Paycor

ManualsBrandsProLiant DL980 servers ManualsPrint & ScanIT Case Study

Paycor tackles peak workloads with newest

HP ProLiant servers

To enable continued growth, payroll service provider slashes peak utilization

from 75-80% to 20%

“One-to-one client service is the foundation we

build our business on. With the DL980 servers,

our associates now are supported by the level of

performance they need to provide clients with

quick, accurate information.”

– Kevin Armour, chief technology officer, Paycor, Inc.

HP customer case study:

HP ProLiant DL980

Servers, HP Proactive 24

Service

Industry: Financial

services

The challenge: 130,000 paychecks

per day

Founded in 1990, Paycor is one of the largest

independently owned payroll processing service

providers in the United States. Based in Cincinnati,

Ohio, the company has 19,000 clients, offices in

18 cities, and 550 employees.

On a busy day, Paycor executes 1.7 million

transactions to support the processing of 130,000

paychecks. Payroll processing is characterized by

peak work loads. Many employers are on a weekly

payroll schedule. They process the previous week’s

payroll data early in the week, so checks can be

issued on Friday. For Paycor, the result is a very short

Objective

Support continued business growth, retain clients

by delivering a high level of one-to-one service,

and ensure the availability of critical payroll

processing applications during peak loads

Approach

Undertake an effort—including benchmark reviews

and vendor visits—to identify and implement a

powerful new IT infrastructure within four months

Business technology improvements

Peak-period CPU utilization reduced from •

75-80% to 20%, providing headroom for growth

Time to download large data files from clients •

reduced by 50%, boosting staff productivity

DL980 servers can support a large number •

of virtual servers, resulting in reduced

power, cooling, and data center floor space

requirements.

Business outcomes

Boost in processing power provides the •

headroom needed to support continued growth.

Faster response times results in higher client •

satisfaction and retention.

Improved application performance allows client •

service associates to be more productive, for

cost savings and better service.

High system availability reduces risk of lost •

business due to outages or missed deadlines.

Summary of content (4 pages)

PAGE 1
Paycor tackles peak workloads with newest HP ProLiant servers To enable continued growth, payroll service provider slashes peak utilization from 75-80% to 20% “One-to-one client service is the foundation we build our business on. With the DL980 servers, our associates now are supported by the level of performance they need to provide clients with quick, accurate information.” – Kevin Armour, chief technology officer, Paycor, Inc.
PAGE 2
“We are getting comments about the system being ‘a lot faster.’ Some large database downloads from starting new clients that used to take 10 minutes now take half that time. Overall, I think that downloads are actually more than 50% faster.” – Kevin Armour, chief technology officer, Paycor, Inc. production window in which to take in the data, generate checks, process secure direct deposits, and handle wire transfers to banks. “A 12-hour period each week—from 1:00 to 5:00 p.m.
PAGE 3
Customer at a glance Name: Paycor, Inc. Headquarters: Cincinnati, Ohio Telephone: 800-381-0053 Number of employees: 550 URL: www.paycor.com “We liked the opportunity to move back to ProLiant,” continues Armour. “All our other infrastructure is ProLiant, so it was a natural fit.” The expected availability of the DL980 also fit Paycor’s schedule. Armour told Agilysys, “If we have it by mid-November, it will work.
PAGE 4
After this initial production run, the rest of the week went just as well. “The peak 30 seconds never went more than 22% utilization,” Armour asserts. “All the indicators were extremely low. Now we know we have the horsepower to support peak loads.” Armour notes that user response has been enthusiastic. “We are getting comments about the system being ‘a lot faster,’” he says. ”Some large database downloads from starting new clients that used to take 10 minutes now take half that time.