Using data to analyze and tell a company's story

Use this as a guide to help you think critically about the game you played and use the resulting database to help analyze data to support your questions and/or help you discover patterns in the data to help you derive appropriate questions that help tell your company's story.

Telling a Story With Data

Telling a story about any company involves the following things:

1. Come up with a meaningful question related to your company's game performance.

i) A question or a statement about your company's game performance should involve Inputs and Outputs. For example, "Did our company's advertising strategy combined with our pricing of products achieve competitive sales with a high margin? "

ii) If you cannot come up with a question you may want to use the Data-Driven approach described below to derive one.

2. Show data in a visual manner that supports your claim. -- develop a pivot chart or graph to help visualize the result. Label inputs and outputs in the graph appropriately. The graph should be easily interpreted without additional explanation.

3. Explain--either in person in front of a group of stakeholders, or in words that accompany the graph--how exactly the graph backs up your declaration. Interpret the results - make sure the value of this result to the company is revealed. If there is no value of the analyzed result to the company then you should start with a different question.

Data-Driven Approach

In this approach, you jump into the deep end, and create a graph first, even before you know what your declaration will be. You hope that the graph you create will have an obvious trend, spike, difference, or other interesting tidbit that you can then convert to a statement.

This approach can make things really easy and straightforward. if there is an obvious tale to tell in the data!
It also may require some trial-and-error in creating several graphs before you hit upon something you can use. For example, if you are trying to tell a story about Marketing, you may need to create graphs for each Product, or each Region, before you notice something interesting in the data.

1. Select one or more Inputs. For example, Price (either for a single product or all products).

2. Create a graph that shows your company's

3. Identify a pattern in the data. Typically, you make a chart and notice major spikes or valleys in the graph, or areas where your company's line deviates from those of competitors.

4. If you cannot identify a pattern or a spike/valley, then redo the graph, but this time, add in the same data for your competitors.

5. If you still cannot identify something noteworthy, then go back to step #1 above and retry with a different Input.

6. Once you find something interesting, think about what the impact of that trend or difference might be. For example, if you notice that prices for all products started low for all companies, and crept up steadily over the run of the simulation, you might think that might have had an impact on Sales.

7. Based on your "hunch" in Step 6, pick one or more Outputs.

8. Graph the Output(s) you have selected over time in the same manner as you did in Step #3.

9. Ideally, if your hunch is correct, you should notice a trend/difference in your Output closely matching or following the trend/difference in your Inputs. For example, if your theory is that your company sold more when its prices were low than when they were high, you should notice that the Output of Sales was high when the Input of your Price was low, and vice versa.

Theory-Driven Approach

The data-driven approach can be useful, but it can also be frustrating: it's possible to create a dozen graphs without noticing a trend you feel you can use.

With the "theory-driven" approach, you have a "gut feel" or have an observation based on having actually played the game. Perhaps you feel that pricing your product high for five rounds contributed to losing revenue to competitors. Maybe you noticed during the game that spending more than "5" in marketing, it didn't make a difference, and you want to compare your marketing spend to your competitors (the theory here would be that no other team spent more than 4).

This is the preferred approach, but it is only possible if you start out with a hunch. From there, it's fairly straightforward:

1. Get a hunch or a feeling. This is typically your statement/declaration, or close to it, and must involve one or more Inputs and one or more Outputs.

2. Graph the Input(s) over time. Usually this will be not just for your company but your competitors as well.

3. Graph the Output(s) over time. This might be just your company, or all companies.

4. Hopefully, if your hunch is correct, you will notice the trend/difference.

Attachment:- assignment.rar