Early in life I developed a love for microeconomics. All of my electives at uni were micro or mathematical economics based. At the time this was unfashionable but over the last 15 years it’s risen in stature.
Microeconomics is all about understanding how people react to incentives. When we’re doing People Analytics we’re not dealing with numbers generated by some machine, we’re dealing with information created by people dealing with the constraints they’re facing as they try and find their best solution. To be a good analyst you always have to understand incentives.
It’s worth understanding that in organizations there will never be a complete alignment between the organization’s incentives and those of the individual. Assuming what is right for the business will also be right for the individual is dangerous. Therefore it’s right that we consider explicitly individual-level incentives and identify ways that these can be harnessed to increase the value to the organization.
What’s in it for me?
People often ask me what is the most important thing is to doing great People Analytics. The answer is simple – trust. If your employees don’t trust what you will do with their data they won’t give it to you, or won’t provide relevant, truthful data.
Too often I see HR rushing into analysis thinking only (a) what can they do to improve the business (b) how can they improve their own lives. Whilst the first one is important, critical even, it’s always important to think about what’s in it for the employees.
People Analytics teams always need to be in a position where they are comfortable explaining what they’re doing to the employees. I personally think we can hold our heads high with the work we’re doing. In many instances we’re looking for information to make better, more objective decisions. We’re using data to reduce bias.
Give people a choice between acting in the best interest of their employer and acting in their own best interest and most of the time they’ll pick the latter. Of course sometimes the two are aligned but this isn’t always true.
With any decision there are likely to be winners and losers. Prospect theory suggests that those who are going to lose out will be more vocal than those who will benefit. Those whose careers have been built on politics will feel threatened by moving to an objective world.
However, this isn’t a reason to not being objective. Company’s can afford to be more objective. It’s right to try and maximise the overall benefit for all stakeholders.
For any People Analytics work to be sustainable (and thus maximising the benefit over time) it needs to benefit all stakeholders. You’ll need employees to help give you great quality data. To do this you have to ensure they can see why it’s in their own personal interest.
The vast proportion of our analytics projects can be explained as being good for employees. Attrition modelling is all about reducing the economic cost of people leaving. People leave because they’re unhappy. You can only reduce attrition by increasing this happiness.
We can think of other topics. Sickness, health & safety (reducing accidents), increasing customer experience, better alignment of teams, diversity, employee experience. What is good for the employer is almost always good for the overall employee population. However the question will often need to be reframed to make this explicit.
People Analytics teams need to be proud of what they’re doing. They should communicate what they’re doing in as open a way as possible. They should engage with works councils where these exist.
If you don’t do this your work won’t be sustainable. You should assume that employees will find out what you’re doing with their data. It’s better for you to tell them in advance. It’s important to build trust.
How thinking about employees’ incentives means better data
Data acquisition is a fundamental part of any analysis work, and in many instances you will rely on employees providing that data. Again, if you want your work to be sustainable you need to think about what’s in it for them to support the initiative.
Again, we have a split between what is in the company’s interest, and what are the incentives for the employee. Unless you consider the latter, and build a convincing case for them personally benefiting, you won’t get great quality data.
For those who are entering the data as part of their job maybe the incentives are aligned. For many other instances there has to be a good reason why someone should provide data.
One of the models we use to describe how improving incentive alignment can improve analysis we call the Virtuous Circle of Data Quality. This describes how employees will contribute more information, or take better care with providing information, if they feel that the analysis provides benefits to them.
One area close to my heart is when asking for feedback. Employees are clearly motivated to provide feedback if they feel that their feedback will make a difference. We see this from both engaged employees – who care deeply about an organization and want tit to improve – and disengaged employees – who want to tell you what’s not working.
It’s worth thinking, however, how you can take this further by providing individual-level reporting. One early proponent of this is Towers Watson who offer individual engagement reports. Their implementation seems very similar to how psychometric test developers often provide reporting to test participants (an almost identical incentive problem). I think this is a great first step. I suspect that user-centred design combined with advanced analytics is the way to build upon this approach.
Help people meet their objectives
The explosion of workforce data has created great opportunities for helping individuals to manage their work in ways that help meet their own personal goals. As analysts and technology providers we need to now build tools that meet this need.
As people analytics evolved we’ve matured from solving problems which were of most interest to HR teams to addressing issues that met business objectives. The next phase will be about providing tools that enable individuals to meet their objectives, even if their objectives aren’t explicitly the same as the organisations’.
We could think about using recommender systems to provide advice, both in terms of identifying knowledge or proposing contacts and networks. We could be thinking about how people like them manage their careers or realise more value from the vast amount of non-pay benefits that are often available, but not well known within organizations.
What is common across almost all of these potential solutions is that (a) there needs to be a recognition that employees’ and organizations’ incentives don’t perfectly align and (b) they use technology, data and analysis to reduce friction including search costs.
Analysis becomes sustainable when it addresses the needs of multiple stakeholders. HR teams should ask themselves the question of ‘what’s in it for the employees?’ If they can answer this question effectively it greatly increases the chance of success.