Term Project Alternative 1

Customer Service Survey Project

See also: Term Project Alternative 2, Case Analysis General Guidance.

1. Identify a company or activity that will let you work with their customers to conduct a customer survey and subsequently analyze the information. It may be bricks and mortar or an e-commerce business (activity). But you must have both placement and access to observe the business, survey the customers and employees, talk to and work with the managers.

2. You may work in groups (five-seven to a group) or individually. Each project is to cover a different company or activity (if the company is large enough, a different branch of the company). The important thing is that customer service is related to the point of contact with the customer. A McDonald's in Belgium will differ from one in Germany or the UK. The legal office at Baumholder will be different from the one in Heidelberg or Ramstein. All the organizational policies and training in the world will not change that reality. Hence, unless you are physically close to others classmates, or select an e-commerce business, group work will be limited.

3. Many firms have long standing programs of customer surveys and may not want you to interfere (or pollute) that effort, so be aware of the sensitivities.

NOTE: In the past I have found that financial institutions on U.S. bases and posts, the military exchanges and commissaries are difficult clients to work with. They operate under a number of constraints, have they own surveys, are tightly controlled and often have very limited time and resources to share with students.

4. If you are going to conduct a survey on a military base, CHECK with Base Legal, Public Affairs or the Services folks, to make sure you have the proper permissions beforehand. Upon request, I will provide a letter of introduction to the proper authorities. (Make sure you give me the correct fax number or email address, name, rank/grade, position, organization, and address of the recepient.)

5. Your project will have four distinct phases:

You will submit a formal report on both the project and the method of determining the suggested improvements; i.e., your formal process improvement report will have both content area and methodological sections.

Your report will be fully documented and presented in technical report writing style.

 

Phase 1: Know your client and the customer. Describe what exists at present in detail. Make sure you understand the system and processes within which customer service is supposed to occur. Map the customer service process using a Deployment Flowcharts or Swimlane Diagrams. Before your fully develop your survey for Phase 2, make sure you have refined and developed the details of your understanding. Then revise your Phase 1 submission and send it to me along with your Phase 2 input. (due 7 March 2003)	For the 10 Feb deadline, this understanding will admittedly be more limited that it must be in the end. I understand that; because of time constraints, I want to force the pace a little and get you out on the street. Your appreciation of the current situation will grow as you progress through the survey, so be satisfied at first with a general description of the firm, its general strategy, products and services, main competitive advantage in the marketplace, customer service policy and the perceived response of customers to the firm, its products and services. Update your Process Map at the end of Phase 2-4 as need be to reflect your improved understanding of the situation.
Phase 2: Develop your survey instrument related to business processes. Focus on the issues Anton raises. See his survey instrument (Anton, pg 86-87) for help in getting yours started. Before you develop even one question, you must identify the business processes that are directly related to the quality drivers that motivate the activity's customers. You must then identify the customer expected attributes related to each process. (See Anton, Fig 4-7 pg. 53.) You will notice that Anton begins but does not complete the example in Chapter 4 and 5 of his book. For example, he shows in Fig 4.7 only four of many customer expected attributes related to warranty claims. Each process will have 3-5 attributes much as each quality driver will have four or more processes underpinning its effectiveness. Thus for the four quality drivers which Anton proposes in Fig 4-3, you could find more than 20 business processes and more than 80 customer expected attributes, all arranged in a more complex and complete version of the root and branch diagram he shows in Fig 4-7. If a firm hopes to use the information from the survey, it has to understand its own internal operations and how they relate to the customer expected attributes. To help them do that you need to relate each attribute to a further 3-5 internal metrics (including customer complaints) such as mistakes on forms, misplaced files, rework on assembled products, absentism, tardiness, poor appearance, etc. Internal metrics must be observable, specific, measureable, and attributable.	Remember to caution the company that if the customer is asked her or his opinion on an issue, the firm and its managers have a responsibility to listen or the customers may rebel and stop trading with them. Unless you are going to enlist the aid of family and friends in the survey, keep your survey objectives under control. I suggest everyone conduct a telephone or curb side survey. Notice that Anton only asked 25 questions in his survey. Some will think that is a lot, but we need some information to work with. Even relatively small populations require hundreds of random surveys if we are going to have sufficient data to work with. If the company will help collect the information, then one set of problems will be replaced by another. If not, you will have to make a judgement on how many surveys you can complete. If you cannot reach the magic number for statistical validity and reliability for saying something about the population as a whole, we can at least describe the reactions of the sub-set of people surveyed. You will still be able to practice the process even if you only secure 30-50 completed surveys. Before you test your survey, send it to me please. Once I OK it, test it out and make changes that seem reasonable. Again, let me know the feedback and comments from the test subjects. Then you should be ready for Phase 3.
Phase 3: Physically conduct using random survey techniques the survey you have constructed. Begin your data analysis. Remember several important things: Don't force or bribe. No dead people taking your survey. No more than one survey from any one person. A telephone survey of customer's desire to visit a particular recreation park (for example) will not be random, for some customers may not have a listed telephone number or evan a telephone. Standing outside a store between 4pm and 8pm Friday nights and selecting some or even all of the people who walk by will not generate a random survey. Stick to random survey methods as and when you can. A word about random methods. Leaving the survey on the counter for anyone to take and fill out is not a random method. Stopping every 5th person walking out of the store is not a random method. Calling every 7th person in a list is not a random method. But, Using a table of random numbers to select time segments to conduct 100% interviews or random interviews is a random method (divide the opening hours each week into 15-30 minute segments, number each segment consecutively so each has an equal chance of being selected). Flipping a coin to select from two alternatives is a random method. Rolling a die to select for the same number of options as sides of the die is a random method. Using a table of random numbers to select names from an alpha roster or addresses from a street directory is random as long as every one in the target population has an equal chance of being selected. Systematically document the operational phase of your survey - number of people approached, number responding, times, dates, etc. and include this information in the metholodogy portion of your project report. (A simple log and tally sheet is often best for recording this information.) If you have help conducting the survey, include information on their training or orientation for conducting the survey as well as details of their activities. Include specifically the terms of reference and operational definitions to which they were instructed to adhere.	In an operational survey, you would not be satisfied with less than random survey methods, hitting the correct numbers, etc. We are here to learn the process. I know that each of you will try to reach the survey target number using random techniques. I know also how difficult it is to do just that - and that the scope of work required actually far exceeds the normal workload for a 3 sem hr course. So, . . . If you do not reach the magic number or you do not follow random selection methods, then make sure your client knows that the results of the survey cannot be used to describe the population from which the sample was taken. The results can be used to describe the sample group itself (those people who did take the survey). This can be valuable. If your client needs a proper random survey, perhaps he or she can help with people or other resources to make it happen. Analysis of the survey responses - and extrapolation to the larger population - is dependent upon a diligent application of random survey methods. If you have a math book of tables with a random number table in it, you are likely to have very nearly a set of random numbers, but that too is not certain. The algorithm for the random number generator in the old Apple II computer produced a list of numbers that appeared random - until someone ran a very long list of them and the sequence started repeating exactly. If you set up your spreadsheet function and generate a list of random numbers, you can produce a list yourself. Again, the algorithm may not be perfect, but it will be close enough to random for our purposes.
Phase 4: Analysis and Reporting First, use Anton to guide your analysis. We do not use SPSS in this course although if you have access to a licensed copy of SPSS, you are free to use it. Your class reports must however be turned-in in MS Office and Excel formats. I will pass out a template upon which you can build your analaysis and do everything that Anton does with SPSS albeit not as elegantly. You are to tie together your preliminary work (on business processes, attributes, internal metrics) and the qualitative results of the customer survey using the quantitative methods explained in Anton's book and recommend a course of action for the client. You are to prepare a technical report of your results and your methods (for an outline, see below). This will be a formal report of your research and results for the client and me of your activities, conclusions and recommendations based on that activity.	This is not a course in technical report writing so I do not propose to review with you here how to construct a project or research report properly. See your Technical Report Writing book, on-line guides, etc. If your client requests a separate report, please send me a copy of that report. Otherwise, provide your client with a copy of what you send me. (Out of courtesy, address the report to both of us.) If you have severe misgivings about the quality of your method, conclusions, or recommendations, you may delay providing your report to your client until after you receive my comments.
As a minimum your report should include: Title Page ToC Executive Summary Purpose Background Your Project activities Phase 1 summary Phase 2 summary Emphasis on why and how you made the choices you made. Population description and size Theoretical Sample size needed vs. Sample size used. Discussion of random survey methods required for reliability and validity vs. the methods you finally applied in the field. Copy of your final Fig 5-5 model (in appendix) Copy of your final Fig 6-4 model (in appendix) Copy of actual survey instrument(s) used (in appendix) Phase 3 activities in detail Survey details - when, where, how, how many. Describe your method of obtaining a random sample Justify any variation from random and explain the impact on your findings. Include sufficient detail so that another researcher, following your instructions, could obtain a random survey result. Total number of surveys attempted, number returned, number spoiled (mistakes in completion), number good surveys returned and used in your analysis. Analytical methods Explain how you analyzed the data (no, not "I plugged numbers into the spreadsheet Phil sent me.") Provide both raw and processed data (not the actual completed survey instruments, but the Excel database you created) Resources used vs. required Report of Analysis Fully explained tables in the form of Anton's Tables 7-4 and 7-5 (Anton, pg 118) Impact Performance charts of the form of Figures 8-1 thru Fig 8-4 (Anton, pgs 122-124) [This includes identification of statistically significant Impact variables and the decision matrix identifying which attributes are the most likely candidates for immediate attention, which can be monitored for the present, etc. If your survey is not random, still apply the techniques, but your results will not provide a basis for definitive management action. They may nonetheless inform future customer opinion research efforts.] Analysis of detailed responses, question by question as appropriate. Analysis of customer write-in comments. Where applicable, ANOVA and Chi-Square analysis of stratified survey results (tabulated by pre-established sub-groups, data which some surveys collected) Conclusions (substantive) What did you determine about the attitudes of the customers you surveyed? How did your findings differ from what you initially expected to find at the start of the project?For a random survey, generalize to the population. For non-random surveys, apply your findings to the sample group that completed the survey and suggest how those insights might be put to good use. Aspects of your Fig 5-5 analysis that yet need to be investigated New areas for investigation arising from the survey results. Management practices that, if modified, might result in an improvement in customer attitudes. (See Anton pg 125-126 for example.) Internal Metrics to monitor, or create and monitor. Need for further customer opinion surveys. Conclusions (procedural) Lessons learned for future customer survey efforts concerning Survey Planning Linkage of quality drivers to processes to attributes to metrics Continual assessment requirements Periodic survey requirements Resource requirements (used vs. required) Survey Conduct Use of selected vs. random samples Improved methods for random sampling Methods that would make analysis better, simpler, or faster Survey Analysis Suggestions for improvement Qualitative vs. Quantitative methods Specific comments on Anton's methods Recommendations For use of your survey data and results For management action to improve customer perceptions For future customer surveys Appendices as noted above Works Cited or Bibliography Added items as you see fit. Provide professional credit to the works of others throughout your report. This is not a cast-iron outline. It does describe the main elements I expect to see in your reports. You may use any approved writing guide.

See the assignment schedule for key dates. Plan and work ahead, please.

Document your efforts. Your report must include an appendix that details your methodology.

Group projects will normally be larger in scope than individual projects; hence you may need to begin your field work early.

Unless otherwise excused, use the MS Office suite of programs to prepare your report. Diagrams made in programs such as Visio or Inspiration must be converted to .gif or .png files before inclusion in your report. (Use .png or .jpg for complex color pictures.) For ease of posting to WT or e-mailing to me, you may need to use several different files. If you do, make sure they are well referenced internally so that it is absolutely clear how the whole project comes together and interrelates.

Use as many of the tools and strategies that you already know or will learn in this course as possible. However, the intent of this customer service project is for you to demonstrate what you will have learned from the concepts and techniques presented in this course. I want to see diagrams and charts, not just words. Note in Anton how words are used to amplify, tables, charts and diagrams - and diagrams are used to capture and communicate so much more than simple words can do.

Where sources other than your texts are used, please include a complete bibliographic reference and final summary of works cited. I should be able to find any quote or paraphrase with relative ease in the original source. Just as importantly, unless the idea, details, graphics, etc. are uniquely your own (and presented for the first time in this report) or in that body of knowledge termed 'general' by a non-specialist, make sure you cite your sources. If you use some else's idea, exact words, graphics, etc., give them credit. If you use a paper you or someone else has written for a prior class or purpose, include it in your references. Be careful. Do not unintentionally plagiarize what has been called 'intellectual property.'