Long gone are the days of ad-hoc analysis and basic reporting. Data is more and more plentiful, available and leveraged by a melange of analytical software. As such, data analysts are no longer historians, picking through past data to understand it. Rather, they’re problem solvers, forecasters. It’s a strategic discipline, not simply a support role. So, the expertise we look for needs to reflect this. In this article, we’ll take a look at the key skills of a data analyst, and how to differentiate between ‘good’ and ‘great’ expertise.
To read about general data team building tips, read our article here.
—
The skill set is broadening
The role of a data analyst was in some ways more well defined previously. Where there are fewer possibilities, there are fewer expectations. Now, by the proliferation of tools and expansion of expectation, it blends increasingly with that of a data scientist. Methodologies which were previously used by specialist mathematicians, statisticians and computer scientists are now increasingly accessible. It’s now essential to have expertise in one or more programming languages, as well as database querying methodologies, both of which we’ll explore in more detail.
It is however important to not totally neglect tried and tested tools in the pursuit of shiny new ones. Excel is still widely used by data analysts for basic data analysis. Similarly, SAS is still relevant in certain contexts like healthcare, finance and government sectors. This is particularly true for long standing organisations who have built their data infrastructure, models, and processes around it. Data analysts with expertise in SAS are valuable for organisations who work with large and complex data sets in industries with strict regulatory requirements. However, for smaller, less established organisations in other contexts, it may be less important.
—
R and Python
These languages are consistently used for their flexibility, extensive community support, and integration with emerging technologies like machine learning and big data frameworks.
The explosion of digital data has transformed the landscape. Data analysts now work with huge datasets generated from various sources, including social media, sensors, and online transactions. This requires proficiency in R and Python, as well as knowing how to utilise DBT (data building tool). More on the latter below.
SQL
Data analysts work with databases, so they absolutely need familiarity with writing SQL queries. The specific dialect depends on the setting, typically for most contexts MySQL will be suitable, but there are a few other common tools worth mentioning, as listed below:
- PostgreSQL: Another open-source tool known for its advanced features, extensibility, and strong community support.
- Oracle SQL: Used with the Oracle Database.
- Microsoft SQL Server: A database management system developed by Microsoft, commonly used in Windows-based environments.
Additionally, data analysts can benefit from expertise with SQL-related technologies and frameworks, such as:
- SQLAlchemy: An Object-Relational Mapping (ORM) library for Python that provides a high-level interface to SQL databases.
- Apache Spark SQL: A module in the Apache Spark ecosystem that allows querying structured and semi-structured data using SQL-like syntax.
- HiveQL: A SQL-like query language used with Apache Hive, which provides a data warehousing infrastructure on top of Hadoop.
Data Visualisation and Storytelling tools
The ability to communicate insights effectively has become crucial. Data analysts now employ interactive visualisations and storytelling techniques to present complex findings to non-technical stakeholders. Organisations have embraced self-service analytics platforms like Tableau and Power BI to do so. These tools allow data analysts to create interactive dashboards and reports, empowering business users to explore and analyse data independently.
Data management
Working with databases is one thing, but data analysts need more than SQL. Data analysts now handle more complex and diverse datasets, including structured, semi-structured, and unstructured data. It helps if they’re skilled in data manipulation and management techniques to clean, transform, and integrate data from various sources. This includes knowledge of data extraction, transformation, and loading (ETL) processes, database management systems, and data governance principles.
It is important to note that in the past few years there has been a significant shift from ETL to ELT processes. A Data Building Tool (DBT) is something any competent data analyst should know their way around. The former allows the user to take the code, compile it to SQL, and then run it against your database. It is all done without having to resort to old ETL tools, which is why such expertise is a must these days.
Leveraging technical expertise
It’s a trend within the data analytics profession to present a long list of technical skills. We’ve elucidated some of the core technical competencies, but technical expertise must be coupled with subject matter knowledge, domain area expertise and business acumen.
Advanced Data Analysis
Proficiency in statistical analysis is crucial for data analysts. They should have a good understanding of statistical concepts, hypothesis testing, regression analysis, and other methods to draw accurate conclusions from data.
Why? Because this builds intuition. Using pre-built statistical functions and libraries is helpful, but it’s important to understand the limitations of the tools and how to accurately interpret the results and qualify the outputs.
As an example, suppose a data analyst is asked to optimise inventory levels based on demand patterns. While R provides various forecasting models, understanding statistical theory would allow them to adapt and extend these models to incorporate additional factors like seasonality or trend. They could potentially develop custom algorithms or hybrid models that integrate multiple forecasting techniques, ensuring accurate inventory planning and minimising costs.
Domain Expertise and business acumen
Understanding the specific terminology, data sources, and business processes allows analysts to provide more contextually relevant insights. They need to bridge the gap between technical analysis and business objectives, aligning their work with the organisation’s goals and making data-driven recommendations.
—
Summary
The role of a data analyst has transitioned from a primarily descriptive and manual task to a more strategic and value-driven position. As data continues to play a central role in decision-making, the role of data analysts will continue to evolve, requiring a blend of technical expertise, theoretical understanding and commercial awareness.