Published on 30-Mar-2012
Carter’s is the leading brand of children’s clothing, gifts and accessories in America, selling more than 10 products for every child born in the U.S.
Carter's was concerned that their existing Oracle database wouldn't be able to scale, and the client was already facing challenges with Oracle's data structure, performance limitations and administrative overhead. Carter's was missing its SLAs and could not capture detailed level data (SKU, store, or transaction level details). Carter's also wanted to gain comprehensive visibility into its Merchandising, Allocation, Planning, and Business Intelligence (BI) environments.
Carter's deployed their Netezza 1000 appliance as their EDW (Enterprise Data Warehouse) with a QuantiSense retail model. After evaluating Netezza, Oracle Exadata, and Teradata, Carter's decided that Netezza was the right fit for them and migrated from Oracle to Netezza in less than 6 weeks.
Carter's reduced: their 12 hour analytical data reformat down to 90 minutes on Netezza; and they reduced the number of reports from 120 to 15. And, they now have a 360 degree view of their business, as opposed to the single view on their Oracle system. Carter's has better, faster analytics. New features include: fresh sales trend data, new market basket analysis, new imminent stock out analysis, and meeting all SLAs.
At NRF 2011, Mark McSwain of Carter's shared the story of Carter's evaluation, selection, and deployment of the IBM Netezza data warehouse appliance for cross-channel, multi-brand analysis. After evaluating Netezza, Oracle Exadata, and Teradata, Carter's decided that Netezza was the right fit for them and migrated from Oracle to Netezza in less than 6 weeks. This video shares highlights from Mark McSwain's presentation.
Mark McSwain, Director-Information Technology, Carter's
Retail BI for us is a strategic undertaking and we really run the business using BI. So every day, to get operational reports, to be able to drill down into those by store and region is very, very powerful for Carter’s. SKU, store, transactional level data is very important. Most transactional systems don’t carry that level of data for any length of time. We had a requirement to carry it for a minimum of 2 years.
We needed a level of visibility to a single view of merchandising for both planning, allocations and BI. We obviously had to re-platform and build that structure and we needed integrated data for multiple brands and multiple channels.
We began to take on a project to evaluate Netezza, Oracle Exadata and Teradata. At the end of everything that we did, Netezza came out the winner for a number of reasons. It was the most straightforward application to put in, to learn, and to execute on. We ended up doing a proof of concept with Netezza, it was very quick. They brought the box in, set up a test, very successful results. Some of the challenges that we had with platforms like Exadata, it’s essentially, we viewed it as an Oracle platform just on multiple machines. So the same issues you had with scalability, maintenance, administration carried right over into that environment. And we weren’t willing to take that on.
Teradata – actually they were in transition from a big DW player to an appliance vendor and we just didn’t feel comfortable that they owned that space yet. We felt Netezza was right for us. We were very surprised to see that we could actually re-platform from Oracle to Netezza in a very short period of time - actually less than 6 weeks to redo that entire ETL, data model and reporting on top of it.
We load 72 weeks of sales history nightly into an allocations system, turn that – that same process is the one that I was talking about, it would take 12 hours in Oracle. We rewrote it in Netezza and it runs in 45 minutes. We gave Netezza a challenge. We actually gave MicroStrategy QuantiSense a challenge. We said here's what we need to do, it's a quick project. It's running along with three other high visibility significant projects.
From a business impact, obviously we’re now meeting our SLAs. We can react to selling trends much faster because we have a nightly reporting of information. Most importantly, a single platform supporting multiple brands and channels and enabled market basket analysis with QuantiSense and MicroStrategy platform that I quite honestly don’t think that Oracle could have supported that type of analytics at the transactional level for what we needed to do.
Products and services used
IBM products and services that were used in this case study.
IBM Netezza 1000