[Author’s Note: This story is the great embarrassment of my time at The Standard. I helped to hype this company, and less than three years after the story was published, Sparks.com went completely and dramatically out of business. I’d already left The Standard by that point, but my former editor, Jane Goldman, offered me the chance to write the company’s obituary. It is to my lasting regret that I turned her offer down.]
On a summer evening in 1997, Felicia Lindau, a bright young advertising executive, was driving home from work and realized the next day was her mother’s birthday. She had no time to buy a card, and as she imagined the ordeal of sifting through endless generic cards under a drugstore’s fluorescent lights, Lindau stumbled on what may prove to be e-commerce’s last next-big-idea – selling greeting cards over the Net. The saga of what happened next is a model of how a new Web business comes into being – including the marketing hurdles, the technical nightmares and the tough lessons learned.
Her basic concept was this: Keep customers’ address books on file and remind them of important dates, carry a large selection of cards and guarantee same-day mailing. It doesn’t sound sexy until you consider the implications. The greeting card industry in the United States rakes in more than $7.4 billion in revenue each year. The average greeting card consumer is a repeat customer, a 25- to 54-year-old woman who buys between 30 and 40 cards annually: for birthdays, get-well wishes, anniversaries, promotions, deaths, Mother’s Day.
The average drugstore can only provide consumers with a limited assortment of cards – consistent sellers that appeal to a wide variety of customers. But if you combine the convenience of ordering from the office with a wide range of cards (from dirty, to clean, to downright obscure) and the ability to further personalize the card, well, maybe you have something. If the service works, someone like Sun (SUNW) Microsystems might want to send 10,000 personalized cards – at $3.95 each – to its Java group every year.
Lindau looked for other companies already working on what seemed to her like an exceptionally obvious idea. Surely, Hallmark or American Greetings (AM) must have thought of this. But as it turned out, some companies have made half-hearted attempts at selling cards over the Web, but none were leveraging what Lindau realized was the great potential of the Internet: personalization. This was when the company called Sparks.com was invented. This winter it expects to go live.
The appeal of the service is partially its ambition to carry an unusually wide selection of cards. But the real selling point is the system’s ability to handle complex tasks, like shipping the same card to multiple recipients without requiring you to fill out a different form for each one. A search engine, built from scratch by Phoenix Pop Productions in San Francisco, allows customers to search free-form for themes they’d like to see in a card. Search for “apology” and “affair,” and Sparks.com throws up a card showing a house being tossed into the air by a hurricane, with the message “Things have been a little crazy lately.” Nothing is printed inside, tastefully enough. Can you ask Hallmark for such a thing?
“We expect that 80 percent of our customers will buy 20 percent of our card selection,” Lindau explains. “But when that 1 percent finds the exact type of card they’re looking for, they’ll know why we’re the best option around.”
Sparks.com will keep their customers’ address books for them, and will e-mail them reminders like “Grandma’s birthday on Wednesday.” Eventually, Sparks.com hopes to strike deals with a company like PalmPilot to acquire customers’ digital signatures. The partnership opportunities are boundless: Companies like Virtual Vineyards could bundle gifts and cards, for instance.
Sparks.com was so named for a few reasons. First, as Amazon.com (AMZN ) proved, an online retailer must have the “.com” designation in the name, so anyone who has heard of the company knows its address. Second, Sparks.com is intended to denote a connection, a “spark” between people. Initially, Lindau wanted to name the company Jot.com – symbolic of the convenience and speed of the service – but early focus groups shot the name down. “You’re in the business of assuaging my guilt,” said one member. “Your name should be patting me on the back for being good and thoughtful. Don’t make it seem that by using your service I’m just tossing off a card. You’re making me a better friend.”
Lindau, 30, is an 11-year advertising veteran. At Foote, Cone & Belding she worked with Excite (ATHM ) and then with Amazon.com when each company was just starting up and looking for funding. A prospective Sparks.com partner introduced Lindau to Benchmark Capital’s Kevin Harvey, who introduced her to Benchmark’s Robert Kagle, the company’s primary backer. Kagle sank a little less than $3 million into the project on the basis of his affinity for Lindau. “She’s relatively young, and this is her first general management role, but that profile has worked well for us in the past,” Kagle says. He sees gold not just in Lindau’s original idea, but in the new category of gift portals. “The whole idea of being a gift portal, of capturing that place in the customer’s mind, is a great idea,” he says. “The card could be the forethought to the gift, rather than vice versa.”
In May, Lindau took on a partner, Jason Monberg, whom she met when FCB was working with CKS, an interactive agency. Monberg, 25, had worked as a Java programmer at CKS, part of the young code-building elite who found themselves making extraordinary amounts of money only a few years out of college. At Sparks.com he is the chief technical officer, a man for whom the 100-hour work week seems designed. “If you work at this pace for years, with no end in sight, that’s how you burn out,” Monberg says. “But if you work on one well-defined project after another, you can sustain your energy forever.”
The company’s office is a cavernous warehouse in San Francisco’s Potrero Hill neighborhood. Lindau and her staff occupy a wooden loft, as temporary shelving is erected below them. “We appreciate the dial-up connection, but we really need the T-1,” Monberg deadpans into the phone to Sparks.com’s ISP. “Not having it has been disappointing.”
Across the warehouse, the company has thrown together a passably clean conference room, complete with chairs, a carpet and a large wooden table. “That table was my dad’s,” Lindau says. Her late father was a self-made entrepreneur, who built a construction company from the ground up. “It’s going to bring us good luck,” she says.
The air in the office is like that of an underground resistance group listening to bombs rumbling above them and plotting their next move.
“The VP of engineering at eBay (EBAY) just told me they get 37 million hits a day,” Monberg announces.
“Oh my God,” Lindau says. “Is that more than Amazon?
“I think that’s more than Amazon.”
The online bookseller is Sparks.com’s greatest nemesis and Lindau’s worst nightmare. Pictures of Jeff Bezos are everywhere. Amazon.com’s founder smiles from within picture frames all over Lindau’s desk and from press clippings on a factory-floor bulletin board. In one picture, taken in January of this year, Bezos and Lindau stand together, beaming at the camera. Little did he know what machinations were at work to his right. “In two weeks, they could begin offering this service,” Lindau says to me as we cross the warehouse floor. “But our service will be better.”
Greeting cards are an ideal product for an online store. They are wonderfully uniform, they won’t be returned because they don’t fit, and they mail easily. But they’re not invulnerable. Each night, the company’s inventory is covered in plastic sheets to ward off moisture and dust, and when the cards are selected from the shelves, workers use surgical gloves to avoid leaving fingerprints.
Lindau says Sparks.com is built to handle 10,000 shipments a day. Serving the Web pages and taking orders online is difficult enough. The start-up’s process grows more complicated when someone needs to find a card amid rows of thousands, personalize the message and mail it off – 10,000 times a day. The wrong card with the wrong message to the wrong person, and a customer doesn’t come back.
Shipping tiny parcels to multiple customers is an intricate business. Systems like Dell (DELL )’s or Amazon.com’s are designed for customers who select merchandise and have it shipped to themselves. But Sparks.com’s system separates the customer from the recipient. When Lindau and Monberg first began looking at firms providing e-commerce solutions that might be able to handle the complexities of their concept, they narrowed their list down to two candidates: InterWorld (INTW ), a Manhattan-based company, and Intel (INTC ) and SAP (SAP )’s combined e-commerce arm, Pandesic.
InterWorld promised a tailor-made solution that was designed to be adaptable, and which would allow Sparks.com to own and improve the servers and equipment as they saw fit, but only after an intense, months-long period of studying the company’s business plan and expectations. Pandesic guaranteed a four-month turnaround to get the system up and running, but required that the company essentially hand their business over. Pandesic would run the business off servers belonging to SAP. Also, as Pandesic does with small start-up clients, the company charges a relatively small $25,000 up front, and then takes a percentage of revenue as payment, starting at 6 percent and decreasing along a sliding scale.
They chose InterWorld. Bruce Falck, CEO of Phoenix Pop, the company that is creating Sparks.com’s user interface, says Pandesic’s model wasn’t suitable. “In order to succeed here, you have to own every piece of everything you rely on, so you can make upgrades when you need to. With Pandesic, if you want to add a new feature, you need to call them and run it by their developers, because SAP is a really awkward system. Every new feature must be SAP-compliant.” Peter Wolcott, Pandesic’s president, agrees that his business is about delivering “standard e-process,” but points out it comes at a significantly lower price than more flexible systems.
The final technical lineup includes Phoenix Pop, InterWorld, which provides the transactional software, and Frontier Globalcenter, which hosts the site.
In late September, two weeks before the site would go live for a dress-rehearsal run among friends and family, a content planning meeting convened. Four people – Lindau, Monberg, Chris Lindau (Felicia’s husband) and Lizzie Nichols, VP of sales and marketing, gather in a conference room at the top of a narrow flight of metal stairs. There is an endless supply of good will and affirmation among the staff. Lindau is given to saying, “Agreed. Perfect,” when someone makes a suggestion or comment. Lindau and her husband get into a back-and-forth about the extent of what’s searchable on the site’s database.
“Can I search by the color of the card?” asks Chris, an advertising veteran.
“You can search by the color of the card, the color of the envelope and the paper type,” Lindau replies.
“Can I search by the skin color of the person pictured?”
“You can search by ethnicity, by gender, by the paper’s weight, by everything and anything,” answers Lindau.
Chris sits back in his chair, satisfied.
The group reassures itself that it has all the answers, although everyone knows they won’t be able to account for every contingency. Ask Monberg whether he’s considered all possible outcomes – including bankruptcy, lack of market interest, system failure – and he stares back disquietingly. “Either we’re going to succeed,” he says, “or a lot of other stuff might happen that’s all inconsequential to me right now.”
This is the era when start-up companies must be run like veteran institutions from the day they begin, Monberg lectures by e-mail. “There’s no room for error here.”
At a meeting that convenes later in the month, the conversation turns to whether Sparks.com should sell the names and addresses of their customers to advertisers. Lindau will not budge: Her answer has always been no. “The whole basis of what we do is trust with the consumer. We want people to trust us. There’s no way.”
“What if we sell them when we fold, in a big going-out-of-business sale?” Monberg asks, to laughter. But with a straight face, he answers his own question, ending the meeting. “Naw, we’ll owe all the money to the VCs anyway. Why worry about it?”
The biggest risk, according to Benchmark’s Kagle, isn’t that people won’t want to try the product, but that the company’s untested infrastructure won’t be able to handle the traffic. “Lindau knows how to draw people in to try the system out,” he says, “but if it doesn’t work the first time, we won’t get the repeat customers that will allow us to turn a profit.”
Sparks.com is a small start-up with one big strength: It’s a great idea that’s terribly difficult to implement. “I don’t think Hallmark will be interested in doing what we do,” says Nichols. “They sell a limited selection of their cards online, but have an enormous channel conflict as it is. Wal-Mart sells their cards, Barnes & Noble sells their cards. If they moved that operation online, the conflict would be too much.”
What about big scary Amazon.com? “We’re not in direct competition with them,” says Lizzie. “If we were, that would be a problem. If you look at Amazon’s portal deals, you’ll see they’ve bought every possible piece of relevant real estate on every important portal in the industry.
“So sure, if we were directly competing, they’d be all over us. For some reason, I don’t think that Amazon will see this as a big opportunity. It’s not a question of just retailing, which they’re definitely interested in, it’s all the other stuff – the reminder services, the personalization – that would deter them. But even if they decide we’re a real challenge, they’ll probably just acquire us.”
As October ends and Sparks.com readies a beta version of the site, the pace in the office grows more frenetic. Monberg carries monitors to and fro; the senior producer holds his head in his hands over a coding problem; Lizzie chats amiably on the phone; and a 10-person staff of temporary workers scans in new cards and indexes them. In the meantime, Lindau woos potential investors and infuses the office with her particular, preternatural brand of unflagging energy.
Order fulfillment, transactional systems and the search engine are all tested and as ready as they can be. Market interest, competitive pressures and traffic load are not. Who knows if the market will be there?
Technical considerations aside, the goal of Sparks.com is enormous: to create an automated system that will encompass every type of relationship in Western society. The content reads like a summary of modern human existence: “poetry, etiquette, superstitions and folklore, famous people, cancer benefits, new babies, Valentines, Dilbert.”
Are consumers ready to make a leap of emotional faith and superimpose their best wishes and deepest sympathies onto a card they’ve never actually touched bearing a computer-printed message they didn’t pen themselves? Sparks.com is betting on it.