Big data and predictive analytics might be buzzwords right now, but their power is very real.
By Katie Kuehner-Hebert
Metrics such as time-to-fill, cost-per-hire, and candidate diversity are always critical for recruitment outsource providers to meet. But now buyers are increasingly asking RPO firms to provide more sophisticated analytics to help them better compete for qualified and productive candidates who will stay longer and make greater contributions to the bottom line.
Some of the data and key performance indicators associated with the delivery of the RPO under a service level agreement have become very basic and might not necessarily be representative of the “nouveau” term for analytics or predictive analytics, says Glen Johnson, director, staffing and recruitment at Delta Air Lines in Atlanta. “You’re not only talking about historical trends, but also predicting future trends on the migration of talent,” Johnson says. “Historically you might have had trouble filling jobs or finding talent in specific locations. With predictive analytics, you can also deliver information back to the business and hopefully tie in with what they are doing.”
For instance, Delta hires hundreds of people at numerous airports during peak seasons, and the airline might use people who are already part of its seasonal workforce pool. “Using analytics, I want to be able to see how we have trended summer over summer at each airport,” Johnson says, “but I would also like to find out from leadership any anticipated short-term or long-term operational adjustments or other internal or external factors that might impact these trends and put those into the calculator, as well to see if we need a different or more flexible hiring strategy. For example, if the leadership has determined that we are not going to fly a certain route as frequently, that might impact staffing levels at that airport.”
Johnson is working on a plan to provide hiring data per role and location to the airline’s network planning team, which has predictive analytics tools on scheduling routes per customer demand. That way, Johnson can marry up historical staffing data and trend analysis with network trends to better determine Delta’s workforce planning needs. “I am not likely to invest in expensive workplace planning tools,” he says. “At Delta, most of our investments are in our onboard product and the customer experience, so when it comes to back-office tools, especially in the HR world, we have to be creative and innovative.”
Where Providers Come In
RPO providers can play an important role in the use of predictive analytics, says Johnson. Delta works with PeopleScout in Chicago. RPO providers have the ability to look across their entire client base and see broader market trends in terms of talent shifts and talent migration, which can give their clients a much bigger picture than just looking at their own data, Johnson says. “For example, an RPO provider can use data from all of their clients to see trends happening in the marketplace, such as engineers disappearing out of Arizona and a bunch of engineers coming to Texas, and then inform each client about these trends as they are pertinent to the business,” he says.
Rebecca Callahan, president of RPO for Randstad Sourceright says predictive analytics is becoming increasingly important for understanding how candidates choose to apply for jobs on client websites. For example, analytics can be used to determine whether a candidate clicked a link to a client’s website because the job opening was posted on a job board or because the client positioned its brand visibly within social media, backed up by the right ads in the right places. “Organizations now realize the types of analytics they need to support recruiting efforts, and data that comes from social media and search engines ultimately helps create strategies that drive candidates to their career sites,” Callahan says. “They want to know where they should place ads for the types of candidates they want to attract, and how to enhance and put employment branding out there to attract the right candidates.”
Analytics can also demonstrate whether a particular source produces qualified candidates who actually get hired, compared to sources that just produce reams of applications, says William Sebra, president North America, FutureStep in Houston.
For example, a large financial institution recently asked FutureStep to analyze its spend using the RPO’s decision-support analytics tool, Sebra says. While the number-one source of candidates came from one of the larger nationally known job boards, FutureStep determined that the job board was actually the fifth-largest source for producing hires. The number-one source for actual hires was employee referrals. “The financial institution was spending $250,000 a year with the job board only to produce a very low number of hires,” he says. “So in relationship to the client’s spend and the real ROI, there was a big disconnect.”
The analysis led to a decision by the client’s talent executive to move funds from the job board contract over to fund an employee referral program that did not exist at the time, Sebra says. The overall talent acquisition budget did not change, but the client used the analysis to make a solid business case that produced even a greater number of hires in subsequent time frames.
Turnover Risk Analysis
John Wilson, chief executive officer of WilsonHCG, says RPO providers also use predictive analytics to determine the risk of certain employees leaving their clients due to the probability of extending a leave of absence, underperformance, or the opposite (stars who are prime candidates for cherry-picking).
“We started seeing from clients the desire to know what was going to happen, just as CEOs forecast earnings for the next quarter, by looking at what was happening with their current headcount,” Wilson says. “For us, as an RPO that has to help clients quickly scale up or down, it is important for us to understand what kind of support our clients will need from us.”
For example, after talking with a client’s unit hiring manager about each worker in the unit’s current headcount, a WilsonHCG recruiter might put a number or a notation next to each worker’s name on an Excel sheet, and the risk of every worker would be identified. That data would then be sent to the client’s senior leadership.
Ten years ago the technology and tools to aid in this type of analysis did not exist, but now there are more sophisticated reporting capabilities within applicant tracking systems (ATS) that also incorporate data from HRIS systems, Wilson says.
Taryn Owen, PeopleScout’s senior vice president for global operations, says the RPO firm captures market and competitor data for its clients using a variety of tools, including analytics within its proprietary ATS and through the reporting capabilities within its business intelligence platform. PeopleScout also provides clients with a variety of analytics that are driven by the needs of their business.
Take Covance, a global drug development company based in Princeton, N.J., that fills some 4,000 positions a year across the globe in about 60 countries, according to Lisa Calicchio, its vice president of recruitment and talent services. Covance performs development research for pharmaceutical and biotechnology firms who are seeking to bring new drugs to market, and as such Covance hires many scientists, technicians, and statisticians to analyze the safety and efficacy of the new products.
Calicchio not only expects PeopleScout to provide standard reporting analytics to measure how efficiently and effectively positions were filled, but also to help Covance determine how to get the most return on investment in the firm’s most critical job positions through predictive modeling. While all 4,000 positions are important, Calicchio and her team are currently identifying the 10 most critical position types in terms of revenue generation and ability to deliver outstanding service to clients. “We need to focus on the top 10 position types to ensure we’ve prioritized our investments appropriately where we’ll maximize ROI,” she says. “We want to make sure we are on top of our game as much as possible, especially in these areas. We’re working hand-in-hand with PeopleScout on this effort, as they handle a large part of our recruitment portfolio.”
PeopleScout also helps Covance use analytics to help make decisions on how to best source talent and drive more personalized experiences tailored toward candidates, Calicchio says. For example, Covance and PeopleScout might choose to recruit young people for entry-level lab technician positions using social media, while using job boards or techniques that capitalize on word-of-mouth and networking or on job boards for older candidates who would be interested in the position of finance director.
PeopleScout also works with Calicchio and her team to develop general hiring themes by analyzing data from reports derived from the RPO’s ATS, follow-up candidate surveys, and market intelligence data, as well as from brainstorming sessions that incorporate benchmarking against best-in-class companies. The partners weekly discuss hiring trends for each country. “For example, Germany is one of the most difficult countries in the world to recruit in, because there is such a strong loyalty factor,” Calicchio says. “Similarly, Japan is a very difficult country to recruit in, as non-Japanese companies are generally not perceived to be as desirable as Japanese companies. It’s a culture thing, not a Covance thing. In these ongoing conversations with PeopleScout, we share anecdotal information, and we collectively figure out whether we need to do things differently.”
Quality Control
RPO providers also use analytics to hit quality-of-hire goals with “embedded” assessment programs, such as those offered by Partners, SHL in Deerfield Beach, Fla., says vice president Jeff Holman. SHL uses predictive analytics in its psychometric assessments of employees to determine how people will perform for certain job roles based on their behavioral traits, cognitive reasoning abilities, and other skills and capabilities. “Many second- and third-generation clients of RPO providers have really refined their services as low as they can go on a cost-per-hire basis, and they’ve essentially sucked all the dead air out of process efficiency,” Holman says. “Now we’re moving to a more mature RPO market, and we’re seeing clients asking for very specific terms on quality of talent, and measurements aligned to psychometric analytics. For example, contact center agents are some of the most tightly measured performance roles in the workplace and we can positively affect areas such as average handling time, revenue per agent or retention of top performers.”
Pinstripe, Inc., is a Milwaukee-based RPO provider that has experience integrating an SHL assessment solution into a client’s talent acquisition process, according to Angela Hills, Pinstripe’s executive vice president and commercial business leader. “Whenever we can add science to our talent acquisition processes, such as the additional facts that assessments provide, we can truly validate that our practices are aligned to our talent strategies,” she says.
Pinstripe has also built its own analytics tools to provide consultative advice to clients on the supply and demand of the talent pool, she says. “For example, if we’re looking to hire pediatric clinical nurses in Philadelphia, we would use the data within our analytics tools to assess how many nurses are located in the state of Pennsylvania and how many are within a 30-mile radius of our client’s location,” Hills says. “We’ll also consider other factors such as what other organizations in the market are competing for the same talent, and how much they are willing to pay. This data-driven assessment tells us whether we are looking for a needle in a haystack or for readily available talent. With this information, we help our clients make informed decisions about how to tackle that search.”
While large employers have the resources to perform the most sophisticated analytics on how to best attract candidates to their websites, mid-market employers are looking for analytics to figure out how to compete, Callahan says: “A client with a call center job in a certain market needs to know what other jobs are out there. That includes jobs that are posted online, but also jobs that are profiled on LinkedIn or Facebook. If the client posts a call center position in a saturated market, they need to understand the competitive landscape. We help them with analytics such as their competitors’ average pay rate, and other incentives that make them more attractive to candidates, and might impact a candidate’s decision to work with our client.”
Big Data, Big Challenge
RPO analytics, particularly those that require more sophisticated predictive modeling capabilities, can be challenging as client databases become more unwieldy in the era of “big data,” says Tony Ashton, vice president of product management for workforce planning and analytics at SuccessFactors, an SAP company. Software-as-a-Service, or operations in the cloud can help greatly, particularly for global clients that have multiple HR systems in different countries, Ashton says.
“We offer the ability to pull data from lots of different sources, as the reality is that clients have many different systems, and we need to make sense of that,” he says. “Often companies have data scientists, whose role is to be inserted between the hodgepodge of databases and managers—the filter to make sense out of the chaos. We certainly want to make sure those core analysts have all the right tools to drive centers of excellence, but we also need to provide information in a way that is simple and easy to understand for end-users, to help them make better decisions.”
According to Ashton, SuccessFactors’ cloud solution has an extensive library of questions and metrics, organized under topics, such as the degree of referral conversion, the relationship between different hiring sources and performance, time to productivity, and whether people from different sources get promoted faster than others or have longer retention and contribute more to business effectiveness. The provider helps clients determine which questions might help solve their particular concerns and then develops metrics around those questions.
PeopleScout also provides clients with analytics to measure hiring manager satisfaction, Owen says: “We ask our hiring managers, ‘Do you consider the candidates hired from PeopleScout an upgrade to the person who held the position previously?’ In the case of Burger King, 88 percent of the hiring managers answered yes to this question. In a client engagement, PeopleScout can achieve every KPI [key performance indicator] and SLA [service level agreement] metric—such as time-to-slate, time-to-accept and time-to-fill—but at the end of the day, if the hiring manager doesn’t achieve the quality of hire that they expect, the SLA achievement is irrelevant.”
In the era of big data, there is no shortage of statistics, but the trick for an RPO firm is to work with clients to figure out what type of data is relevant to their bottom line and which analytics would be meaningful to their C-suite, Sebra says. “Futurestep is looking at how to gather the right data to make decisions that will evolve the recruitment function from reactive to proactive, to better align with the business strategy and to drive as much predictability into the system as possible,” he says.
What does the future hold for RPO analytics—how much more sophisticated will it get?
“The Nirvana that people want is to find out which channels provide more productive performers than other channels, but these analytics are still being developed in the market,” Callahan says. “More sophisticated models and analytics might be developed, but turnaround can also be caused by ineffective management, and you can’t always solve for all those soft pieces in the middle.”
