4 Key Steps to Build Impactful Visualisations for HR analytics
Whilst people may tell you to not judge a book by its cover, let me tell you something: in HR analytics, looks do matter. As the human brain is wired to process visual information more effectively, visualisations can make a huge impact on how people perceive and interpret data, and businesses are gradually waking up to the immense potential of visual forms of data.
Visualisations are not just about charts and graphs. A sophisticated dashboard with powerful analytics tools and interactive features can be your best bet in helping stakeholders dive deep into seemingly esoteric employee data, even with limited technical know-how.
Imagine a massive reservoir of HR data, with lists of numbers, feedback, and survey responses over a long period of time, can span over pages in excels, and become a messy data glut that few can comprehend.
This is where visualisations come in, summarising a deluge of data into succinct, interactive illustrations. With the pool of data ever expanding, their importance will only increase.
The one-million-dollar question is: How to maximise the potential of visualisations? How to create dashboards that are not only aesthetically pleasing, but also informative and intuitive?
Don’t worry — as a data visualisation developer, I struggled with the same questions. Below are a few tips to create a powerful dashboard I have consolidated for you.
1 | Make your problem, objective and goal as clear as possible.
Dashboard and visualisations are usually built around solving a particular problem or business need. Hence, it is important to define your problem well. Who is the audience? What is the problem? How big is it in scope? What is your objective?
Knowing where you are going and how you would like to get there will allow you to map out the best route, and in this case — the most suitable type of dashboard.
For a top-level CEO who wants an overview of the headcount, a headcount split dashboard showing multiple charts with headcount split by different features such as location or organisations might be adequate.
For an HR manager looking to solve a specific problem, such as rising attrition, or the efficiency of the recruitment process, we may need to look further for more inputs from your stakeholders. Prepare to have conversations with:
1. the client(s) about their business needs and what metrics they want to see,
2. your visualization team members, who brainstorm ideas and metrics of their own and propose them to the client(s). More often than not, having worked with the data, the DataViz team will have useful input towards solving the client’s needs that even the client might not have considered yet, and
3. industry people, to explore referencing industry standard metrics for reporting on similar problems.
2 | Define your metrics accurately.
Whilst this step can be swept under Step 1, it is important that we do not only treat metrics as mere benchmarks for measurement. The dynamics and interactions of metrics really matter. You first need to choose the right metrics — what you want to measure, and then think about the relationship between the metrics you choose — how they interact with each other, what you can develop from that and the influence of this dynamics on your final object.
Let’s take the attrition rate as an example. A typical process for deciding on an attrition overview chart might look like this:
We don’t just want a simple number of leavers, but rather a percentage of leavers against headcount. Also, we don’t necessarily just want a block figure in a static period, as that doesn’t consider the seasonality of attrition, so we want to have a chart of attrition over time. We also want to get some idea of who the leavers are, i.e. are these new hires, medium-tenure hires, or people with longer tenure who are leaving the company?
So, let’s have a chart of attrition over time, where for every month we show three values, one for attrition due to leavers with less than 3-month tenure, for leavers with less than a year of tenure, and for total leavers. We can then use the same process to see how attrition for all three values differs by department, geographic location, job grade, gender, and so much more.
Paying attention to metrics dynamics can help you choose and connect your metrics in the most accurate way, and this will result in precise analytics that can help you gain insights into your company’s performance.
3 | Build interactivity into your dashboard, and choose the right visual pattern.
Interactivity is critical — it puts the power of analytics into the hands of managers, and makes data actionable. While this is technically challenging, it is incredibly rewarding for the clients to be able to drill down and slice and dice into the different subgroups under the population, to look out for areas of concern, as well as recognise the groups where the Key Performance Indicators are good (meaning that the policies adopted in that group have been successful) — all by themselves.
While building interactivity, the guiding policy should be to give the clients as many options as possible, with the only constraints being the intuitiveness of the interaction and neatness of the presentation. For instance, it may be useful in some situations for the client to have the capability to apply individual filters to each chart on a dashboard. However, that might lead to too much confusion and clutter on the dashboard, and you might decide to only apply filters globally for the entire dashboard. These decisions involve both the UX and the Business Requirements teams to sit down and decide what the optimal trade-off for the client would be.
4 | Implementation
Finally, it is time to do the dirty work of actually implementing the dashboard. Although the technology space has a deluge of libraries and tools attempting to ease this task for the web, finding the right fit in terms of design quality, code stability, and flexibility to accommodate all of the user’s needs is no easy feat. You don’t want to end up having to compromise on any of these, just because you are limited by the tools. In a future article, we’ll provide an overview of the tools in the market and explain how we tackled the technical problems and the lessons we learnt.
I hope these tips will help you make your dashboards more impactful. However, if you feel a bit lost with creating dashboards for HR use, fret not.
My team and I decided to build Panalyt — an easy-to-use engine that integrates HR data from any source and puts the resulting analytics into intuitive, aesthetic and informative dashboards for managers those who are yet to be familiar with data-driven analytics.
Instead of having headaches over creating different dashboards and starting anew every time you have a project, now you can create your own data visualisations with just a few toggles and clicks. All highly actionable HR data analytics — attrition, gender ratio, key performers, recruitment insights, benchmark/ benchmarking are put into illustrations that are comprehensive, allowing you to obtain trend forecasts that put you at the forefront of the talent market.
Panalyt – we combine beautiful visualisations with powerful analytics, and give you the best of both worlds.
Note: This post was originally written by Mohammad Sharique Zaman, Front-End Developer at Panalyt