Why your HR data doesn’t have to be 100% accurate
As HR professionals and executives are trying to move the workforce analytics agenda forward, we are still facing challenges. Although the urgency for data-driven decision making in HR is increasing, the maturity of workforce analytics is still quite low in most organizations. Why is this?
The primary reasons I’ve found working with HR leaders are:
- HR staff are not trained to do the work and lack an analytics skillset and mindset;
- HR thinks it is all about HR and doesn’t involve business stakeholders to make it business relevant;
- Data quality and availability is low and is dispersed among different systems
In this article, I would like to address one specific bottleneck mentioned: Data quality and availability. What can be done to improve data and where do we start?
The perfect workforce data warehouse
The first thing organizations consider an important precondition to doing proper analytics is solid data management.
Most businesses have implemented IT solutions to help them store and utilize their business information more effectively. Cloud computing is the standard, big data solutions are on the rise, and organizations have centralized – or are starting to centralize – all their data in a data warehouse. However, the creation of valuable business-relevant insights about their people is not on the agenda. Why? Organizations usually indicate the following two reasons:
- There is a primary focus on business analytics and these projects always have priority. The human capital analytics projects remain on the backlog.
- The quality of employee-related data is considered insufficient to be used for analytics. Information from employees may be incomplete or has not been controlled for consistency and uniformity.
Yes, it makes sense from a business standpoint; most analytical efforts focus on customers, the market and process efficiency. And if your data is unreliable and incomplete, the outcome of analysis is unreliable, and your senior business stakeholder will dismiss the results.
Related (free) resource ahead! Continue reading below ↓
People Analytics Resource Library
Download our list of key HR Analytics resources (90+) that will help you improve your expertise and initiatives. Your one-stop-shop for People Analytics!
The opposing argument is, however, that you will never need all of your available workforce information for an analysis. Besides, you will need a very convincing business case to justify a project with a data engineer in order to get all your data standardized, up-to-date and available. A different approach is needed to start creating valuable insights from your workforce in the short-term.
Start with a simple yet relevant question
Instead of focusing on the data, start with focusing on a specific need for information your organization has regarding the workforce. Perhaps your HR reporting is up to a level that you are able to generate solid data for specific metrics such as absenteeism, turnover or performance ratings.
It gets interesting when there is a specific business-relevant question where there is a need to combine certain data sources.
For example, a regional senior executive indicates they are seeing a decline in customer satisfaction and says they might actually lose a big client. This is not perceived as an HR issue and it very well could be that it isn’t. But it is reason enough to explore all possible causes.
You have performance data that is reliable, consisting of a performance rating and a potential score and you ask an analyst to analyze these data and slice-and-dice by region, management level and explore further (if it’s a dedicated analyst he/she will want to explore further in my experience).
Perhaps you can show a link between regional performance and potential ratings from account managers and customer satisfaction or (and this is also often the case) you discover something else in the analysis that doesn’t impact the issue directly but is something that does need attention. For example, if a specific group scores high on ‘has reached potential’.
When you have done one of these analyses and you have shown your business partners what insights this can provide (be aware that you might not get results from your first try, we are talking about research and hypotheses can be proven or disproven) it will generate support for any future endeavors with workforce analytics and they will value your efforts and contributions better.
After a few valuable ad-hoc analyses key stakeholders might be more inclined to free up resources for continuous efforts on workforce data quality and availability.
Once they know it might lead to insights that create business value and contributes to decision-making on a fundamental level for the organization, priorities may shift.
The key is to take it one step at a time. Don’t try to convince senior leaders to build a factory if they are unsure of the product it will generate. Build one machine first and show them what it can do. If you can produce some good products, they will want to build more machines and possibly a factory soon enough.
For example, I saw a presentation about Big Data once where an IT specialist from a big online webshop had a hard time convincing management they should invest in a Big Data platform. So, he decided to show them, by taking some old servers, building a small prototype platform with a few other specialists and presenting how this could affect search results and – eventually – customer satisfaction. They were blown away.
If you can present a few real small-scale cases with tangible results, your stakeholders will start seeing the value of data and – equally as important – governing the data quality and improving data availability.
The more clean and reliable data you can process into your analyses, the more insightful the outcomes can be. This is the message you need to get across after you have shown the potential of the analytics approach.
We are seeing a shift in the mindset of HR professionals and executives and we are seeing pioneering efforts in getting data-driven decision making to the next level, but you’ll have to start small to make big steps.