Project Overview:

The goal of this project was to identify the top 3 candies for Halloween based on specific criteria, such as a combination of features, winpercent, sugarpercent range, and price-to-win ratio. The final result provided the best combination of candies that would appeal to a diverse audience while maximizing value.

Criteria for Selection:

1. At least 3 Features Per Candy – To ensure diversity in candy characteristics, each candy must have at least 3 features (e.g., chocolate, fruity, caramel, etc.).
2. All 9 Features Present – The selected 3-candy combination had to cover all 9 unique candy features.
3. High Winpercent – The combination had to have an average winpercent higher than 60% to guarantee popularity among consumers.
4. Wide Sugarpercent Range – A wide range of sugarpercent ensured that the combination would cater to a variety of tastes, from very sweet to moderately sweet candies.
5. Low Price-to-Win Ratio – The final selection had to be cost-effective, providing good value for money.

Step-by-Step Process:

1. Generating Unique Candy Combinations (Python)
   • Tool: Python
   • Objective:
     • The first task was to create all possible unique 3-candy combinations, ensuring that each candy in the combination had at least 3 features. Python was used to automate this process.
   • Approach:
     • A filtering process in Python helped ensure that only candies with at least 3 features were included in the combinations.
2. Filtering Combinations with All 9 Features (Tableau)
   • Tool: Tableau
   • Objective:
     • From the combinations created in the first step, only those combinations that collectively covered all 9 candy features were retained.
   • Approach:
     • Tableau bubble chart was used to filter out the combinations and retain only those that included all 9 features.
3. Filtering by Winpercent (Tableau)
   • Tool: Tableau
   • Objective:
     • The next step was to narrow down the combinations to those with an average winpercent greater than 60%.
   • Approach:
     • A Histogram was used to visualize winpercent distribution, only combinations with an average winpercent 60% or above were selected.
4. Selecting Combinations with the Widest Sugarpercent Range (Tableau)
   • Tool: Tableau
   • Objective:
     • To ensure that the final combination appealed to a broad range of tastes, the widest sugarpercent range was selected from the remaining combinations.
   • Approach:
     • Tableau was used to sort and visualize the combinations based on their sugarpercent range, selecting the one with the widest range.
5. Selecting the Most Price-Efficient Combination (Tableau)
   • Tool: Tableau
   • Objective:
     • The combination with the lowest price-to-win ratio was chosen to ensure value for money.
   • Approach:
     • Tableau was used to create scatter plot and rank remaining combinations based on price-to-win ratio, the combination with the lowest ratio was selected.

Results:

• Top 3 Candies: Kit Kat, Nerds, Snickers
  • 3 Candies average Win Percentage: 69.60%
  • Sugar Percent Range: 31.30% to 53.50% (Wide range to appeal to diverse tastes)
  • Price-to-Win Ratio: 0.007 (good value for money)

Tools Summary:

Python: Used only for generating unique combinations of candies with at least 3 features.

Tableau: Utilized for all other stages, including filtering combinations, applying the winpercent threshold, selecting the widest sugarpercent range, and choosing the most price-efficient combination.

Click here for interactive dashboard ->

https://public.tableau.com/views/HalloweenBusinessPlan/HalloweenBusinessPlan?:language=en-US&:sid=&:redirect=auth&:display_count=n&:origin=viz_share_link