Behind Kozmo.com's Fall: A Direct Marketer's Nightmare
"We've got to get our average order sizes up," says one executive during a meeting shortly after the dot-com bubble burst in April 2000.
In summer 2000, just after the board ousted co-founder and CEO Joseph Park, newly placed executives attempted to change Kozmo's cash-guzzling ways and pay serious attention to profit-oriented metrics, according to sources.
And while overly aggressive expansion plans and a backbreaking $150 million deal for a store presence with Starbucks are well-documented reasons for the company's demise, behind the scenes, direct marketing-related troubles continuously stymied the company's attempt at a turnaround, the sources said.
For example, near the end of Kozmo's run, according to one source, executives calculated that if they had 25,000 users per hub -- there were 18 hubs in 11 markets -- Kozmo would be profitable. But they soon found out that they had the 25,000 users, but still weren't turning a profit. The problem? They defined a user as anyone who had ever registered at Kozmo.com.
To try to get a realistic handle on the situation, the new senior vice president of marketing, Andrew Resnick, drew on his operations background from Blockbuster Video.
He defined "active members" as people who had used the service in 30 days.
To boost that number, Kozmo's previously mainly inbound 50-telerep call center began calling people who hadn't used the service in 30 days, offering them $5 worth of so-called Kozmo Points to use the service again.
"That worked phenomenally well," Resnick said, estimating the offer garnered a 30 percent redemption rate.
Resnick also estimated Kozmo drove its customer acquisition costs down to about $50 each by fall 2000 from $175 in July 1999.
He also slashed Kozmo's marketing budget from $40 million to less than $10 million.
"We went from burning $20 million a month to $1 million a month," Resnick said.
But Kozmo's homegrown technology created another direct marketing-related barrier. The company was unable to test promotions other than its standard free pint of ice cream to first-time buyers without manually circumventing its software. As a result, Kozmo was unable to bundle its products into special-event packs or try many of the gimmicks common to other direct marketing companies.
"They couldn't do promotions, which as you know is a significant handicap in direct marketing," said Cliff Sloan, president of The Sloan Group, the New York ad agency that took over the Kozmo account in August 2000.
What's more, Kozmo Points could only be used for rental products.
"So when we tried to become less dependent on rental products, we couldn't even incent you to come in," Resnick said. Even the $5 reactivation campaign that had worked so well could only work on rental products.
Kozmo's technology also was reportedly incapable of database marketing. "We had no ability to do that, or we absolutely would have," Resnick said.
Kozmo also reportedly did little to exploit its customer e-mail list.
"They were terrified of e-mailing their customers," said Ludmila Palasin, vice president of marketing at The Sloan Group. "They didn't want to be labeled spammers."
Kozmo sent one local e-mail per month to customers in its 11 markets, and one national one. What's more, they hadn't experimented with anything beyond text e-mails until a rogue manager tested HTML in the spring of 2001, she said.
The clarity of Kozmo's marketing message was also an issue. The company's tag line, which had been "Want some happy?" was changed to "What you want. Delivered today."
"I don't think anyone knew who these guys were," said Sloan. "So clarification was important."
In fall 2000, with outbound telemarketing, a free-standing insert, a radio campaign, some e-mail and co-op advertising part of the mix, the company's efforts began to show promise.
Christmas 2000 broke all of Kozmo's sales records, and Valentine's Day 2001 broke the Christmas records.
"Valentine's Day was probably our first really well-executed marketing event where the merchandising team, the marketing team and the e-commerce team really synched up," Resnick said. "It was by far our biggest single business day, and I would say our first real success."
Sales were so brisk, "everybody including the CEO was down there [in the hubs] picking and packing," Palasin said.
But a low average order size and insufficient population density surrounding some of its hubs would continue to dog Kozmo.
"The prevailing [Kozmo customer] profile was people who had bought a video or CD -- low-ticket items," Sloan said. "They needed people who would spend more money."
In an effort to prospect for higher-dollar customers, Kozmo had The Sloan Group design a catalog for a March drop.
The company's goal was to drop 4 million books and make $5 million in sales, according to internal documents.
Kozmo's average order size was about $25 (up from $9 in early 2000). If that figure stayed steady, and the catalog garnered a 1 percent response, the effort would bring in $1 million -- far below goal.
The mail plan included names from a slew of different lists in an effort to profile higher-spending customers and target buyers like them in a May drop. But the May drop never happened.
The 24-page, digest-sized book offered a hodgepodge of items -- videos, magazines, cat food, pasta, aspirin -- seemingly related only by the promise that they could be delivered within an hour.
The March 2001 catalog dropped the same week as the company shut down, and hit prospects' mailboxes after it was too late to order.
"We inherited something that was going to be out of business in two months, and we got 10 months out of it," Resnick said. "And we came within two months of probably being able to make it work."