform better than others.

Zulily’s audience might not realize it—in fact, most don’t—but every consumer-facing experience, whether it’s on, on the mobile app, or through the daily email, is personally tailored down to a granular level, in real time. For example, if you have an 8-year-old boy, and you sometimes buy him clothes, Zulily’s algorithms won’t show you dresses for 3-year-old girls. What you see on your screen is engineered to be as close a proxy to you in real life as possible. This kind of one-to-one personalization is constantly being refined: Based on the way you browse, the software can infer the age and gender of your children; what kind of pets you have; what size you wear; and even whether or not you shy away from gluten. Like Amazon, the site is designed to feel intensely personal for millions of people. “We’ll have focus groups, and moms will say, “I’m your perfect customer, because everything you launch is exactly what we like,’” says Vadon. “Well, yeah.”

These multivariate gains aren’t huge: Typically the change hovers between 0.2% to 0.4%. Yet those fractions of percentages add up, particularly when you consider the sheer amount of content Zulily publishes every day. It is fluidly adaptive. Like snowflakes, it’s never the same for any two customers.

Running millions of these tests every day is, in Zulily’s estimation, one of its core advantages in online retail, helping its buyers attenuate to the fickle needs and wants of moms as their children grow up. Children’s clothes, for example, are no longer Zulily’s biggest-selling category; now it’s womenswear. It has been an incremental evolution that would not have happened thanks to Zulily’s enormous trove of buyer data. “The other week I was talking to a scientist about their research cycles and why they use flies for testing,” says Cavens. “It’s because they multiply so fast! The life cycle is so short, that they can do a lot of testing.”

He applies the same philosophy to Zulily. “We get to launch a new e-commerce website every day,” he says. “Right now we’ve had a thousand different versions of the site that we’ve been able to launch. Imagine somebody who does that once a quarter. They get four times a year to figure out if they did it right and adjust! We get 365 days a year to do that testing. So we get to learn in a way. We get to do that sort of life-cycle research kind of fast.”

Flash Is Fast, But Can It Be Cool?

Zulily has to operate fast, because what customers—and now investors and analysts—expect from e-commerce keeps changing. For customers, the number-one complaint leveled at Zulily is its excruciatingly slow shipping times. Customers weaned on free two-day Amazon Prime bemoan having to wait two to three weeks (and sometimes longer) for their orders, which isn’t free, either. Zulily’s cofounders admit its languid shipping times can be frustrating. But, they argue, it is a deliberate concession necessary to keep prices low and the product lineup diverse. Cavens and Vadon claim it is something that customers in focus groups, when forced to choose between lower prices or faster shipping times, opt for the good deal.

Customers still hate it. “Shipping takes forrrrreeeevvvvvveeer,” my friend Brooke Tawanna exclaims when I ask her via email about shopping for her 2-year-old on Zulily. “I’ve ordered clothes for Quinn that no longer fit her by the time they arrived.” Merchants, too, can have their doubts. “With us holding inventory in Pakistan, by the time we get the goods to Zulily, sometimes it can be two to three weeks [before a customer gets their order],” says Adnan K. Mehmood, whose kids clothing company, Jelly the Pug, went from selling a few hundred units in relative obscurity to a $4.5 million business since partnering with Zulily in 2011. “With that propensity, customers could go somewhere else.”

Skeptical analysts and investors have battered Zulily’s stock down from $73.50 to $18.72 as of February 5, 2015, below its $20 IPO price. They are wary that new customer growth is slowing, which will require more marketing. Cavens has acknowledged a problem with Zulily’s in-house email marketing platform, and new costs—such as Zulily’s first TV spots—did nothing to assuage any concerns. Furthermore, some analysts are wary of Zulily’s hyper growth. “There’s something fishy about their numbers,” says one analyst from a top research firm who spoke candidly but asked to not be named after our phone conversation. “Their top-line growth seems a little too aggressive in a world where no one else is growing. Most of the time, if it looks too good to be true, it’s not.” Not everyone shares this view. Zulily’s stock saw a spike in January when the investment banking firm Stifel called the company “an appealing long-term asset.”

Although Zulily is the only e-commerce company with a flash-sales model to grow this large and have an IPO, I wonder whether Zulily is paying a bit for the sins of its peers. I had to ask Cavens and Vadon: Where did other flash-sale sites go wrong? Unsurprisingly, they have a few thoughts on the matter. “When you look at Fab or some of the other things that got started and flamed out, I think there’s a lack of cost control,” Vadon offers. He seems to relish the role of the crotchety retail veteran (he also sits on the board of Home Depot—”for fun”), shooting a smile at Zulily’s now-nervous communications chief. “We’ve always been really cost conscious. We realize that if you save a nickel, you can pass that savings onto consumers, which generates more demand.” Cavens says he attributes much of Zulily’s success so far to “obsessive discipline,” such as nailing the unit cost economics on fulfillment and marketing early on.

Vadon can’t help but mention an old Fab video on YouTube, where cofounder Bradford Shellhammer gives a TechCrunch reporter a tour of the company’s headquarters. “They’re showing the scratch-and-sniff wallpaper,” Vadon says, incredulous at the thought. “And I’m like, you’re a two-year old company. You’re basically running through venture money to buy scratch-and-sniff wallpaper. Is that smart?”

This Little Piggy

During our talk, Vadon mentions a book by Harvard business professor Youngme Moon called Different: Escaping the Competitive Herd. He was so moved after reading it that he asked Moon to fly out to Seattle for a meeting. Soon after, he and Cavens asked her to join Zulily’s board. (She did.)

In the book, Moon argues that after spending time and effort on consumer research, many companies focus their effort on fixing the things that consumers complain about, instead of what they are good at. This, in Vadon’s worldview, is a grave mistake. “True outliers in the auto industry, like Ferrari, aren’t worried about how the baby seat fits in the car,” he says. “Or you’re Audi, and customers say, ‘I love the excitement but I wish you were safer.’ Then what happens is everything starts to homogenize and become the same.”

Talk to Zulily execs and longtime employees and before long they’ll tell you the rubber pig story to illuminate how this philosophy plays out at the company. About three months after the site went live in 2010, when you could count all of Zulily’s employees on one hand, the website suffered a near-fatal error. An assistant buyer uploaded something that caused every product on the site to free-fall to the very attractive price of $1. Within minutes, everything sold out. The first developer freaked out. Cavens, who was doing a significant chunk of the coding himself at that point, pulled him into a conference room, told him it was going to be all right, and together, they spent the next three hours piecing the website back together, while the other employees called up customers, telling them there had been a technical glitch. To Cavens, it was business as usual.

Now, whenever the website goes down, the engineer or whoever was responsible for the breakage has to display this ugly rubber pig on their desk. It’s pink, and a bit larger than a football, with a furrowed brow and traces of dirt in its crevices. Touching it would seem to require hand sanitizer. It looks angry, and worn, having been with the company for four offices now. This pig has seen things.

The ugly rubber pig

Displaying the pig is not meant to shame them, per se, but rather to remind Zulily-ians that it’s okay to try things as long as you learn from it. “I don’t know who picked the mascot out,” Cavens says. “It was an element of trying to put a lighthearted moment on something that’s oftentimes taken way too seriously. What I didn’t want to have happen is to have somebody break something and then them slow down to a crawl, and not release something quickly—to not break something again.”

Zulily is aware of its challenges, but the company also has a very clear vision of what its strengths are. Investing in infrastructure—whether it’s building the sets in-house for creative; or entrusting its photographers and designers to produce dozens of magazine-style spreads a day; or giving its engineers the freedom to break things; or building an expensive CMS that can automatically compound fraction-of-a-percentage clickiness to boost sales; or going out of its way with customer service—is what has allowed Zulily to scale to its size without buckling. Each stratified layer helps brace the chaotic one above it. It’s about moving fast. Ironically, they’re not worried about how the baby seat fits in the car.

Whether Zulily is a true outlier or a flash in the pan remains to be seen, but its executives seem confident that they’re in it for the long run. “It’s not doing one big thing to improve, so much as a hundred little things better,” Cavens says. “It’s like the photo studios downstairs, where that’s iteration 10 or 12. They made it a little better, and a little better, and a little better.”

It’s nearly impossible to not read the building’s five floors as a physical manifestation of Zulily’s structural integrity. Cavens, in his noisy bottom-floor office, with the technology team just within earshot, provides the cornerstone that keeps the cathedral from crumbling. They’re there in the dark, making the business a little better, and a little better, and a little better, every single morning. Anything outgrown gets thrown away.

Corrections: An earlier version of this article misstated the number of products Zulily publishes every day, which is 9,000. Zulily also had a $1 billion run rate as of Q3 2014, not Q4. We regret the errors.


About the author

Chris is a staff writer at Fast Company, where he covers business and tech. He has also written for The Week, TIME, Men's Journal, The Atlantic, and more