[Data science project title]
Become less tool-focused and more impact-driven. Summarize a data science project in 10 minutes with the sections below.
Addresses a specific problem that links to a strategic goal/mission/vision
- "Enable data-driven marketing to get ahead of competitors"
- "Automate fraud detection for affiliate programs to make marketing focusing on core tasks"
- "Build automated monthly demand forecast to safeguard company expansion".
- Build x to solve/achieve/improve y
- Create x to become data-driven
- Daily usage of reports y
[TODO: Define objective]
Lists measurable outcomes that mark progress towards achieving the objective
- "80% of marketing team use a dashboard daily"
- "Cover 75% of affiliate fraud compared to previous 3 month average"
- "Cut 'out-of-stock' warnings by 50%, compared to previous year average"
- Increase/reduce/improve X by y%
- Increase/reduce x% on y
- Predict X with accuracy Y
[TODO: List key results]
Describes properties the ideal or available dataset
- "Transaction-level data of the last 2 years with details, such as timestamp, ip and user agent"
- "Product-level sales including metadata, such as location, store details, receipt id or customer id"
[TODO: Sketch data properties]
Puts the project into a business perspective by visualizing value, feasibility and uncertainties around it. The larger the range of points, the more certain is the project's value or feasibility.
Estimated uncertain value from 2-7/10
Estimated low feasibility from 1-5/10
[TODO: Estimate value-feasibility]
Sketches follow-up projects and puts the project into a broader perspective.
- apply sales prediction to marketing campaigns features
- use cloud architecture as blueprint for sales prediction pipeline
[TODO: Map possibilities]