How are the primary tools for People Analytics used day-to-day?
TL;DR: Knowing how to build and maintain data sets using SQL, analyze data using either Python or R and visualize data using Tableau will make any HR professional competitive for most People Analytics jobs.
In the last post, we established that Excel continues to form a foundation for many professionals working in the HR/People Analytics space. However, after surveying 22 recent job descriptions on Indeed.com, it was clear that the HR generalist who wanted to transition to more analytical work needed additional tools such as: a database query language (SQL), an object-oriented language (Python/R) and a tool for visualization (Tableau). There are, of course many more tools, including HRIS systems specific to each organization, but these four can make you competitive for most jobs posted.
But then the question becomes, what do people actually do with those tools on a day-to-day basis?
This post builds on the previous one where I try to provide some context for how the core four technologies mentioned previously are used in a People Analytics role.
To answer this question, I went through the same 22 job descriptions from the previous post and analyzed all day-to-day tasks and responsibilities for each position. I analyzed the job descriptions to highlight what I felt were distinct themes.
(Note: These are not mutually exclusive themes, but they provide a way to somewhat separate tasks requiring “hard” skills vs “soft” skills).
Here are the themes that came up and a tally of how many times each were mentioned across the 22 job descriptions:
Themes (Frequency)
- Collaborate with internal stakeholders (39)
- Analyze Data for trends, relationships, and prediction (36*)
- Inform Business Decisions (32)
- Inform HR Function(s) / People Decisions (30)
- Communicate to diverse stakeholders (30)
- Create Visualization(s) (dashboard / reporting / scorecard) (29*)
- Build and Maintain Dataset (27*)
- Provide analytical support (25)
- Manage complex projects and workflow (14)
- Train algorithms (maintain code base) (10*)
- Improve efficiency, automation of work processes (8)
- Exercise judgement and discretion (sensitive data) (5)
- Establish metrics / measurement4Capacity building of analytics team (2)
- Gather market data / trends (2)
- These are the themes that can shed light on how the core tools like SQL, Tableau, R/Python are used.
Build and Maintain data sets [SQL]
The People Analytics job candidate, much like data scientists, will be expected to build large, complex data sets from multiple sources, often containing both structured and unstructured data. They’ll gather data internally, cross-functionally (HRIS/HCM; other departments: operations, finance, sales etc.) as well as externally (market data, social media, web scraping, etc.).
This work is a collaborative effort with IT and the data architect, but the candidate is generally expected to be self-sufficient with the technical skills to restructure the data as needed.
Knowledge of SQL (structured query language) will go a long way towards allow you to collect, combine and/or restructure data from multiple sources. This process is broadly known as performing ETL (data extract, transform, load) for data warehousing and knowledge of SQL (MySQL) and comfort with CSV files are essential.
There’s much more to this responsibility than combining data sets. Once data sets are built, the candidate may be expected to lead/participate in data governance efforts to ensure data confidentiality, quality, consistency, integrity, validity, accuracy and promote general risk-awareness.
Analyze data for trends, relationships and prediction [Python / R]
This theme is very much connected to the previous one.
After a data warehouse is created, a candidate will be expected to query and retrieve data from the database and perform exploration, manipulation, wrangling, cleaning and analysis on the data to extract meaningful insights for the business and/or major HR functions (recruitment, retention, development, comps and benefits etc.)
A competitive candidate is expected to know when a business question requires reporting on data or building models for statistical analyses (even better the difference between statistics and analytics). The candidate may be mining data to form hypotheses or running statistical analyses to test hypotheses, all in the context of a strategic HR or business question.
Depending on the organization’s analytics maturity, the candidate could be shifting from reporting to predictive analytic modeling.
Beyond descriptive statistics, the candidate may be building models where techniques such as Chi-Square test, ANOVA, simple linear regression, logistic regression, decision trees, clustering, trend analysis and neural networks are used depending on the question.
For these tasks, the tools most often mentioned in various job descriptions are Python or R.
For organizations with very large data sets, the candidate will be expected to split the data set, where one is used to create models and the other portion to validate the model.
For advanced predictive analytics, the organization may have the analytics professional train algorithms to apply to new data sets.
All of the above can be achieved with Python and R.
Create visualizations [Tableau]
Arguably, none of the above matters unless that results can be communicated to a non-technical audience in a way that is business relevant and actionable.
Communication is a theme in and of itself. Creating compelling, easy-to-understand visualizations is a sub-set of communicating data results.
Visualization include Dashboards, Report and Scorecards all of which and required periodically to senior leadership and other stakeholders or, on an ad hoc basis.
The competitive candidate will produce recurring reports/dashboards with Tableau (or another visualization tool like Microsoft BI, R Shiny etc) and should be able to produce reports and dashboard in an ad hoc fashion as well.
The candidate will understand user-centered design, visualization and a flair for data-informed storytelling when creating visualizations.
Now you know how the most popular technologies are used in People Analytics. In future posts, we’ll get into even more specific details of what is involved, including both hard and soft skills needed and how to get them.
You can sign-up for regular updates here: http://getwyze.com/
