Magazine Article | April 1, 2003

What's Your BI IQ?

Source: Innovative Retail Technologies

You know your customers. You know your store-level sales. Inventory is in check. Your suppliers are complying. But if you don't pull it all together, you don't have BI (business intelligence).

Integrated Solutions For Retailers, April 2003

With all eyes on the bottom line, vendors of BI (business intelligence) solutions are making a big splash in retail. That's because retailers are using BI to save - and make - money. Here, Guy Amisano, president of Salient Corporation, and Jeff Buck, principal at QuantiSense, tell us how.

1. What do retailers demand of BI solutions?

Amisano: Retailers are trying to understand how to improve their bottom lines. With a fixed asset, a store, or a section of shelf space, the key question is, How do I maximize the profit for each square foot? Using POS technology, retailers are now collecting enormous quantities of data about what is selling, for how much, and when. If they are utilizing a loyalty program, they even know who is doing the buying. Good BI solutions are able to tie all of their data together, both transactional (sales, costs, customers, products, categories, time, and employees) and non-transactional (promotional events, floor and shelf space, population, temperature, head count, etc.).

We see a strong demand from retailers for the ability to answer questions like, Who are my most profitable customers? What type of promotion works best to attract more of this type of customer? What are their key demographics? What types of products or categories of products do they buy? What don't they buy? Do I have enough inventory so I do not lose opportunity sales? What is the optimal pricing that will allow me to maximize our margins? Am I allocating my floor and shelf space optimally?

Buck: In the past, only large retailers like Target, Wal-Mart, and American Stores implemented BI solutions because they could afford the high cost and long implementation cycles, as well as a project ROI which was not quickly realized. Today, as mid-sized retailers struggle to compete with larger retailers and with each other, we have found there is a huge demand for BI. Every retailer has BI functionality to some extent, though how that is achieved varies greatly. In many retail organizations we have seen creative variations of hard copy spreadsheets and manual data entry that make up processes that serve as labor-intensive, ad hoc BI solutions.

2. What kind of results statements could a typical retail customer make a year after deploying BI?

Amisano: We have heard statements like:

  • There were stores that weren't even carrying the feature product.
  • A product that has 12% of our shelf space is only contributing 4% to our profit.

In less than a year, we would expect the average retailer that is really using a quality BI solution to see increased revenue per customer, higher sales and margins per square foot, improved promotion effectiveness, reduced inventory, and reduced supply chain costs.

Buck: Most retail BI solutions concentrate first on merchandising. We've heard testimonials like:

  • Having a detailed look at weeks of supply at the SKU (stock keeping unit)/store level enabled us to reduce stock-outs by 75%, which brought roughly a 10% gain to our top line. Because we were able to transfer inventory based on geographical demand and cancel orders for overstocked SKUs, we were able to reduce our inventory carrying costs by 25% and increase our profit margin due to fewer markdowns on stagnant items.
  • Our ability to analyze historical pricing trends side-by-side with weekly unit sales numbers has enabled our merchandisers to determine the price "sweet spot," thus our profit margin has increased on key items.
  • In-depth vendor analysis has shown us the true picture of vendor profit. Because we have a tool to create reports with sales, margin, markdown, and return information, we are better able to negotiate and manage our vendor relationships. This has translated into higher gross margins and improved customer service.

3. What are the three most important features any retailer should look for in a BI solution?

Amisano: Decision makers cannot anticipate every question they will want to ask when they first design a BI solution. Thus, they need to be able to ask any question and receive an immediate response. They can't wait for one report to generate a second report. They need the ability to have a dialogue with their data! They need to have access to very detailed data, not an aggregated level of data, to make these critical decisions. For example, one retail executive I worked with was confident in the fact that his 'average' customer spent approximately $23 per visit to his store. When he looked at this information using the detailed data that retailers spend thousands of dollars to collect, he was able to see that certain category shoppers actually spent an average of greater than $57. With this type of information available, the store modified promotions to attract more of this type of customer, improving revenue and margins. When you use summarized data, your decisions may be clouded by the 'average.'

Buck: Retailers must be able to trust the numbers. The BI solution must be able to take data from many different systems, reconcile the data, and produce trusted numbers to produce reports everyone in the organization can trust. That means one definition of GMROI (gross margin return on investment), margin, weeks of supply, and markdowns.

Retailers need access to detailed information because they cannot run their business with aggregate numbers alone. To understand what makes a department successful or unsuccessful, you must be able to look at the underlying reasons. To take action, retail professionals need access to SKU/store/week historical information. Since retailers often sell a lot of products across a wide geography, obtaining the details is not a simple challenge.

A range of information users with varying degrees of computer literacy and comfort staffs every retail organization. The BI solution must have an easy-to-use, flexible interface that provides aggressive users with rich data manipulation capabilities and average users with an effective means of obtaining the information that they need. In addition, BI solutions must have a deployment plan for users or the solution will likely be underused.