McDonald's Case Study - Determining Sampling Size

McDonald's Case Study: Determining Sampling Size

McDonald's from the beginning has been committed to doing the right thing. They believe that the customer comes first and their experience will bring them back. They have a global strategy called Plan to Win, which centers on an exceptional customer experience including: People, Products, Place, Price and Promotion as well as being committed to continuously improving, enhancing the customers experience and giving back to the community. So when it comes to finding out ways to improve upon the restaurant or menu, they speak with the people who matter most the customer.

Today McDonald's is one of the world's leading global fast-food restaurants with more than 33,500 locations and serving nearly 68 million customers in 119 countries each day (McDonald's, 2012). But let's base it on the McDonald's Case Study at 32,000 stores in 117 countries and they wanted to find out where they rank according to fast food excellence. Let's say that they want to survey 1,250 of their restaurants worldwide out of the 33,500, including the usage of at least four locations in each of the 117 countries where they are located. What needs to be determined is if the 1,250 gives them the 95% confidence level of covering 4 stores from 117 countries representing their 32,000 stores. In determining this we need to figure out (n) the required survey size. In solving (n) we use the following formula: n=ơ2z2/D2 where ơ is standard deviation (figured by 117*4/6)=78 stores, z=1.96 from table for 95% CL) and D=5 so n=(78*1.96/.05)2 or n=935 stores for a 95% confidence rate sampling. Since McDonalds plans on surveying 1250 stores they will cover the necessary confidence rate to ensure they survey at least 4 stores from each of the 117 countries or specifically 468 stores with the 95% confidence rate.

"Stratified sampling uses a two-step process, producing a probability rather than a convenience or judgment sample (Malhotra, 2012, p. 348)." Stratified sampling would be a good direction for McDonald's in the 117 foreign countries. The target population would be focused on the family unit ranging from 16 years of age to 70 years of age. In the target population let's select a geographic area around the 468 fast-food chains to have the surveys completed. By setting up sections of the population it will show ranges in groups; such as age and male or female results. This will allow McDonald's to see the variations in both sides to see how they rate in quality of food, service, cleanliness and value. This will allow McDonald's to see the areas they are successful in and the areas that could use improvement.

For a survey company they would have to figure what characteristics should be used to subdivide the population, how many strata should be constructed and how many observations should be taken in east stratum. The company would have to figure out the population around each McDonald's to determine the proportional allocation to survey. The other issue would be if McDonald's gave a fixed budget for the survey, this would help in determining the size.

The questions that would need to be addressed are quality of food, service, cleanliness and value. Setting up these questions would be to the point and would not be leading. It would have to be set for each country to where it would not mislead or upset the different cultures. McDonald's would be smart to use field work agencies. This would lower costs of employment needs, training and could assist with the agency having employees already set in the countries that McDonald's would need. Field-workers would be placed in the geographic area to speak to families. The questions would be to the point; have you ever visited a McDonald's? Some would have factors on what do you consider important when visiting McDonald's, and would outline product, price, food quality, coupons, nutrition, waiting time, service, convenience, cleanliness, and value. They would be able to rate these from 1 being the lowest to 5 being the highest. This would allow the participant to have a choice for each one instead of choosing just one. By having an agency handle should ensure that the data has been prepared in a comparable manner across countries or culture units and would increase the precision of data collected.

With the survey there would be numbers, values or responses to the questions allowing for additions or changes. Considering that maybe they would want to find out how many people pay with cash, check, debit card or other and now they