I’ve been studying People Analytics for almost 20 years now and this world has really changed. In this article I’d like to give you some insights on the explosive growth, and explain some of the new research we just published.
1. People Analytics has grown up – it is now an established discipline in business.
For years the discipline of HR analytics, training analytics, or people analytics was considered a nichy, backwater part of Human Resources. You may have had an I/O Psychologist analyzing engagement data, or a data analyst looking at the impact of various training programs, or you analyzed job advertising to figure out which ads produced better candidates.
Each of these projects was typically done by a technical individual, often a person doing it in their spare time, and there was a small budget for analysis tools (spreadsheets were always the most popular) or a few specialized tools that helped collect data in that domain of HR.
The projects were often used as a way to “cost-justify HR investments” so we looked at the ROI of various training programs, variations in employee engagement and where managers needed to focus on employees, or perhaps how pay was unevenly distributed, the distribution of performance ratings, or other topics which helped the company improve its various HR programs.
I met with dozens of groups like this over the years and always did fantastic work – but the two issues they typically raised were (A) we don’t have enough budget to expand this work into an enterprise-wide program, and (B) the quality of our data is poor and we don’t have the money, infrastructure, or IT support to build a complete data warehouse (now called a “data lake”).
Today, I’m happy to say, all this has changed. With the increased focus on measuring diversity, gender pay equity, skills gaps, labor utilization, retention rates, real-time feedback, and even organizational network analysis, CEOs and CHROs now understand that people analytics is a vital part of running a high performing company.
Witness the following data from our High-Impact People Analytics study: this year 69% of companies are integrating data to build a People Analytics database . In prior years this was always about 10-15% of the organizations we surveyed – this huge change in investment is a sign that this discipline has grown up.
I recently completed a trip to Asia and spent a few hours with the People Analytics team at a large financial institution. I was exhilarated to find myself surrounded by 15 senior HR and technical professionals, an HR VP and senior Director, and a set of business partners dedicated to helping business unit leaders take advantage of the people-related data in their organizations. This kind of effort is now happening in large companies everywhere.
2. The problems of data quality, integration, and integrity are being addressed.
In all the research we’ve done on this topic (including the study we just published), the problem of data quality, consistent definitions, and data integrity (ie. We don’t have multiple copies of the same data in different forms) have been major obstacles. In the 2017 Deloitte Human Capital Trends we found that 39% of business people believe their company has “very good” or “good” quality data for people-related decision-making and 31% understand what “best-in-class” people analytics looks like. This is an astoundingly high number and I believe the 2018 data will show even more progress.
(Our research also found that level 4 companies are twice as likely to have a data council responsible for data governance – a critical success factor in keeping people analytics data reliable and useful over time.)
In our new High-Impact People Analytics maturity model we found that 90% of the companies at level 4 believe they have accurate people-related data, 95% believe they have strong practices for data privacy and security, and 75% believe they have consistent data definitions. While this is still a small number of companies, you can see that “world-class” is now easy to define.
There are two reasons this has happened. First, the need for high-quality data is urgent, as CEOs and CHROs are being asked to report on pay equity, diversity, and skills gaps by the board. Second, there’s a new generation of integrated cloud HCM systems (approximately 40% of companies now have a cloud-based HCM system) that require a company to implement a more consistent system of record.
I am not saying this effort is easy. According to the latest Sierra-Cedar survey on HR systems, the average company now has more than 7 “systems of record” for people related data. (Payroll, learning, recruiting, performance, engagement, wellbeing, and others.) But what has gotten easier is integrating this data – a wide set of new tools is now available to help integrate data in an easier way than ever before, and most big companies now have Hadoop clusters and data lakes they can set up to bring all this data together.
3. Companies are greatly expanding the type, nature, and level of data to analyze.
We now live in a world where employee-related data is everywhere, and it is expanding day by day. Most companies have lots of data about pay, performance, learning, job candidates, recruitment, talent mobility, and organizational compliance.
But they now also have near real-time data about employee engagement (coming from pulse surveys or continuous performance management tools), employee recognition (from social recognition systems), employee communications and teams (through organizational network analysis and email metadata analysis systems), travel and location (through time and expense, employee badge readers, or phone location data), employee wellbeing (through wellbeing applications and voluntary data shared about exercise and fitness), and even trust and employee sentiment (through “mood analysis” of survey responses and emails).
Our research shows that advanced companies now use 7 different “methods” for capturing data, including looking at internal and external social media, ERP systems, surveys, and analyzing information in business communication tools. Many of the new email systems being offered now enable “organizational network analysis” (ONA) to look at email metadata, so this data is easier and easier to collect.
I know it sounds a little creepy, but several vendors now sell software that reads email and identifies the “mood” or “changes in mood” in team or organizational communication. One of them showed me data that can spot “stress” in the organization and has proven that its algorithms can pinpoint areas of potential fraud or client projects that are going poorly. We now have access to many tools that measure stress in our voice: I would not be surprised to see systems in the workforce in 2018 that listen to meetings and identify areas of stress. (Note that Amazon just announced Alexa for Business – one could easily build a “skill” that listens to meetings and applies off the shelf AI algorithms to analyze the conversations.)
One of our clients told me about a project they did to analyze the performance of their engineering teams. They asked a set of engineers to wear smart badges and join a project to understand “what makes engineers happy and productive at work.” After several months of analysis they found that the “happiest” engineers were the ones who moved around the most – they had more physical activity, more relationships, and spent more time meeting with others. This was important data used to reorganize the facilities, change the way meetings were handled, and improve management practices to encourage engineers to spend more time with their peers. Almost every company now has the ability to do this type of analysis.
In our newest research we highlight how JetBlue uses many sources of data to understand attrition patterns, drivers of engagement, and causes of flight delays and low productivity. They integrate feedback data, crew and customer complaints, HRIS data, training data, and data about employee flight activity into an integrated system that gives the company a total picture of employee satisfaction, engagement, and customer service. Intuit is doing the same.
4. Data and analytics literacy has become an imperative for HR professionals.
I remember a meeting with a CHRO several years ago where he told me “I’m tired of hiring HR professionals that don’t know the difference between a median and a mean. I’m thinking of asking all my HR teams to take a course in statistics.”
Well that dream is starting to become a reality. Our new research shows that one of the biggest factors that predicts success in People Analytics is not just the skills of the analytics team – it’s the skill set of the HR business partners, analysts, and staff. In fact we found that level 4 companies have a sharply higher set of skills among their general HR population than those at lower levels of maturity – and I’d venture to say that this is a new “bar” they have raised for their teams. (Level 4 companies report that 63% of their HR professionals have strong analytics literacy, vs 20% in Level 1 companies.)
So HR professionals out there: it’s time to become data geeks!
The reason for this is simple. In the world of People Analytics today, the power of the analysis is not always the line manager looking at a dashboard to figure out why someone is likely to leave their group (they don’t have the time or inclination to do this). Rather it’s the HR business partner, HR VP, or HR consultant who comes to the senior leader and shows him or her data which points out that their team has bias, poor work practices, weak skills, failing culture, or other problems that can be proven with data.
Most business people have learned that data is key to their success: they will not listen to an HR professional waving their hands about how bad the “culture” may be or how “biased” or “unfair” the organization has become. They want data to prove what is wrong and they want data-driven recommendations for improvement. If the HR professional cannot make that case, show the data in a clear and understandable way, and defend their analysis, the line leader simply will not listen.
I know that in my case as a business leader I always ask people to give me a data-driven explanation for why they recommend a course of action. If they can’t show me the data I always wonder where they came up with the advice. This has now become the new world of HR: if you can’t put data behind your work, business leaders just will not pay attention.
So the problem is not just “having the data” but “knowing how to use it” and understanding how to explain it, visualize it, and put it into action in front of a business leader. And the business leader may have an MBA or background in statistics and is very likely to ask you “where did this data come from” and “how did you come to that conclusion.”
HR teams are not there yet – I still hear continuously that HR business partners are not analytic enough. But if there’s one thing you should think about in 2018, it’s “energizing your HR organization” with a good set of courses, programs, and exercises in statistics, data analysis, and the effective communication of data-driven recommendations.
In our research we detail the program Chevron developed to build global HR skills in analytics: it has been extremely effective in their organization, and serves as an example of how important it is to take data literacy seriously within the HR function.
5. AI and Machine Learning have arrived – and People Analytics teams are using these algorithms to partner with the business
The final change I’d like to point out is the fact that advanced statistics, neural networks, and other forms of machine learning have arrived. LinkedIn just published a study that shows skills in “machine learning” are now the hottest in the marketplace, and a new study by a team of AI leaders shows that courses in AI are exploding in popularity. These professionals are now in the workforce and they are itching to look at interesting data problems in business.
(By the way, if you dig into machine learning, you find that it’s essentially a lot of math. People Analytics teams are going to be able to develop or use these algorithms from public domain APIs, so this technology is available to any company.) I’ve now talked with HR departments who are looking at attrition patterns, prediction models for performance and retention, models for employee absence and grievances, and analysis of many other forms of employee productivity – all based on the People Analytics data available within their organizations. These companies are starting to correlate this data against data available from external social networks and can now learn things about their companies they never before thought possible.
One vendor, for example, now has a tool that reads comments from bi-annual engagement surveys and automatically recommends direct behavioral changes to managers to help improve the engagement and productivity of his or her team. Another company has built a machine-learning algorithm that identifies the behavior of their best sales people to help understand how to train others to perform at a higher rate. Many professional services firms are looking at communication patterns and travel schedules for the highest performing consultants to figure out what others can learn.
We used to think the secret to productivity at work was “skills.” Now, through the use of machine learning, we can understand that the secret is also “behaviors,” “habits,” and “patterns” that highly successful people adopt. Many of these are unconscious by the experts, but can be analyzed and understood by software.
Our research shows that the highest-performing people analytics teams are now partnering directly with the business, serving as internal consultants, and bringing their analytics expertise to bear to focus on productivity, performance, safety, and direct facing work related problems.
As the head of analytics at a large technology firm put it:
“I am not in the curiosity business. We need to know the relevance to the business before we spend time and energy to work on a problem.”
This is the new mantra we see taking hold.
Let me summarize by saying this has been quite a journey. I’ve been studying this domain for almost 20 years now and the domain of people analytics has reached the C-Suite. For those of you still wondering how to proceed, I suggest 2018 should be your year to consider investing in these technologies. I look forward to hearing your stories and would love to help any organization understand how to take advantage of this critical new business imperative.This has been a long journey, and it continues.