The Talent Analytics Market Heats Up With New Cloud Offerings

Workday Introduces Predictive Analytics: Workday Insights Applications

The market for corporate talent analytics has gotten red hot in the last two years. Almost every major organization we talk with wants to build a “data-driven HR” program to help understand who to hire, identify the factors that drive retention, analyze employee engagement, and apply data to make better decisions about who to promote, how much to pay people, and just about every other talent-related decision.

As in all the other areas of business analytics and Big Data (marketing analytics, financial analysis, analysis of customer buying patterns, supply chain analytics), organizations have to build internal capability, get help from IT to bring the data together, and then move down the learning curve to better understand their data and what it means.

In the case of Human Resources, which deals with some of the most difficult decisions we make (who to hire, for example), most companies are behind the curve. Our research shows that only about 4% of HR departments are really using any form of predictive analytics today and more than 60% are struggling with a mess to just get good reports (many cannot even tell you how many people are on the payroll in a given day, and most don’t track hourly workers well).

Well this world is about to change. Almost all the new cloud-based HR systems vendors now offer what I’d call “embedded analytics” solutions built into their software, ready to turn on and analyze employee data right out of the box. This includes solutions from Oracle, SAP, ADP, IBM, Ultimate Software, Saba, Skillsoft, and soon CornerstoneOnDemand (which just acquired Evolv, an established talent analytics firm, to accelerate their analytics solutions).

Recently Workday, a fast growing cloud-based HR and Financial Management vendor, announced Workday Insights Applications, built on technology and expertise they acquired earlier this year from Identified. This product is one of a new breed of solutions to come, delivering sophisticated “out of the box” talent analytics which companies can use almost immediately.

What’s new about Workday’s offering is that it is built on technology that “categorizes people” into roles and positions for comparison. (Workday calls this feature “Industry Trees” or “Job Classification Taxonomies,” built on the SYMAN technology acquired through Identified.)

So rather than just produce correlations that might say “these are the factors that seem to correlate with people who leave our company,” it actually categorizes people into job roles and makes even smarter decisions by analyzing things like “people who are mid-level project managers in IT should not take this kind of career path, because they are likely to quit.”

Consider the problem of classifying jobs: if every job in every company had a job level and a stable job title and description, it would be easy. But of course that’s not true: every few years we reorganize, change job titles, and change roles, so “job categories” are very hard to keep track of. And if you try to compare someone within your company with someone outside (ie. for recruiting), the matching is even harder.  This is the technology Identified developed over the years.

The analytics applications introduced by Workday help you predict retention (a huge topic in business) by trying to figure out which factors correlate to employee departures (ie. salary? job rotation? commute distance? manager?). The company also introduced an application that analyzes effective and ineffective career paths, a big issue in business.  (ie. Which career moves seem to cause people to leave the company and which would help retain them?).

These embedded analytics applications are part of a category of such solutions which work right out of the box. They crawl through your employee data (they don’t leverage external information yet, more on this below), run correlations and look for causality, and try to tell you what is going on. And your team can then add external data to make the analysis richer.


Fig 1:  Workday Insights Retention Analytics

For example, the system might tell you that “Software engineers who don’t get raises every 12 months and live more than 20 miles from their work location are twice as likely to leave as IT managers who don’t get a raise every year” (kind of an obvious example). Or another example is “IT managers who work in helpdesk who are moved into operations are very likely to quit, but those who are promoted into software engineering or operations management are very likely to stay, and likely to succeed.”


Fig 2:  Workday Retention Risk vs. Performance

These seem like obvious findings once you see them, but most of these kinds of decisions are very hard to make on a daily basis. Every day we are deciding who to hire, who to promote or move into a new role, how much of a raise to give people, and who to move into management – with almost nothing to go on but our own gut feel and experience. Over and over again we talk with companies that look at the data and discover that “gut feel works” but “data works much better.”

Look at it this way. If you’re running your company on “gut feel” people decisions but your competitor is using predictive talent analytics, wouldn’t you start to worry what he knows about your people that you don’t know? You should. Talent is now the most scarce and valuable commodity on earth, so companies who really understand how to attract, retain, and manage people will win. Eric Schmidt and Jonathan Rosenberg’s new book “How Google Works” does a great job of disclosing how Google uses some pretty out-of-the-box strategies for people management that have helped the company recruit some of the best minds in business (and computer science).


Fig 3:  Workday Career Path Prediction

Workday’s solution is based on the company’s Big Data Analytics platform launched a year ago, so it not only analyzes all your internal data, it can be loaded with external data (employee job history, social profile data, anything else you can think of!). This means the solution is really both a set of out of the box applications as well as a toolbox and platform for building more advanced analytics applications.

Workday’s announcement is exciting, but it follows many similar products from other vendors:

  • Oracle has offered OBIE for many years (Oracle’s business analytics solution) and it is widely used and has been significantly enhanced with predictive analytics applications in the last year.  Oracle also offers “out of the box” retention prediction, for example.
  • SAP offers SuccessFactors Workforce Analytics, an advanced system which not only predicts retention but also has extensive modelling capabilities for workforce planning.  SAP also introduced predictive analytics for learning, which will “recommend learning activities” to each employee based on what other successful employees in a role have done, similar to Netflix or
  • ADP introduced a powerful analytics solution this year which not only analyzes your own employee data but compares it with 630,000 other companies in the ADP payroll database. This is a very useful solution: one which lets you compare your retention against competitors in your precise geographic area, for example, or quickly compare salary or HR spending data between your company and other similar companies by location.
  • IBM Kenexa Talent Analytics, which was just released recently, brings people-related data from IBM Kenexa software and other sources and uses Watson to deliver easy to use analysis with an english language interface.  IBM’s Watson technology focuses on cognitive processing, letting HR managers type english questions and interact with the system in a conversational manner.
  • Saba is introducing predictive analytics for learning and compensation, which not only recommends learning but also recommends who you should be connected to and when and how much of a raise individuals should get.
  • Ultimate Software, a leading HR vendor for mid-sized companies, offers a set of out-of-the box retention index based on various employee data.
  • CornerstoneOnDemand just acquired Evolv, an established Big Data talent analytics vendor, with the intention of developing something similar to Workday, and SAP.

Where is all this going?  While there are many fantastic talent analytics solutions out there (Visier, Vestrics, OrcaEyes, as well as Tableau, ClickView, Xcelsius, and the traditional business intelligence tools), the market is shifting toward integrated applications. Just as Salesforce created a set of integrated analytics for all areas of CRM embedded in their platform, so will the larger HR vendors develop integrated analytics applications with their transactional software.

This does not get HR off the hook: companies still have to clean up their data, manage it well, and understand what it means. And very few companies have only one HR system, so we still need to do a lot of consolidation and build a global data dictionary so we know what all our people-related data means. But the problem of “building a data warehouse” and “building your own analytics infrastructure” is starting to go away, and these solutions have highly advanced algorithms for modeling, categorization, statistical analysis, and correlation.

Now more than ever HR organizations must have an individual with the title “VP or Director of People Analytics” to bring this all together. Our research shows that people-related data can be an extremely powerful tool for better decision-making, but it’s not all in one place and we need experienced professionals to learn how to make sense of it. People and HR data is seasonal, it varies widely in meaning based on job and a myriad of demographic factors, and it’s all over the internet.  The discipline of learning to manage and model people-related data is in its early days, so you need someone to jump in and really run this new business function.

As I’ve written before over the last year or so, I believe this whole area is still in its early days, and we are probably in a ten year journey through the process of what I call “The Datafication of HR.”  And while these big vendors are all very focused in this area, there are many creative startups with groundbreaking solutions as well.

A new startup by the name of HiQLabs, for example, just started to sell a solution that predicts retention using external data alone (where you live, your commute distance, job title, social data, etc) and they tell me their predictions are four times more accurate than predictions developed using internal data. Large companies subscribe to their service to get a “heat map” of which high performers are most at risk. This is only one example of the ways people-related data will be used for better decision-making in business, and I think we have a lot more innovation to come.

The Workday announcement brings the company up to par with its competitors and further pushes the industry toward what we could consider “intelligent talent management software – tools that don’t just automate processes, but actually help you make decisions in real time. Just as Google “suggests” search query results based on your activity and Amazon suggests books, it is now possible for your complex HR software to suggest who to hire, who to give a raise, and who to move into what position. The algorithms won’t be perfect, but they’ll be better informed than simple “gut feel.”

As I discuss in the “Ten Most Disruptive Trends in HR Technology:  Ignore Them At Your Peril,” embedded analytics is one of the ten disruptive technologies in the HR software market today. If the system can’t immediately recommend or help you make better decisions based on the data it contains, it starts to look like a somewhat boring application. So the bar is raised, and we will see even more exciting value created as talent analytics enters the mainstream in our efforts to better staff, hire, and manage our people.

(More information on The Datafication of HR and a slideshare presentation here.)