Last Wednesday I attended the 2nd HR Analytics conference from the CIPD attended by close to 150 delegates. All were hungry to understand how to crack the HR Analytics nut. The day was arranged with a set of presentations from the leading firms driving the implementation of HR analytics. The message from them was clear; the path from collecting and storing the right data to gaining relevant insight takes a lot of time and effort. Every delegate went away thinking they needed to build a data warehouse, a database used for reporting and data analysis, and that to do this required the hiring of highly skilled and educated individuals. Unfortunately this is representative of a space individuals and organisations are really struggling to understand and get to grips with. I believe that when it comes to HR analytics, building a data warehouse is expensive and time consuming. It is time this approach was revolutionised.
The data warehouse approach is not working
The first presenter, from a major bank, was up first because it was perceived he was furthest along the HR analytics journey, rating his organisation’s progress at 5 out of 10; they had agreed what data to collect, were building a data warehouse, had developed dashboards using data visualisation software and were beginning to get valuable insights from their data. During his presentation he hammered home the fact that building a data warehouse is hard, and that those going down the HR analytics route were in for a long, tough ride. To put this in context, he estimated that his company would take 7 years with a team of one hundred to complete their data warehouse. Yes, you read correctly. That’s 7 years with 100 people – 700 man years! Assuming these are highly skilled people with an average annual cost of £65k ($100k), then that reads as £45m ($70m) in monetary terms. These figures may be wrong and one hundred people may not be consistent throughout the seven years, it might just be a peak number. Even so, the order of magnitude is crazy. It is remarkable that out of six companies prioritising analytics the one which was furthest along only rated their progress at 5 out of 10 and envisioned many more years of pain.
HR should stay away from data warehouses
At Concentra, the firm I run, we work with and make extensive use of data warehouses. We have a team of BI developers who create data warehouses in some of the most complex settings to solve some of the most complex issues. For example we are currently helping a global FMCG firm take out over $500m of working capital across 180 markets through inventory optimisation and better forecasting. This is using the same sort of data visualisation software that the presenter was describing. However, there are two problems with translating this approach to the HR world.
- The type of team you need for this work is extremely skilled and hard to attract – I should know given the amount of applications we sift through year on year.
- HR data is not big data. It is fundamentally different and so data warehousing is not the right tool for the job.
BIG Data vs HR data
I want to focus on the second point because if a data warehouse is unnecessary for HR analytics, which I believe it is, then the first issue of attracting a workforce to create it becomes obsolete. First, let’s be clear about the difference between big data and HR data. Big data involves looking at millions of transactions and thousands of products.HR data is small, messy data which is constantly changing. Thus, while big data requires relationship databases this approach is completely impractical for HR data. What is needed is graph databases in order to allow for flexibility and to view the data from all angles at the touch of a button. The point is use the right tool for the right job.
The alternative to the Data Warehouse
When we developed Orgvue we saw the huge need to solve the data issue. Going down the data warehouse approach is a hugely complex, long and painful journey only solved by brute force and herculean commitment. But it does not need to be like this. Using new technology with thinking from a wide variety of backgrounds means we have been able to approach the data from a different perspective. The result is that the speed at which you get results and insight from building a data warehouse can be increased exponentially. I am confident we have got to the point where Orgvue could reduce the speaker’s 700 many year project down to just 1.Orgvue can produce outputs and analysis from your data in a minute rather than weeks or months. To take an example, one consulting client did a project in 2 months with us when it previously took them 12 for the same amount of work.
Join the Revolution
The world of technology is constantly going through revolutions, and like all revolutions, it takes a while to get your head round the possibilities. The first cars were designed to look like a horse and carriage because that was the paradigm at the time. In contrast OrgVue does not look like a data warehouse but to understand what it offers you have to compare it to one. The revolution is in the making; I implore you open your mind to the possibilities and escape the warehouse.
Latest posts by Rupert Morrison (see all)
- 3 simple steps to workforce planning - November 30, 2017
- What’s the Difference Between Target Operating Model and Organisational Design? - October 20, 2017
- RACI, RASIC, or RAPID – throw-out the complexity - October 20, 2017