The first step in the data analysis of any data is to identify the required work outcome.
Start by asking yourself, what business problem am I trying to solve, or what does the company want to know? The client might ask, 'Why are we losing customers?' but investigating a more specific issue, difficulty, contradiction, or gap in knowledge may be more useful.
So, instead of asking, 'Why are we losing customers?' you may need to reframe the question to 'Which factors are negatively affecting the customer experience?' so you can hypothesise.
Objectives may be described in a problem statement.
Note: A problem statement is a short statement explaining the need and reason for your research.
If the objective is ambiguous, ask a manager to identify their goals.
Alternatively, look for work outcomes in your job brief or communication from the client.
Work outcomes may relate to:
- what the business wants to know
- when the analysis is required
- the format and length of the report
Key Performance Indicators (KPI) may identify the expected Return on Investment (ROI) from your analysis.
Examples of KPI are available here.
Typically, data analysts are employed to achieve at least one of four work outcomes.
There is no agreed method to prioritise, but stakeholder involvement is critical. The approach might also involve policy decisions about statistical work outcomes (which data sets to interpret and analyse) and operational decisions regarding implementation (methodology and reporting).
Ask how and when your analyses will be used to confirm the priorities. Ask a manager to identify the most urgent request if competing demands exist.
