In 2005, Exmplar sent out more than 1 billion e-mails, "and not one piece of spam," boasts Alex Sukhenko, chief technology officer of the multichannel direct marketing company in Fairfax, Va.
Indeed, Exmplar is no garden-variety direct marketer, blasting millions of e-mail come-ons to unsuspecting consumers in the hope that some fractional percentage will "click through" to a Web site and buy something. Not only were all 1 billion of its 2005 e-mails sent to its clients' existing customers — who opted to receive e-mail promotions — but each was personalized to the recipient based on past purchases and demographics.
A survivor from the late 1990s dot-com era that spawned, then claimed, hundreds of e-commerce startups, Exmplar has leveraged technology and expertise developed over those years to reinvent itself. Founded in 1998 as an e-mail marketer called Shop2U that created customized electronic catalogs for its clients, Exmplar has evolved into a multichannel marketer that helps retail clients create individually personalized shopping experiences for their consumers. Using proprietary software and an IT infrastructure that scales to support huge campaigns for clients such as 1-800-Flowers.com, Champion Sportswear and New Line Cinema, Exmplar delivers targeted offers via personalized e-mail, Web sites and printed catalogs or flyers.
"Sending a promotion to a million-plus people and hoping that they'll come into your store isn't very effective," Sukhenko says. "If you target, they're more likely to act on it."
Exmplar calls such targeting, which is based on customer demographics and transaction history, "brand personalization." It goes beyond Amazon.com's well-known technique of recommending items to shoppers based on their past purchases and those of others who've bought similar merchandise. The company applies a complex set of proprietary business "rules" to determine the offer or marketing message that each customer receives, whether it's an e-mail, a customized Web storefront or a printed piece sent in the mail. Exmplar has used its technology to rise, like a phoenix, from the ashes of the dot-com meltdown, but it also illustrates the power and maturity of data-collection and analysis tools that other businesses, particularly retailers, can use to grow.
"What they're doing from a personalization approach is very much in the current mind-set," says Patti Freeman Evans, a retail analyst with Jupiter Research in New York. Most retailers don't have the internal expertise to do the complex analytics that are embedded in Exmplar's rules and software algorithms, she notes; nor did they have in the early days of e-commerce the granular customer and transaction data used in Exmplar's system. "Now, retailers have a lot more data on their customers, so we can make the rules, track the actions and make more informed decisions about how to go forward."
"If you talk to the more sophisticated merchants, they're moving in that direction, so that works in Exmplar's favor," agrees Lauren Freedman, president and founder of the e-tailing group, an e-commerce and merchandising consulting company. "Most of the market probably wasn't ready for the segmentation and personalization — they were just trying to figure out how to get their e-mails deployed," she says of the first e-commerce wave. "But once they find it works, they want more."
Unlike traditional direct marketing campaigns that distribute the same promotion or catalog to all recipients, Exmplar uses client-supplied data — including detailed customer demographics and two years of transaction history — to customize the message sent to each customer. For example, a customer who previously bought a Gore-Tex jacket might receive an e-mail notice of a sale featuring the matching fleece liner for the jacket. Getting personal is more expensive than typical mass e-mail marketing: five cents or more per e-mail, depending on volume, versus traditional outsourced e-mail campaigns that cost as little as a fraction of a penny per message. But the results are markedly better, Sukhenko says.
"It increases [sales] from 50 to 200 percent or more, depending on the strategy," he says.
Exmplar's personalization rules use variables such as the retailer's objective (up-selling, cross-selling, inventory liquidation, new product introductions, etc.), the targeting method (e.g., choosing customers to receive a promotion based on demonstrated brand or product affinity), customer purchasing history (frequency of buying, average dollar amount and how recent was the last purchase, etc.) and the specifics of the promotions being offered (discount percentages, two-for-one offers, product bundles, etc.).
The secret sauce of Exmplar's brand personalization service is its proprietary software, written from scratch by in-house programmers. The technology staff, almost one-quarter of the company's 40 employees, also maintains its 10 servers — configured in five database clusters, each with a primary and a backup server — running Microsoft's Structured Query Language Server and the open-source JBoss Java application server. Last fall the company migrated from disk arrays to an EMC CLARiiON storage area network to provide greater reliability and flexibility in configuring its database storage.
"Making an investment in a storage network has been a huge benefit to the company," Sukhenko says. "The cost, although higher, more than pays for itself when it comes to management."
Lessons from a survivor of the dot-com crash:
Exmplar's applications don't require real-time data access or absolute 100 percent uptime, like a hospital or a financial institution, he acknowledges. Yet when processing e-mail campaigns for clients that send 25 million e-mails and use 30 different types of campaigns per month, reliability and flexibility are critical. "If a machine was to be kicked and a plug came out, that becomes a big problem. We can't afford to face downtime."
CEO Keith Wardell calls the IT team "critical," and Sukhenko sits on the company's three-person executive committee. Even after seven years in business, the company's culture is that of a fast-paced startup, Sukhenko says, but there's not a refrigerator full of Jolt or employees working 12-hour days. "We're not working like we're going to vaporize." With 2005 revenues of about $5 million, the company is just breaking even, but Wardell expects to receive new funding soon, which will help fuel growth of 50 to 60 percent per year, doubling revenues within the next few years.
Wardell says he learned valuable survival techniques from the dot-com collapse and the company's resulting restructuring. Among them was the importance a strong foundation and the folly of the late 1990s mantra that startups should get big fast to secure "first mover" advantage regardless of cost.
"Make sure you have a business model with a proven value equation, and don't get large until you can consistently prove the value equation," Wardell advises would-be entrepreneurs, noting that Exmplar spent seven years figuring out how to use its personalization rules. "We focused on the methodology before we ramped up growth. A lot of companies in the dot-com era got big too quickly, and they didn't make it."