If the relative simplicity of using data warehousing appliance technology makes the hardware and software bundles...
seem like a potential fit for your organization, the next step typically is assembling the business case for a potential deployment. And to get support for investing in appliances, analysts and IT professionals said, the No. 1 thing to demonstrate is that they are a better choice than traditional data warehouse software is.
“First and foremost, be sure that what you want to do can’t be done with a traditional [system],” advised Eric Williams, who was executive vice president and chief information officer at Catalina Marketing Corp. in St. Petersburg, Fla., before retiring from the company in January. Catalina uses more than 30 data warehouse appliances to help it collect and store point-of-sale data, which it then analyzes for retailers and uses to help them target coupon offers to customers during the checkout process.
While cost savings are often the primary driver of data warehouse appliance purchases, that isn’t the most important consideration to Williams. “Many people think appliances will automatically save them money, but the focus needs to be not on saving money but on delivering something the business needs,” he said.
More news and advice on data warehousing technology and issues
Read a buyer’s guide on evaluating and selecting data warehouse tools
Learn why Forrester Research expects more vendors to begin offering data warehouse software bundles
Get Gartner’s take on why data warehouse appliances are no cure-all for performance ills
In Catalina’s case, an extreme need for speed -- to deliver near-real-time responses while ensuring that its system delivers appropriate coupons to consumers as they complete their purchases -- was the biggest factor in the adoption of packaged appliances to replace homegrown ones. “Like other organizations, we can equate speed with an increased value,” Williams noted. “Going from a 10-second response to six seconds delivers money to the bottom line.” As a result, he said, Catalina is constantly looking at which appliances deliver the best speed and total cost of ownership.
Julie Lockner, an analyst at Enterprise Strategy Group in Milford, Mass., said she has found that companies often want to improve their business agility by deploying new platforms such as data warehousing appliances. That element also introduces the issue of portability, she added: “You don’t want to invest in technology that won’t work tomorrow.”
Like Williams, Lockner recommends starting the business justification process by focusing on the ultimate goal of a deployment. “I know it can be hard for people to hear that over and over,” she said. “But if you don’t have a measurable goal, it doesn’t matter whether or not it is an appliance that you want.”
Data warehouse appliance concern: Tying knots, or getting snarled?
Data integration issues also need to be taken into account in assessing a potential appliance purchase, according to Lockner, who said that integration can be the trickiest task in any data warehousing and business intelligence project. “People lose sight of that,” she added. “It gets so complex with multiple data sources and trying to figure out the relationships [between data]. If you are going down the appliance route, make sure it doesn’t further complicate your data integration process.”
Gartner Inc. analyst Merv Adrian said that there are two types of organizations: the ones that already have data warehouses and perhaps some data marts, and the ones that don’t. “Those are two different propositions,” he said, adding that deciding whether to install data warehouse appliances often “depends on where you come from.”
For companies that don’t have existing data warehouse architectures in place, buying appliances is a potential way to “keep a large number of issues off of the table,” Adrian said. For instance, there’s no need to worry about configuring storage and CPU capacity or the number of channels between a system’s memory resources. Putting together an unbundled stack of data warehousing technologies can be a substantial undertaking, especially “if you are trying to get to high levels of performance or volume or [user] concurrency or some combination of those things,” he said.
Lyndsay Wise, founder and president of consulting company WiseAnalytics in Toronto, agreed with Adrian that existing data warehouse systems can influence decisions about whether to buy appliances. “I always believe in seeing what the business hopes to achieve, but with data warehouse appliances you are also looking at what you have in-house and the potential integration opportunities,” Wise said.
An appliance decision might also touch on matters of storage strategy, depending on the databases involved, she added. And that makes it even more important, Wise advised, for companies to assess the business case and think things through in the fullest possible context before committing to any appliance purchases.
Alan R. Earls is a Boston-area freelance writer focused on business and technology.