Outback Steakhouse Case Study
Autor: andrey • March 8, 2011 • Essay • 995 Words (4 Pages) • 919 Views
Case Study: Outback Steakhouse
Outback Steakhouse grills spam with Proofpoint
clogged inbox complaints were rampant
Outback Steakhouse, Inc. headquartered in Tampa, Fla. was founded in 1988 and operates
Outback Steakhouse units throughout the U.S. as well as Carrabba's Italian Grill,
Lee Roy Selmon's, Cheeseburger in Paradise, Bonefish Grill and Paul Lee's Chinese
Kitchen. The company has 1,100 company-owned stores.
According to Outback Steakhouse Director of Technology Infrastructure, Michael Boykins,
"We were getting a ton of calls from employees about being inundated with a high volume
of all types of junk mail. This created a number of problems for the company."
The first problem was lost productivity as workers had to spend large blocks of time
sifting through junk mail every morning. This problem was compounded by a drain on
IT resources as IT staff attempted to manually deal with the spam problem and IT infrastructure
(such as mail servers, network bandwidth and storage systems) was becoming
strained by the high volume of unsolicited messages. Lastly, the company felt it was
important to protect its employees from exposure to offensive email content.
Any message protection solution Outback Steakhouse chose had to be extremely accurate
due to the nature of the company's business. Many legitimate emails received by
Outback employees deal with subjects and products commonly found in spam emails. For
example, messages that contain content about "beer," "liquor," "cocktails" and other food
and beverage terms can be inadvertently blocked by less sophisticated anti-spam techniques
such as keyword matching or simple Bayesian techniques. A solution that filtered
these emails incorrectly, creating "false positives," would cause more harm than good.
taking spam off the menu
Realizing that something had to be done, Outback Steakhouse's IT team began evaluating
a variety of anti-spam solutions, including a test deployment of a server-based