Fast-casual restaurants are a relatively new industry segment that offers counter service and high-quality food in an upscale ambiance. They provide fast service with made-to-order but affordable food, and they are growing.
In fact, fast-casual restaurants are growing so fast that they are starting to eat into the market share of leading fast-food chains. According to Technomic’s 2014 Top 500 chain restaurant report, sales for fast-casual chains grewby 11 percent. However, even with all the success and popularity, fast-casual restaurants have been slow in adopting advanced analytics and building big data infrastructures – an exclusive resource once reserved for the market giants that can deliver the segment unprecedented ROI and growth opportunities.
With the advent of scalable cloud technology and the “analytics-as-a-service” industry, what once was too large and expensive for a growing fast-casual chain is now readily available and affordable. The fast-casual chains are finally able to glean insights beyond the business intelligence (BI) their point-of-sale (POS) systems provide and can now use advanced analytics and data science to drive performance.
The key difference between BI and advanced analytics is that BI provides information on what happened in the past but not the root cause correlative to the understanding of why things happened or their drivers. This “why” is key to learning how to predict using predictive analytics, which is closer to predicting the future than any other form of technology. In the past, a restaurant had to rely on instincts and experience. Today, it’s possible for a restaurant to get outside, unbiased, data-driven information about the future.
The Danger of Being Fooled by Averages
There’s a statistical joke that goes like this: A statistician confidently tried to cross a river that was one meter deep on average; he drowned. Restaurants can be fooled by averages from their POS system. Without slicing and dicing the data further and taking a multi-faceted, dimensional look, data can be deceiving.
For example, looking at average sales by product for the month or quarter might show significant results for a particular product. However, diving deeper might show that most of those sales are on promotion days at deep discounts. Thus, the sales volume might be high on average, but profits might be flat or even at a net loss when drilled down to a more granular view.
Moving up the Spectrum of Analytics
Restaurants collect and store their data in their POS systems, relational databases, and enterprise data warehouses. The first stage in the spectrum of analytics is reporting and descriptive analytics, which explains to the restaurants what has happened. For example, how many dinner entrees were sold last week, last month, and last quarter? The next level would be to dive deeper into root causes.
