Moving from Gut Instinct to Advanced Analytics Is a Winning Strategy
There may be some situations in life in which making decisions on the basis of gut instinct is useful or even necessary. Business-critical decisions in today's light-speed, ultra competitive environment, however, are best made in a much more informed manner.
This is certainly the case for PlayPhone, a leading global media company that provides entertainment content including ringtones, mobile games, wallpaper, videos and more to mobile consumers in over 20 countries.
"When you're running a business like ours," explains Simon Rose, PlayPhone director of infrastructure for its European, Middle East and Africa region (EMEA), "success is all about generating volume. That requires the planning and execution of simultaneous, large-scale marketing campaigns. If you have no way to plan, execute and evaluate these campaigns rapidly, based on accurate, current data on customer behavior and purchasing histories, and other relevant metrics, you can run through a lot of marketing money very quickly, only to learn after the fact that a particular campaign or multiple campaigns weren't successful. With so much on the line, you really don't want to rely on gut instinct to make critical business decisions. In order to run more efficient and cost-effective campaigns and achieve our goal of becoming the largest subscription-based mobile content provider, we knew we needed a robust, reliable and affordable business intelligence capability."
Rose and his IT colleagues in PlayPhone's EMEA division, operating out of London, had earlier addressed one critical success factor – the implementation of a transactional system geared to very high performance for processing customer orders. Mobile customers, not surprisingly, expect WAP (wireless application protocol) pages on their devices to render immediately and are not very patient as their orders are processed. They don't care about the multiple actions that have to run behind the scenes, such as confirming their identities, verifying credit information and actually processing the transactions. In this real-time environment, customers can grow dissatisfied if they're made to wait more than a second or so between their actions and the expected response from PlayPhone.
With a background in financial services IT, Rose made a decision early on to implement Sybase Adaptive Server Enterprise (ASE) to ensure high performance, reliability, security and low administrative support requirements. That choice proved to be a good one from both PlayPhone's and PlayPhone's customers' perspectives.
Gaining a Comprehensive View of Customers and Campaign Effectiveness
The next challenge Rose and his colleagues faced was gaining a better and more rapid understanding of PlayPhone's customers and of the effectiveness of its marketing campaigns. The company had been running nightly and weekly reports on a batch basis to provide it with a view of its business performance. Running these reports against the operational database, however, had several drawbacks. PlayPhone EMEA could not run all of the reports it wanted to run: Several of the reports it did run took quite a bit of time because of the large volumes of data being queried. And finally, running reports against the transactional database ran the risk of slowing down the performance of the transactional database and possibly frustrating customers.
To solve this problem, PlayPhone EMEA again turned to Sybase technology – in this case to Sybase IQ, the company's industry-leading column-based analytics server.
"Sybase IQ Had All of the Advanced Analytic Capabilities We Needed"
"We needed a way to support our analysis and decision-making processes regarding which customers to target with which campaigns, and to evaluate the effectiveness of our campaigns in something closer to real-time," explains Rose. "We wanted to be able to analyze customer acquisition costs and customer profitability, as well as understand which campaigns produced the best responses among various groups of customers in different countries. Of course, we also wanted to be able to minimize customer churn. We knew that the more of this kind of information we had available, the better equipped we'd be to maximize the impact of our marketing spend and grow our business."
PlayPhone's criteria for an analytics server included ease of support and low support overhead (to allow staff to focus on the data, not the server), high performance, data security, the ability to run on Red Hat Linux, the ability to run on small specification machines, efficient storage and proven reliability. Rose was aware of Sybase IQ, with its reputation of delivering dramatically faster results at a low cost.
"I was familiar with Sybase IQ and had seen a demo of it at a show that knocked my socks off," Rose says. "It seemed to have all of the advanced analytic capabilities we needed to improve on our ‘gut feeling,' enabling more empirical decision-making based on access to large volumes of current transactional and historical data. With the ability to query this data and receive results within seconds or minutes versus hours or days, we could target customers more effectively and fine tune marketing campaigns to maximize the impact of those campaigns."
Proof-of-Concept an Unqualified Success
PlayPhone EMEA engaged Sybase Professional Services to assist it in creating a proof-of-concept with Sybase IQ running on Red Hat Linux on a commodity server and employing the SAS Business Intelligence Server as its front end.
"The proof of concept very quickly demonstrated the power of this solution," says Rose. "Standard reports involving large joins of large volumes of data that previously took up to an hour ran in less than a minute with Sybase IQ. There were other reports that we couldn't even run before having Sybase IQ. These new reports provided us with answers we needed to make better predictions and decisions."
Based on this success, PlayPhone rolled out the Sybase IQ-powered business intelligence system across its EMEA enterprise. Currently, all of the division's transactional and archival data – approximately half a terabyte and growing – resides in Sybase IQ.
"Sybase IQ has enabled us to gain a single view of our data for reporting and data mining purposes," says Rose. "It's been extremely stable and reliable. I don't think we've ever had to call customer service, which really speaks volumes about its quality and reliability. It's really been faultless during our implementation. Also, having the ability to use the developer edition of Sybase IQ has allowed us to try different things before actually implementing them.
Adds Rose, "Sybase has provided us with a great product suite. ASE is very stable and has a very low DBA to server ratio compared to the competition. Sybase IQ is a brilliant analytics engine. Coupled with a good reporting and data manipulation tool, it's hard to beat. I don't know that the business would still be here today without our Sybase solutions. Because you've got to have a lot more than gut instinct to guide you in the current business environment."