Software AG Case Studies Top Loyalty Provider Adds Customers by Performing Transactions in Real Time

Edit This Case Study Record

	Top Loyalty Provider Adds Customers by Performing Transactions in Real Time Software AG

Top Loyalty Provider Adds Customers by Performing Transactions in Real Time

Software AG

Technology Category	Application Infrastructure & Middleware - Data Exchange & Integration
Applicable Industries	Retail
Applicable Functions	Sales & Marketing
Use Cases	Real-Time Location System (RTLS)
Services	System Integration
Challenge	The company, a top loyalty program provider, was facing a challenge in meeting the aggressive end-to-end Service Level Agreement (SLA) of 500 milliseconds for every transaction, as demanded by large retail prospects. The company's data center took an average of 20 percent longer than that — 600 milliseconds — just to query upwards of 32GB of customer data in a central, disk-bound database. Furthermore, Java-related garbage collection pauses caused unexpected spikes in response times that would have resulted in financial penalties for failure to meet the SLA, and in unhappy customers. In order to win new business from larger retailers, the company realized it had to move its data into fast machine memory. Read More
About Customer	The customer is a top loyalty program provider in the marketing services and loyalty industry. They communicate with retailers through a point-of-sale application that manages activities such as adding points to a customer’s account, subtracting redemptions, and presenting relevant marketing offers — all in real time while the customer waits. The company was looking to win new business from larger retailers, who required the company to commit to an aggressive end-to-end Service Level Agreement (SLA) of 500 milliseconds for every transaction. Read More
Solution	The company considered Terracotta BigMemory and Oracle Coherence as options for achieving its performance and scale targets. After a successful proof-of-concept implementation, the company chose Terracotta for its full support for the Ehcache API, better Hibernate support, simpler deployment, a smaller server footprint, and motivated, knowledgeable engineers. The company brought the first retailer into production with all 32 GB of customer information in BigMemory, and immediately experienced an 83 percent drop in data access times — from 600 to 100 milliseconds. At the same time, transaction throughput jumped by a whopping 400 percent. Garbage collection is no longer an issue, allowing the company to meet the 500 millisecond end-to-end SLA demands for the first time. Read More Log in to view content
Contents

Technology Category

Application Infrastructure & Middleware - Data Exchange & Integration

Applicable Industries

Retail

Applicable Functions

Sales & Marketing

Use Cases

Real-Time Location System (RTLS)

Services

System Integration

Challenge

The company, a top loyalty program provider, was facing a challenge in meeting the aggressive end-to-end Service Level Agreement (SLA) of 500 milliseconds for every transaction, as demanded by large retail prospects. The company's data center took an average of 20 percent longer than that — 600 milliseconds — just to query upwards of 32GB of customer data in a central, disk-bound database. Furthermore, Java-related garbage collection pauses caused unexpected spikes in response times that would have resulted in financial penalties for failure to meet the SLA, and in unhappy customers. In order to win new business from larger retailers, the company realized it had to move its data into fast machine memory.

About Customer

The customer is a top loyalty program provider in the marketing services and loyalty industry. They communicate with retailers through a point-of-sale application that manages activities such as adding points to a customer’s account, subtracting redemptions, and presenting relevant marketing offers — all in real time while the customer waits. The company was looking to win new business from larger retailers, who required the company to commit to an aggressive end-to-end Service Level Agreement (SLA) of 500 milliseconds for every transaction.

Solution

The company considered Terracotta BigMemory and Oracle Coherence as options for achieving its performance and scale targets. After a successful proof-of-concept implementation, the company chose Terracotta for its full support for the Ehcache API, better Hibernate support, simpler deployment, a smaller server footprint, and motivated, knowledgeable engineers. The company brought the first retailer into production with all 32 GB of customer information in BigMemory, and immediately experienced an 83 percent drop in data access times — from 600 to 100 milliseconds. At the same time, transaction throughput jumped by a whopping 400 percent. Garbage collection is no longer an issue, allowing the company to meet the 500 millisecond end-to-end SLA demands for the first time.

Impact #1	New business won with 6 large retailers
Impact #2	83 percent reduction in data access time from 600 to 100 milliseconds
Impact #3	400 percent increase in transaction throughput

Impact #1

New business won with 6 large retailers

Impact #2

83 percent reduction in data access time from 600 to 100 milliseconds

Impact #3

400 percent increase in transaction throughput

Benefit #1	83% reduction in data access time
Benefit #2	400% increase in transaction throughput

Benefit #1

83% reduction in data access time

Benefit #2

400% increase in transaction throughput

Download PDF Version

Overview

Top Loyalty Provider Adds Customers by Performing Transactions in Real Time

Operational Impact

Quantitative Benefit