Human resources leaders have an idea that artificial intelligence and machine learning tools are going to transform the way they use their HR management systems. They are right, in theory.
For years HRMS vendors have been talking about the promise of AI and machine learning and how these tools will soon behave more like Netflix and Amazon, and customers are excited. “They want an individualized experience based on who they are, what they prefer, and what people in similar roles have done,” said David Ludlow, group vice president of SAP SuccessFactors.
They also want predictive analytics that will help HR leaders attract and retain a more effective and sustainable workforce, said Chris Havrilla, vice president of HR technology and solutions provider strategies at Bersin, Deloitte Consulting LLP. But so far their expectations haven’t been met.
Deloitte’s 2019 “Human Capital Trends” report shows less than half of cloud-based HRMS users say their platform delivers the employee experience (39 percent) better data and insights (40 percent) and increased HR tech innovation (32 percent) that they expected. Worse, just 5 percent believe their platforms do an “excellent job” of meeting full time workers’ needs, and even less believe its meeting the needs of alternative workers. “That’s hugely disappointing,” Havrilla said.
Though it’s not entirely the vendors’ fault.
Part of the problem is the lack of consistency and integration across their own human capital databases. Deloitte’s report found the average HR department now has nine different systems of record (up from eight in 2018). This legacy of patchwork HR systems makes it difficult to integrate databases for meaningful analytics.
6% HR leaders believe their HR technology is excellent.
65% HR leaders believe their HR technology is inadequate.
The way companies capture and keep their human capital data is also a problem, said Mike DiClaudio, principal management consulting for KPMG. Outdated data practices and lack of consistent data structures also gets in the way of effective data analysis. “They are putting old data in a new platform and expecting different results,” he said.
Ludlow agrees. Companies can’t even deploy basic automation tools, such as chatbots to answer HR questions, unless the tool can access all the data related to benefits, paid time off, and other HR programs. “There is no point in rolling out a chatbot if only three of your policies are digitized,” he said.
Time to Get Clean
DiClaudio argued that getting the most value from these platforms requires users to go through a digital transformation that includes cleaning up their existing data, breaking down silos between systems and changing the way data is captured and used going forward. “There is no silver bullet to any of this,” he said. “You need all of these pieces in place to predict outcomes.”
To begin this journey, Havrilla suggested creating a list of data transformation goals based on what’s desirable, what’s feasible, and what you can do today to deliver profitable results that will help secure funding for future data projects. That might include cleaning all the data for a single department, or focusing on one area of HR data to integrate and analyze.
“It’s a lot of work, but it will be valuable,” she said. Once users have the right data in place to do predictive analytics, it can provide insights into who’s likely to leave, where the best candidates come from, whether you are meeting talent development goals, and many other talent management issues.
Customers who feel lost should speak to their vendors about what’s required to get to an ideal end state, and consider working with consultants who can help them craft a data transformation plan as part of their roll out.
“Vendors now recognize that the implementation is as valid as the product itself,” DiClaudio said. “They aren’t just selling a product, they are helping their clients solve a problem.”
He added that buyers should talk to their HRMS vendors about what support they can offer in helping them transform their data, and how the platform will evolve to meet their needs over time. “Don’t buy the technology,” he said. “Buy the partner who can help you make it do what you want it to do.”