5 Ways to Kickstart Your Predictive HR Analytics Activities
Every time when we meet with an HR professional who calls us in to discuss building predictive HR Analytics capabilities, we get the standard question: how can we best start in the fastest possible way? Of course, there’s not such a standard response to this question as every organisation is different and has varying challenges with data, analytical capabilities, business and/or people issues, etc. However, below you can find some of our key answers to the most common questions. The iNostix FAQ so to say…enjoy!
1) Outsource your HR analytics support to experts
Organisations often have a short-term need for quick people analyses to support the decisions they make (e.g. at budgeting time, during assessment processes, re-organisations, etc.), without them wanting to worry much about building internal HR analytics competence. At such moments, the presence of an external expert can provide considerable added value. But speed also often has to do with the ability to develop maximum competitive benefits quickly, which was also a reason to outsource for a number of our clients.
Quite a lot of our client organisations are working on building up their internal analytical capabilities. However, this takes reasonably long and in the meantime they call on external experts for HR analytics pilots in order to make progress, gather important experience, get to know their data better, etc. For all learnings that clients acquire during such pilots, see the next point in this blog. Read also all the details with regards to outsourcing in our blog: ’12 Reasons why outsourcing HR Analytics is good for HR’.
2) Start a small pilot!
We always recommend that organisations wanting to quickly start with HR Analytics begin with a ‘low hanging fruit’ pilot and we advise them not to make the mistake of only focusing on the analytical outcomes of the pilot itself, but rather (and foremost) on the wider context of why the pilot was organised. Analytical results are always impatiently awaited but as important as these results itself are:
- gathering analytical experience in general;
- gaining an insight into the data complexity, data ownership, data location (who owns the data, where are the data stored, how to get access to the data,…);
- understanding the quality of the data (evolving from an administrative to an analytical data vision);
- collaborating cross-functionally with the business; (see this great case study about such collaboration)
- learning to work with data analysts (who do analysts report to in HR, where do they fit in the structure?);
- organising the analytical activity: doing it internally or outsourcing it? (see our blog on this subject)
- asking the right business questions;
- understanding analytical methodologies; (see our blog on this subject)
- presenting business cases and learning to improve the storytelling and visualisation of analytical outcomes;
- learning to deal with data governance (privacy, legal, transparency,…),
- working with IT on data gathering, data integration, data protection, etc.
By the way, we always advice ‘starters’ that they shouldn’t wait until all their HR data is in order before starting with HR analytics – a stance (and sometimes an excuse) often heard from HR. It’s a parallel and even interactive process: analysing data and at the same time getting to know its real quality when you start working with it in a pilot. You can’t begin analysing early enough because you’ll probably need to keep track of a whole series of data for longer. So it’s best to start doing this right now. Here you can read more about other learnings from our experts.
3) Think of these 4 approaches to kickstart your HR Analytics
We are being asked regularly what kind of analysis we perform in our predictive HR Analytics projects. For that reason, we made a short overview of the 4 most common approaches or methodologies that we use over and over again in our blog ‘4 Approaches Everyone In HR Analytics Should Be Using‘. To summarize these 4 approaches:
Approach 1: Clustering: With this approach, you can start investigating hidden group patterns with the help of clustering techniques.
Approach 2: Driver analysis: With this approach, you can start understanding hidden relationships. Most of the time, we use the word ‘impact’ to explain relationships between events or people/business characteristics.
Approach 3: Risk analysis: With this approach, we can start understanding probabilities or the likelihood of (important/strategic) occurring events.
Approach 4: Forecasting: With this approach, you can start understanding future trend lines, based on historical patterns. See also our next point in this blog.
You can find more details and examples on these approaches in our blog: 4 Approaches Everyone in HR Analytics Should be Using.
4) Add statistical forecasting to your important trends, such as turnover or absenteeism
Traditional employee turnover or absenteeism analysis is typically executed with retrospective, descriptive reporting. Although this is definitely useful as a start, statistical forecasting will dramatically enrich the insight. By adding a 3, 6 or 12 months statistical forecast to your standard turnover or absenteeism reporting, the insights will become much more actionable and will get more attention from management.
On top, the addition of statistical forecasts (versus financial estimates or assumptions), will help to calculate the costs of future turnover in a much more precise and impactful way. Your CFO will like it! More details on analysing turnover can be found in our series of blogs on employee turnover.
5) No need for big upfront investments in analytical technology
HR doesn’t need to make big investments in data technology to do predictive HR analytics. Of course, when you are planning to kickstart with the help of external experts, you don’t have to invest in technology at all. Let me quote Patrick Coolen (ABN-AMRO) and Esther Bongenaar (Shell) in the interview ‘Debunking 5 predictive HR analytics Myths’:
Patrick Coolen: To start with analytics you do not have to invest in technology immediately and you do not have to create that perfect data cube before starting. My advice is to start small and learn fast. If you have a ‘good’ dataset, you can start with analytics today (Patrick has outsourced his analytical activity).
Esther Bongenaar: If you have a laptop with Excel, you can do correlations and regressions. Add R software for free, and you are ready for advanced analysis. You may invest in statistical or visualization software, but it is not a requirement. We do the majority of our analysis on laptops equipped with Excel and R. Regressions, clustering, and transformations form our basic toolkit (Esther has not outsourced analytical activity).
You can read the full interview with Patrick and Esther in our blog ‘Debunking 5 Predictive HR Analytics Myths’.
Related HR Analytics articles
Interested in kickstarting predictive HR Analytics as a key component in your people or business strategy? Get in touch with CEO and co-founder Luk Smeyers for more information (mailto:firstname.lastname@example.org or via Google+). Or follow iNostix on Twitter and/or Facebook for exciting international articles on HR analytics. And don’t forget to register for this blog!