Data Mining and Fast-Food Dining

The Food Business Today – Turning Bites Into Bytes

Globally, the entire quick service restaurant (QSR) field is facing a formidable challenge: make informed results-oriented decisions or lose out to brands who will. Certainly, it’s a daunting task for restaurants also facing pricing wars, brand relevance, menu popularity, sustainability, social image, and the ever-present consumer demand for value, which is why data gathering and analytics often finds its way to the back burner. Make no mistake, however, that this is actually counterproductive since the lessons learned from a front-and-center analytics system can positively impact almost every other challenge QSR brands are facing in today’s market. So Why Are Brands Postponing or Ignoring Analytics? The short answer is simply the belief that a robust analytics infrastructure is too difficult or too expensive to implement. This is only partly true. Debra Miller, the Vice President of Business Analytics for Church’s Chicken, explains. “In food service in particular, there’s a huge gap between new-out-of-the-box technologies for collecting and processing data and what we call ‘legacy systems’ which are those point-of-sale solutions that were in place prior to some of the more significant advancements in metrics and data.” In most cases, when existing legacy systems were first implemented…
  • It was done so at significant cost to the organization
  • It required a sizable time investment to bring all locations and franchisees onboard
  • It successfully delivered more data than no system at all
Interestingly enough, these are the very same reasons many brands are so fiercely determined to hold on to their old legacy systems even 10, 20, or as many as 30 years later. The idea of investing time and money on a much larger scale just to get “some more data” seems like an irresponsible luxury. Or Is It? One of the common pitfalls surrounding the legacy system switch is the “all or nothing” mentality. There is a prevailing belief that overcoming an existing (or in some cases non-existent) system is cost prohibitive and no other options exist. The truth, according to Miller and other analytics experts is that there are interim solutions that can yield rapid and meaningful results for organizations that choose to implement them. TITLE:    The Costs of Common Sense Analytics At an average cost of $31 million or more, it’s easy to understand why organizations often balk at a complete system overhaul. However, an approximate investment of around $100,000 to enhance what a legacy system is providing is not only a smart decision – it can have long-range impact on profitability that actually helps finance a complete overhaul in years to come. “What quick service restaurant corporations are actually facing is a need to get smarter about the data they have, rather than just switching method for collecting it,” says Miller. It’s very disappointing for a brand to spend millions on a new system only to underutilize it. They end up in a very similar situation to where they were in the legacy days, with little to show for the investment.”

Legacy – Today – Tomorrow – What’s the Difference?

In deciding to take a more proactive approach to business analytics, it is first necessary to understand:
  • What kind of data you need
  • What kind of data you have
  • How to bridge the gap between the two
The Data You Have vs. The Data You Need Most POS systems in use the QSR industry today are capable of delivering fairly robust sets of raw data. Daily or monthly sales figures, average units per sale, average spend per sale, profit/loss information, cost of goods sold, and other figures are essential to ongoing business operations. But they’re only part of the big picture. Consider the following scenario: Deli XYZ has 100 locations nationwide with a legacy-style system. GMs run reports on a weekly or monthly basis and send to corporate, per the terms of the franchise agreement. As expected, Deli XYZ’s sandwich sales are steady month to month across the network. But then, a few stores in the Northeast start turning in lower-than-average total sales numbers. Without knowing all the factors at play, the decision is to lower sandwich pricing in that region to spur more sales. A few months later, those stores are turning in even lower numbers. What gives? This is a prime example of incomplete analytics at work. Yes, under normal conditions a discount or temporary price cut can overcome a temporary dip in the market. But what if something else is at the heart of the situation? Let’s say in the same scenario above that a closer examination of the data shows that well before Deli XYZ’s Northeast stores started struggling, they were actually seeing an increase in soup and salad sales. At first it may seem like an unexpected bonus, but thorough analysis could recognize it as a trend, and perhaps roll out combos or promotions aimed at these eager customers. Instead of cutting away at a profit center (sandwiches) – Deli XYZ’s corporate office might be able to launch soup-and-sandwich or salad-and-sandwich offerings to up-sell to the audience and prevent a profitability downturn. Based on those results, it may even prove to be a move that’s effective across the entire 100-store network. In this case, Deli XYZ had useful sales data, but they needed trend data to identify what was truly happening in the market. In most cases this is not a feature offered by most legacy POS systems… but it’s not exclusively available through new overhauled systems either. Having a dedicated analytics team in place can bridge the gap between what is and what needs to be in a very common-sense way, without the multi-million-dollar expense.

Assembling Your Analytics Team

With any data and POS system, legacy or otherwise, it is absolutely vital to have experienced people attached to the effort. Data doesn’t analyze itself… at least not fully. Even the latest and greatest technology solutions are confined by certain limitations. (SIDE BY SIDE LISTS – DON’T NEED TO SHOW AS CHART)
What POS Systems CAN Do… When POS Systems Need PEOPLE…
-Quickly tabulate large sets of data -Send alerts for upper and lower limits -Identify simple patterns -Provide unbiased neutrality based on cold, hard facts only -Making decisions based on compiled data and figures -Setting parameters for “low” and/or “high” numbers -Recognize trends -Understand when situations go “beyond the numbers”
Clearly, the people handing analytics in your organization are just as (if not more) important than the system you’re using as the foundation for your data operation. For an industry like Quick Service Restaurants this can present a secondary challenge, as the corporate players vary widely from one to another. The Elders:
  • Brands established in the 1950s and 1960s
  • Historically focused on development, market share, financial reporting
  • Typically large, sometimes multinational networks of corporate-owned and franchise locations
  • General lack of uniformity/compliance between locations regarding data structure and warehousing
  • Often set in their ways
Seasoned and Savvy:
  • Brands established after the 1960s
  • OR – Elder Brands that have kept pace with more contemporary technologies
  • Can be large, established networks or those on the verge of expansion
  • Emphasize uniformity/compliance, data structure, and reporting for more than just financial purposes
  • Have made or are willing to make investments of capital and human resources to stay technologically current
Young Upstarts:
  • Recent newcomers to the QSR category
  • Often “born into” technology
  • Latest-available technology is “the norm” available from the day they open
  • Have learned from Elder Brands and Seasoned Savvy Brands what challenges need to be overcome as well as opportunities to seize from industry
In the real world, effective analytics draws from all three types of business models – with team members who understand that an analytics infrastructure must be:
  • Quick
  • Economical
  • Flexible across both Company and Franchise restaurants
  • Mindful of the network’s ability to uniformly comply
  • Able to deliver actionable data for decision-making purposes

Driving Data Forward – A 5-Point Plan

Keeping in mind that a complete POS and data system overhaul can take as many as 5 to 10 years to bring fully online, a Common-Sense plan like Miller developed for the Church’s Chicken network is sound business. In just 1 to 3 years, with minimal capital investment, more insightful analytics can be obtained from an existing structure. As well, when this effort is conducted simultaneously with an overhaul, the transition can actually be quicker and more seamless. “It doesn’t matter whether you had a legacy system in the past, are presently working with one, or are actively searching for the next solution,” offers Miller. “An interim common-sense approach can and should be implemented in the meantime.” For this, she suggests a 5-Point Process: Step 1: Repair Foundation
  • Begin by setting system-wide standards that all franchises and locations follow
  • Use consistent naming, formatting, figures, etc.
  • Eliminate duplication and manual entries wherever possible
  • Ensure data integrity
  • Build solid databases
Step 2:  Build Infrastructure
  • Cleanse existing databases of errors
  • Eliminate unnecessary redundancies
  • Continue to apply consistent standards for data format
  • Optimize time, capital, and human resources
  • Make use of existing systems and/or make adjustments and modifications to increase yield or performance
  • Create changes on corporate/centralized level – avoid incremental roll outs
  • Automate whatever processes possible
  • Compile, process, and analyze data
Step 3: Improve Existing Reporting & Analysis
  • Stress quality of data – accuracy, timeliness, automation
  • Be smart about data quantity – know your parameters
  • Substance – is this data you need to know? How will it help you bridge knowledge gaps?
  • Depth & Breadth – do you have enough data to support decisions? Is it relevant to the entire network or just anecdotal?
Step 4:  Create a Comprehensive Analytics Toolbox
  • Start using data to identify strengths/weaknesses of promotions
  • Begin forecast modeling
  • Integrate marketing and supply-chain aspects
  • Recognize correlations between data sets and analyze sources
  • Notice the difference between business basics, shifts, cycles, and trends
  • Be mindful of efforts that cannibalize profits or top out at incremental improvement only
  • Identify true profitability drivers that can be leveraged time and time again
Step 5: Establish the New Norm
  • Approach all business operations from a more informed perspective
  • Experience a stronger competitive advantage among key customer targets
  • Leverage business analytics into bottom-line financial performance
  • Be better prepared to transition to a new system when the time comes
Most of all, remember that analytics is an ongoing activity. There will always be more to learn and investigate as the QSR and technology industries continue to evolve and shape each other’s futures. The businesses that learn to balance sound analytics practices and expensive hardware/software solutions will not only be able to maximize their investment into any infrastructure, but also potentially gain more powerful insights into the data they are mining… and that’s something that satisfies the appetites of customers, franchisees, investors, and shareholders alike.