Three years ago, Sun Microsystems had a database of millions of email addresses from all over the world. When the company's marketing staff wanted to send an offer out, it used segmenting: Most prospects in a particular country might get one offer, while those who had expressed interest in a particular product or service got another.
While this tactic helped Sun increase the relevance of the offers it made, the company still felt something was missing. Sun wasn't taking advantage of everything it had in its marketing arsenal. Recently, the company did just that when it started taking into consideration both implicit and explicit details about its prospects and customers to send them marketing messages.
"What we do today is look at the actions people are taking on our Web site, look at the profiles that they've filled out and send them the most relevant offer and piece of content based on that information," said Felix Serna, Sun's director of global e-marketing.
Sun, which works with Acxiom Digital on email marketing, has a wealth of information on which to base its decisions, Serna said. For example, the company can make an informed decision about what someone might purchase next by looking at which of Sun's online communities the prospect is involved in, which software they downloaded most recently and where they've been spending time on the Sun site. It does this using a combination of the company's Web analytics, an Oracle database and Acxiom Digital's IMPACT email marketing platform, Serna said.
"People will get a transactional email. They'll get an email if they subscribe to something or join a community. They will get an email when they download something—a targeted cross- and upsell based on that trial. Emails go out to people who have a service plan expiring," he said. "The triggered emails based on behavior happen based on a behavioral score. If someone is clicking around on articles in the newsletter, it will push the score to a threshold where the database will automatically send them another email with content that's most significant [to them]."
In a nutshell, Sun is doing what retailers have done for years: upselling and cross-selling before a single purchase is completed. Later this year, Sun plans on taking its behavioral targeting to its site-based ad serving, hoping to get the same lift, Serna said.
"We look at complete performance when we're serving advertising. For example, when I have a sale, I can look at what vehicles and offers are getting the most action and tie it to reminders and communications," he said.
This means a customer can fit into one marketing communications buc-ket today and a different one tomorrow. Also, customers in the same "segment" may get different offers.
One of the key pieces of the program: Sun never assumes a customer or prospect's life cycle position, Serna said. Even if he thinks a particular segment may respond best to a specific offer, he still does plenty of A/B testing.
Since implementing the program, Serna said, the company's click-through and response rates have been two to three times higher than they had been previously.
"That's the minimum," he said. "Sometimes rates are three and four times higher."