“We were getting literally millions of e-mail messages pointed to our domain,” says John White, vice president of IT at Central Insurance Companies, a regional personal and commercial lines carrier. “We get millions of those from spammers. It's a bandwidth issue and a storage issue. We needed something to filter out all the garbage.”

Central turned to Gateway Defender to deal with the issue. “We had been trying to manage this internally, but Gateway Defender had a solution that fit our needs perfectly,” says White. Central researched other services, but White found them to be more expensive and not as hands-on in dealing with customers. “We have very good access to these folks because they are hungry and they are small,” he asserts. “You don't go through a lot of layers of technical folks to get some information.”

What White likes about Gateway Defender is the service is on a front end Central can work with but doesn't have to maintain. “We're not dedicating ser-vers or resources or anything in here,” he explains. “It's all external, and [e-mail] is filtered on [Gateway's] servers.”

Gateway Defender offers the capability of opening or closing the mailbox as much as the customer wants. “If there are specific senders out there whom we know we are getting bombed with, we just can close off anything from them,” says White. “Or we can say [to Gateway] here is something that isn't passing the filter for some reason, and it's a business partner of ours. We can open the door [for those messages] because we know they are a legitimate source.”

Central provides Gateway with all its valid e-mail addresses. From there, certain filters can be adjusted. “[Gateway] is going to start with a base set of filters. An example would be adult material,” says White. “Gateway is going to screen the header and some of the content looking for keywords. If something related to adult material comes in, [the system] is going to stop that even if [the message] has a [valid] address. It does the same thing with questionable language. If there are any executables that could be dangerous, those are going to be blocked.”

Dealing with spammers is like a game, comments White, with each side trying to stay ahead of the other. “Let's say we have some filters in place, and they are working on 98 percent [of the e-mail],” he says. “In two months, they might be working at 80 percent because the spammers are figuring out how to beat the filters.” When inappropriate material makes it through the filters, he adds, it is returned to a Gateway Defender e-mail account where the company will examine the message and determine how to beef up the filters to stop similar messages.

Some messages get stopped on the Gateway servers and reside there. “We have a process where we review those to make sure there is nothing out there that should come through,” says White. “The nature of the beast is something gets blocked that should have been allowed through. If one of our employees comes to us [with this problem], we can go to the Gateway Web site, scan our list of e-mail, and see whether there is something that got blocked. Our tech systems people can release [the e-mail]. It happens rarely. If you have any type of decent filtering in place, you're not going to get away from that. If there is any downside, it is the small percent that potentially gets blocked that shouldn't get blocked–less than one percent. What's important is we have a mechanism in place where we can go and check that.”

In the first six months of 2005 Gateway Defender blocked more than 4,000 e-mail messages sent to Central that contained viruses. “We have our own virus protection for what does get through, but we want [Gateway] on the perimeter to block as much as it possibly can,” concludes White.