McDonald's uses JMP to explore the delicious world
-- using predictive analysis to better meet customer needs
on April 15, 1955, when Ray Kroc opened his first McDonald's restaurant, a hamburger was only 15 cents and a juice soda was 10 cents. He sold a total of $316.12 on the first day. In 1958, McDonald's sold the millionth hamburger! In 1959, McDonald's 100th restaurant opened. McDonald's launched "Big Mac" hamburger in 1968, three-tier beef hamburger in 1973, and mcnugget in 1975. I'm afraid everyone has heard that "McDonald's has served more than 99billion people in total"
today, McDonald's has more than 31000 restaurants in 118 countries around the world, with more than 58million people eating every day. However, in the past few decades, some voices have always accompanied the growth of McDonald's, such as "fried food is addictive" and so on. Over the past few decades, McDonald's menus have gradually included more complex and healthier taste requirements, and the sales revenue related to these new flavors is gradually rising. Who can successfully predict all this? McDonald's can
mike Cramer is one of the professionals who complete this great work inside McDonald's. As the director of McDonald's operation optimization, his job is to predict and monitor trends, identify and test new opportunities in any restaurant operation, and provide strategic advice to restaurant managers around the world on how to continuously improve customer service. Among the necessary tools he chose, JMP professional analysis software from SAS, the statistical discovery engine of excellent performance, is used every day
meet the increasing and personalized global market demand
McDonald's operations research team uses JMP to obtain historical data, establish statistical models, and predict future trends. And use JMP as a business reporting platform to provide strategic advice to global internal customers
McDonald's global operation optimization team has successfully provided local market trends for restaurants around the world, Improved McDonald's restaurant operation performance
how McDonald's uses JMP
"JMP helps us focus... Its analysis and presentation are lifelike and clear at a glance" -- Mike Cramer, director of McDonald's global operation optimization
JMP's graphic generator is one of Mike's favorite functions, because this function allows users to make almost the most beautiful charts with just a few clicks of the mouse Just like designing concept cars, Mike and his team will help McDonald's design and develop restaurants around the world. When designing, they must consider the current operating conditions and need to predict the future operating conditions
"graph generator" is the favorite function of Mike and his team, because it can be used to randomly, change and reorganize the analysis graph
McDonald's "operation platform" is used to record the whole customer experience of a customer from entering the parking lot to leaving the restaurant. Mike said, "this platform is actually our information system, a human-computer interaction platform for employee work planning, and all the content that connects customers and employees." In the past, decisions about restaurant operations, such as building design and employee training records, were made through individual decisions; Six years ago, Ken kozil, senior vice president of McDonald's worldwide, led a new innovative collaboration system to improve this decision-making model. Ken said: "McDonald's is built on a solid foundation, which is our core menu from all over the world. However, we hope that our restaurants will pay more attention to the needs and changes of local tastes, and reflect them on the local menu as soon as possible. In order to achieve this goal, we start from scratch, first write down our design points, and then focus on the changes in technical means and customer trends, such as us Determine the customer's order mode in today's technological environment and the possible changes in this mode in the near future. " McDonald's divides its global business into 10 markets, and Mike's department helps McDonald's restaurants in these 10 markets establish a transformation path to satisfy local culture. For example, in Japan, rigid cement pipes have another mechanical property, that is, impact performance testing. People use them to pay subway fees and street retail machine shopping fees. "About 4 or 5 years ago, we found this phenomenon and thought it would become a trend in the future. Now, this has indeed become the mainstream model. However, Japanese customers are not inclined to choose customized menus - special requirements such as additional pickles and no onions. Therefore, we have done little work in Japan in terms of customized needs, and mainly focused on grasping the changes in the trend of payment mode, Timely launch a service model that meets customer requirements and is forward-looking
Mike uses the design concept car to compare his operation platform. "You determine the design criteria, develop the concept car, and then use reverse engineering to implement it.". His operations research team is also trying to break the rules. Imprecision is a typical rule: for example, we mismatch the ginger juice with the Chicken McNuggets we ordered. "We strive to make our operation system more accurate, whether it is information system, monitoring system, or entity management system. Our three key points are: growth strategy, concept vehicle implementation method, and breaking the rules, which is our top priority."
"I search for clues and restore the truth according to the data, and then I can easily find the problem" -mike Cramer, director of McDonald's global operation optimization
mike uses the powerful graphic function provided by JMP to reveal the "delicious" details hidden behind the data. Does the colored ellipses in the picture reveal a secret
Mike and his team used JMP to divide the order volume data into 10 groups to study its similarity (as shown in the ellipse in the figure)
Mike said, "we analyze and mine all the data to the greatest extent". Using the visual analysis capabilities provided by JMP, Mike's team increasingly uses exploratory data analysis as an engine to promote the company's growth strategy. "In the past three years, I have been committed to helping our colleagues understand their data and discover the information and value behind their data". Gradually, our work has established an increasing popularity within the company, and company leaders in various regions have also begun to understand their business from a new perspective through data
tests before release
mike has been using JMP for more than 4 years. He was completely conquered by JMP's top visualization ability (i.e. graphics ability). "In the past, I have been trying to describe the information hidden behind the statistical results to my colleagues in the business department, but most of them don't understand statistics and can't understand what I show. JMP happens to be a bridge between statistical experts and business departments. When I first used JMP, I was very surprised to find how powerful and easy this software is. For non statistical professionals, JMP is a good data analysis assistant. Based on careful consideration, Mike also carefully evaluated the software before promoting JMP to all departments within McDonald's. In the evaluation group, there is a very professional statistician, an entry-level analyst, and a business person who does not know statistics at all. He assigned them various tasks of data exploration, analysis and display, and the results of everyone's analysis were satisfactory
the beauty of predictive analysis
operational research mainly includes three aspects: predictive models (mainly used to predict future operational conditions, test possible operational designs, and ensure the most realistic model in the future), rapid implementation (ethnology plus visual analysis of customer behavior in the store, to study how to improve customer experience), And prediction analysis (obtain historical data and current data, and predict the future according to the prediction model). "JMP undertakes our main prediction and analysis work. JMP makes it easy for our staff to obtain data. Through simple mouse click and drag, we can get the analysis results and charts that need a lot of programming and are difficult to achieve in traditional statistical software."
how to show the results
it's no surprise that even the display work can be achieved through JMP. Mike's goal is to replace all excel reports and PPT displays with JMP. Mike said that you can never foresee the problems of every customer, but when it comes to presentation, you can fully prepare in advance. Through JMP analysis and prediction, we can basically find the problems at the operation level and predict the possible conditions in the future. With such preparation, we can basically know what questions people will ask and how we should answer them. Traditional PPT can show a lot of content. The problem is that when the audience is inspired by your carefully prepared content and begins to ask more elaborate questions, you cannot interact, because PPT is dead. JMP is different. It can not only show all the content that PPT can show, but also interact on site, and introduce the audience's questions to a higher level step by step, which makes the audience refreshing: use the graphical process of data analysis and clear and understandable results on the spot to tell him where our past problems are, how to do better now, and what may happen in the future. Not long ago, Mike and senior executives in the U.S. market held a meeting to see whether the workmanship of the torque tester is an internal accessory of the solid torque tester equipment. At the meeting, they wanted to further understand how the new business has an impact on the overall performance growth. Mike clearly described this impact with various dimensional graphics attached to JMP's prediction and analysis model, It also showed the different distribution of this influence in different regional markets. As a result, Mike said, "they immediately reported their thoughts to the president and immediately took action based on my analysis"
mike has made the analysis of different categories into JMP analysis and display templates. His colleagues are more and more familiar with and like such standardized analysis and display templates. After opening the template, his colleagues can easily show the problems and values behind the data from several different analysis angles with just a few clicks of the mouse. "The action plan formulated based on these analysis results is helping Maidang save a lot of operating costs globally because steel mills are buying goods with a buying up rather than buying down mentality. This is a great success," Mike said. (end)
Copyright © 2011 JIN SHI