Businesses looking to deploy SAP’s High-Performance Analytics Appliance (HANA) should plan on building a team with the necessary IT skills to maximize their return on investment, according to consultants.
Customers deploying the HANA in-memory technology will likely have to rely on outside vendors to get the system up and running because the technology has only been on the market since last summer. After that, they should look at transferring those skills to an in-house IT staff to avoid making an already expensive investment even costlier, experts said.
The SAP HANA system for large enterprises consists of an appliance sold directly by a small network of hardware vendors. HANA software, sold by SAP, is then loaded on to the appliance before deployment. Customers can also build custom applications that run on HANA.
“One of the downsides to new products is if people don’t have requisite skills maintaining, optimizing, customizing, you become dependent on your vendor and consulting partners,” said Jon Reed, an independent SAP analyst and founder of JonERP.com. “Every time you have a problem, do you want to get IBM on the phone? Probably not. You’d like to handle that internally.”
To start off, look to SAP hardware vendors
Companies looking to deploy SAP HANA in-memory technology should avoid the temptation to do it on their own and instead get the hardware vendor to get the system up and running, according to Ethan Jewett, a consultant with the German consultancy Business & Decision.
“I don’t believe there is any other way to get a supported HANA setup,” Jewett said.
What happens next with SAP HANA projects?
Companies should then take over the administration of the HANA environment, according to Vijay Vijayasankar, a consultant who specializes in SAP business analytics at IBM Global Business Services in Phoenix.
Supporting the HANA appliance is a straightforward process that involves covering things like connecting to the network and recovery and backup issues, he said.
“The hardware part of it is pretty routine,” Vijayasankar said. “Most people can pick it up in one or two days,”
Application support is a bigger challenge that requires a combination of hardware, data provisioning, reporting and data modeling skills, according to Vijayasankar. Users will also need someone familiar with HANA’s variation of Basis administration.
Large companies are likely to have BI reporting skills in-house, while data modeling skills are rare and take longer to learn, Vijayasankar explained. A company could also try to hire someone with those skills, but that won’t be easy.
“They are hard to come by,” he said. “HANA has only been around for a little while.”
In-memory technology only gets you part of the way
The emphasis on getting the right kind of people with the right kind of skills underscores the limitations of in-memory technology itself, Jewett said.
“A HANA system that is firing on all cylinders is good, but it doesn't inherently create any value for a company,” he said.
Companies have to make sure that their data is accurate and that they know what to do with the information that comes from that data.
“In my view data quality, deep analysis, modeling skills and business domain knowledge are central competencies that companies can use to extract value from their data. Technology is important, but it acts as a force-multiplier for these competencies,” Jewett said. “If these competencies don't exist or are not well-developed, then a company is not going to get the best value of out of HANA or another technology investment.”