formed. This more effectively helps you plan for the holiday season.

Using business intelligence to establish a merchandise strategy must begin well in advance, at least a year. You need to start planning the best merchandise fit for customers during the key season by looking at comparative trends, internal transaction data and syndicated data, such as consumer demographics from companies such as Nielsen.

Avoiding Overstock Sales Starts at the Supply Chain

CIO: Last year, many stores ended the holiday season with leftover stock and were forced to mark down a lot of merchandise. Why is getting that merchandising strategy right so difficult?

Anand: Retailers tend to focus on the front end at the expense of the back end. They effectively plan for how their stores, website and catalogue are going to look. But the back-end that involves reviewing and analyzing customer data for taking effective merchandising decisions is always an issue. It's a race against time in the case of several retailers as their data intelligence strategy is not in place or the response to BI is too slow (non-real-time).

Back-end focus is the key towards a successful sales season. Retailers must focus on the entire retail supply chain—product sourcing, procurement, the order-to-time cycle. They must identify which products will actually sell, and create an entire lifecycle management strategy right from product development to delivery nine to 12 months before a key selling season. And business intelligence plays a huge role in it, because it identifies the problem spots for the company in that entire lifecycle management. It also identifies opportunities to improve, in terms of the sourcing of the products, and actually mapping that sourcing with consumer trends.

Problem: Dirty, Distributed Data

CIO: What is the current state of the retail industry’s use of business intelligence and what are some of the challenges?

Anand: Aberdeen research shows that 71 percent of retail companies are using business intelligence, primarily to manage customers, plan business strategy, and for merchandising. Still, only about half of companies are using BI for merchandising, so I think that trend has to improve. The enormous volumes that retailers have to plan and execute can be challenging.

The main challenge that companies face using business intelligence is that their data is not clean enough for analysis. Forty-nine percent of our survey respondents listed dirty data as a top challenge when initiating or implementing business intelligence or predictive analytics. Another associated challenge is that the data is scattered throughout the organization; it's not centralized. Coming up with good sales analysis, space planning analysis, assortment planning analysis, market basket analysis—all of which can help a company create a good merchandise stack for holiday season—requires integrated data. This is isn't possible without