How to get started with Workplace analytics

Implementing workplace analytics is a process that requires careful planning and consideration. In our last blog about Workplace analytics, we discussed what it is, and showed a 6-step process for getting started with workplace analytics. In this blog, we will go through this process step-by-step:

Identify the workplace strategy and business drivers
Determine the measurement variables
Select the right measurement devices
Collect data and ensure data security and privacy
Analyse the data
Turn insights into actions

1. Identify the workplace strategy and business drivers

The start-off point and the goal of improving workplaces can differ immensely. Human resource managers and facility managers both deal with the same workplace but they look at it from quite different perspectives. The biggest split is between physical workplaces and people. Improving organisational performance through changing a physical workplace is a largely different process than starting off with the change of behaviour of the people who use it. But they’re both angles that are completely justifiable. Nevertheless, it requires different data, data collection, and has different outcomes. So at this stage, it’s already smart to take into account what you’re able to change and try to reach with the workplace analytics program.

Hard data vs. soft data

The split between a focus on the workplace and people strongly connects to hard data or soft data. In this other blog, we dive deep into the differences between hard data and soft data. But it basically means: Hard data is objective data and observable data (e.g. occupancy, room temperature, or wall colour) while soft data is subjective experiential data (e.g. workplace experience, perceived productivity, or happiness rating). Measuring the experience of people or the objective (use of) workplaces is quite a different start-off point, and leads to very different findings and actionables. So, determining the measurement variables is the next step in the process that will support you in the process.

2. Determine the measurement variables

As discussed, the goals of managing and improving workplaces can widely vary. For example, you can have the goal to reduce expenses, improve well-being, or improve work processes. Because of this, the data you want to collect and analyse varies widely. If you deal with data, you deal with variables. Variables are changing values collected by quantitative measurements through e.g. a sensor or survey. The table below shows several potential variables depending on different goals.

However, don’t forget that working on one variable might change the outcomes of another variable. For example, if you are trying to reduce expenses and cut in office space, this will affect well-being and work processes too. So one might decide to also analyse how workplace experience or productivity is affected when cutting in office space. Moreover, there are also other goals (like hygiene, safety, privacy, or design) that affect these goals and also have their own measurement variables attached to them.

Goal

Reduce expenses

Well-being

Work processes

Variable

Utilisation

Occupancy

Asset use

Energy/real estate expenses

Workplace experience

Air quality

Noise

Greenery

Space purpose and activities fit

Perceived productivity

Occupancy

Table 1. Examples of measurement variables based on optimising different workplace goals

What is Workplace Analytics actually?

Find out here

3. Select the right measurement devices

Through several different methods you can collect data. However, your personal goals strongly lie with the measurement methods you should take. In our previous blog, we included an extensive list of different methods but here we’ll discuss them shortly with different goals in mind. If you prefer to intensively deal with people, you can perform a more qualitative data approach using interviews or focus groups. The interviewer gets an in depth insight into the stories (of some) employees, but when it comes to data analysis, it’s not so useful.

From interviews, you can make visual word clouds but much more than that is difficult. One step toward a more data-driven approach where you can collect everyone’s (subjective) thoughts happens through surveys or experience sampling. Here you can collect data from a wide range of people and analyse it easily by stacking up all the responses, and collecting clean data. If you lean more towards hard (workplace) data, you can decide to go towards observations, sensors, or people counters.

Method	Output	Type of data	Data visualisation	Advantages	Disadvantages
Interview	Individual elaborate thoughts, needs and satisfaction	Soft data (Textual transcripts)	Word clouds	Deep insights	Very small and selective reach
Focus group	Group thoughts, needs and satisfaction	Soft data (Textual transcripts)	Word clouds	Discussion/group interaction	Small reach and potential bias through selection and loudest contributors
Survey	Overall thoughts, needs and satisfaction	Soft data (quantitative variables)	Graphs	Individual thoughts over many people	Takes a long time to submit, and depended on quality of questions and analysis
Experience sampling	Experience and activities	Soft data and Hard data (quantitative variables)	Graphs	Mix of hard and soft data, in the moment collection	Requires repetitive motivation of users
Observation	Activities, occupancy, use	Hard data (quantitative variables)	Graphs	Actual objective data of people and space use	Measuring in selective timeframes
Sensors	Occupancy climate,	Hard data (quantitative variables)	Graphs	Continuous data of actual objective variables	Relative lack of context in data

Table 2. Different data collection methods and their specifications, visualisations, and (dis)advantages.

4. Collect data and ensure data security and privacy

For all different methods, you can theoretically collect data yourself. For example, you can send out surveys using Google forms or buy sensors online. However, when it comes to soft data; it’s very important to ask the right questions. The data you collect should not be biased, as there are many ways to ask questions, and the questions should be posed in such a way that they are easily analysed. For hard data, one also needs to use the right methods to get accurate results, and one needs proper ways to analyse the gathered data which could be quite a challenge.

Hiring organisations to measure for you

Of course, there are many different organisations, including us, happy to support you in your journey of measuring and analysing your data. One advantage of these services is that they provide tools and people to collect data properly, ensure data security and privacy, and have sufficient tools for analysing and displaying the data. Their standardised methods make it often also possible to benchmark your own data with other (similar) organisations. This could be interesting to understand how your workplace is doing compared to others. But of course, services come at a certain price.

5. Analyse the data

The collected data, stored in (large) databases with columns, rows and thousands of numbers, need to be transformed into attractive visualisations. For soft data from interviews or surveys, the only common visualisation method is a “word cloud”. In soft survey data, the answers are often grouped into categories and displayed in graphs just with much hard data collected through observations or sensors. The picture below shows several examples of visualisations with different types of data.

Figure 1. Observation studies (observed activities)

Figure 2. Sensor studies (highly detailed)

Figure 3. Experience sampling (perceived in-time disturbances)

Circle diagrams, bar graphs, and line graphs are all basic visual representations that give a clear picture of the data. Getting to more advanced workplace analytics, there are several options available. The first one is with the use of filters and selection methods. Some services (like ours) offer interactive dashboards where you can select parts of the dataset that can hugely affect your insights. Depending on the type of data, you can filter on age group, department, day of the week, and your graphs will be giving you completely different insights.

For example, in one study we saw that the average workplace satisfaction was good (8.3 on a scale of 1-10). But when breaking it down, we saw a 9.0 for office ergonomics and a 7.0 for IT. When breaking IT down into age groups, we found lower satisfaction for younger employees (6.8) compared to older employees (7.2). So this detailed analysis might bring an action when it comes to IT that could improve the retention of young employees even though the average score was quite good.

Advanced workplace analytics strategy

Another more advanced workplace analytics strategy is just to have more advanced graphs. The risk of these complex graphs is the learning curve which can lead to the wrong interpretation of these graphs. But a good example of a clear, but advanced graph is the graph below. It shows how many and which activities (the size of the bubbles) are performed in which location (office, home or elsewhere). While the colour of the bubble represents the perceived performance of the employees in this combination. Furthermore, while hovering over the bubbles you can see the exact scores directly. One insight you can get from this picture is the low performance of digital meetings at the office and elsewhere compared to home.

Figure 4. Bubble graph displaying location (y-axis) and activity (x-axis), the number of occurrences of this combination (size of the bubble), and perceived performance (colour of the bubbles) for this combination.

One thing to realise when analysing the data, however, is the sample you have collected. With survey data, for example, there are always a few people who haven’t submitted their answers. While with observation data the data is only collected for one or two weeks. This is what they mean by the “validity of data” and the importance of missing data. This is good to take into account when analysing yours.

6. Turn insights into actions

It’s easy to get lost in the data by diving deeper and deeper by going to sub-dashboards and using filter after filter. These insights need to be transformed into actions. The good thing about data is that it covers what’s really going on. So it is easy to convince other decision-makers into making a strategy based on these numbers. More detailed insights (through the use of these filters) can lead to more detailed strategies. However, taking action here requires confidence about the actual data gathered. Therefore, it’s important that the first 5 steps of this action plan have been properly realised. Only then you can take effective actions through data-driven decisions.

Do you want to know how to gather your own workplace analytics?

Book a (virtual) coffee meeting