How BI facilitates a decision-making process that saves millions
At the core of every business decision is the desire to drive value for the company – whether that's increased sales, higher margins, elevated profits, or return on investment. Decision makers should use all the resources at their disposal to drive this value, including their business intelligence software, which may include guided analytics (i.e. dashboards), ad-hoc analysis and collaboration capabilities that contribute to informed decision-making. Today I'll explore how BI software facilitates decisions in a retail scenario. But this article isn't just for retailers – anyone can extrapolate this information to their business to see how BI can provide concrete ROI.
arcplan serves a number of customers in the retail industry, including two of the largest grocery chains in the United States. Retailers are well-known for the small net revenue margins – on average, 3% across the globe for all types of retailers – which pose significant challenges on process controls and efficiency in supply chain decisions. One of the key areas of interest for all retailers, especially grocery chains, is the reduction of shrink – the loss of inventory due to product spoilage, waste, theft and other causes. It's estimated to account for 2-3% of overall sales. Perishable shrink even goes up to 5% within a typical grocery chain. So for one of our customers, whose revenue reached $6.25 billion in 2012, a reduction in shrink of just 0.1% means $6.2 million to their bottom line.
So a simple question that would catalyze a decision-making process at this grocery chain might be: How can BI help reduce my shrink by 0.1% while balancing availability of goods and customer satisfaction? They would want to meet high customer expectations without over-ordering, which leads to shrink through spoilage.
The first challenge is the multitude of systems that contain data relevant to the task. Inventory, point of sale (POS), and supply chain data are typically in separate silos. On top of that, they frequently report in different time frequencies (while POS updates on a daily schedule, inventory might update weekly or monthly). So the first challenge is to integrate these different systems to provide a unified view across those systems to assess sales data, inventory development and supply. In essence answering the questions, How much have I sold, how much do I have, and how fast and how much can be restocked?
The topic of data integration would warrant its own article. Let's just say that at arcplan, we believe the data need to be taken from the original systems without adding a time lag and costly intermediate data layers or data warehouses to the overall scenario. Because arcplan enables real-time access to source data, it can provide relevant information in time for managers to make decisions that prevent shrink and improve the company's bottom line.
The next question is: How should I approach the decision process, top down or at the point of decision? While a top down approach may provide an expert view across different locations, most organizations are finding that reordering decisions should be taken at the point of decision with the most context knowledge, i.e. at the store and department level. Let's assume this grocery chain has roughly 3,000 department store managers, responsible for reordering and restocking decisions. Three-thousand department store managers are not going to do their own ad-hoc analysis or data discovery to identify the goods that need to be reordered while balancing low shrinkage with customer satisfaction. But they have the most context knowledge; they just need a suitable framework in which to perform their analyses.
So what kind of BI would provide the most value here? Surely guided analytics would be best. Dashboards with built-in business logic that incorporate many layers of context (day of the week, seasonality, weather), delivered on an iPad or smartphone, provide great value to mobile store managers who don't necessarily sit behind a desk all day. A guided analytics setup would also include customer satisfaction data as another data stream, and would allow users to filter and view data from many viewpoints, such as period, product or store, and drill down into information they find noteworthy.
The inclusion of benchmarking data is also an important motivator at retail chains, enabling employees to not just benchmark their performance versus other stores, but to also set new, more stringent standards for themselves. At our customers, we've seen these standards evolve, with the help of arcplan's collaborative capabilities, into best practices that influence company operations across the board.
Striking the right balance between supply and demand is especially important for perishable grocery produce. The 2012 National Supermarket Shrink Survey confirms that shrink due to poor planning is a significant source of revenue loss in grocery stores, and companies that have implemented measures to improve the planning process gain 6 to 14% further shrink reduction than those who do not. For grocery stores, the tangible business value of analytics is not having to throw away perishable items because they sat on the shelves too long. By using their BI software in a fruitful way (pun intended), store managers can base orders on customer demand and the natural seasonality of goods, so they waste less and sell more. An empowering BI system that enables more informed decisions has direct impact on the bottom line in this scenario.