Hiring, retaining, and developing the best leaders, professionals, and frontline performers forms the common thread that successful leaders from Jobs to Gates, Buffet to Bezos, and Welch to Zuckerberg invoke when asked what led to their success. In books from Built to Last, Good to Great and Great by Choice, Jim Collins argues that getting the right people on the bus, in the right seats comes before working out the details of the brilliant strategic plan. In The Headhunter’s Edge, Jeff Christian made the same argument while outlining a talent-centric investment strategy for VCs, PE firms, family offices, and CEOs. While advances in financial, information, logistics, and engineering technology eat away at those sources of strategy execution risk, there are forces taking human capital risk in the other direction. This post documents what those forces are. A follow-up post covers how innovations in natural language analytics offer hope for Reducing Human Capital Risk.
Sources of Increasing Human Capital Risk
Performance ratings, whether by a single manager or multi-rater, are going away.They are already gone at: Accenture, Adobe, Con Agra, Cargill, Deloitte, Eli Lilly, the Gap, Intel, Juniper Networks, Medtronic, Microsoft, and Sears, among others . Some have argued that performance ratings contained little information about the talent being rated anyways . But the solution to bad data isn’t worse data, as in the frequent collection of team leader micro-opinions on team members, as advocated by Buckingham and others. The solution is unbiased analytics based on big data. More on that Friday.
Background checks are under attack and in retreat, according to an article in the Wall Street Journal . That makes it harder to avoid the kind of Insider Risk that has generated large corporate fines and rumbling about criminal executive culpability. The initial response here has led in the direction of digital forensics using the rusty dull blade of Boolean search strings that have never worked well for finding top talent, much less detecting Insider Risk. The digital investigators grind through a company’s digital data trove of emails, documents, texts, postings, and presentations looking for specific words or phrases or combinations thereof that point to risky behavior or risk propensity. That’s where big data can produce big positive impacts, but Boolean search delivers too many false positives and false negatives to have much detection power. More on Friday concerning what works better.
Trends leading to more hiring mistakes drive the third source of Human Capital Risk. Hiring mistakes not corrected by background checks or performance management get compounded over time, leading to costly and sometimes fatal levels of Strategy Execution Risk. Just ask the former executives at Kodak, RCA, Wang, Compaq, General Foods, Arthur Anderson, Wards, American Motors, and the list goes on. Finding the very best talent for strategically pivotal roles begins with attracting the right talent pool, screening it down to a manageable number, and then selecting the best talent via final decision interviews. This post focuses on the screening and selecting bits.
Accurate screening at scale requires valid online assessment tools. At from 60 to 400 respondents for each posted position, manual phone screening just isn’t feasible. Neither has it proven to be accurate in any case. But here’s the paradox. Candidates will complete short, fun, science-free online assessments. They just aren’t accurate– (i.e. make mostly hiring mistakes with few hits). At the other end, traditional multi-item assessments with long, repetitive, text-dense testing deliver reasonable accuracy, but candidates abandon them in shockingly high numbers. Even when taken on a PC, completion rates top out at around 50%. When taken on a smart phone, completion rates for a 20+ minute process fall to single digits . So savvy hiring managers should be asking themselves, “What are my chances of hiring the top 1, 5, or 10% talent I want, when completion rates are so low?” Good question, and the number isn’t encouraging.Help is coming on Friday.
The final source of rising Human Capital Risk lies in the broken behavioral interview. I should know. I published the initial research and wrote the book on Behavior Description Interviewing back in 1986 . The problem here isn’t that behavioral interviewing doesn’t work. The problem is that hiring managers attend the interviewing courses, eat the doughnuts, and then mostly keep doing what they have always done– which is to pay much more attention to how candidates answer questions than to what they say. Hiring managers remember one or two behavioral questions from the workshop, but retreat to their tried and tired generality and opinion questions. They rarely re-direct candidates who stray from behavioral answers. They take mostly illegible notes, even to themselves, if they take notes at all. They rate answers after the interview if they rate answers at all, which most don’t. Then they wonder why they don’t get the better talent found in the published research on behavioral interviewing when the research participants actually followed best practices. Fortunately, natural language analytics called Latent Semantic Indexing offers proven power to solve the three main sources of increasing Human Capital Risk. But by now you know the drill.Stay tuned. Friday.
Plenarium Chief Science Officer Dr. Tom Janz attended the 2016 Future of Talent annual conference, US Edition, organized by Kevin Wheeler, CEO of Global Learning Resources. Held on the beautiful grounds of the Marconi Conference Center on Tomales Bay, California. This year’s gathering of 18 talent professionals featured many familiar faces from FOTs past, and several new ones. The following paragraphs hold Dr. Tom’s views, which do not necessarily reflect the official view of Plenarium or any other person or corporation.
Kevin Trend | In Kevin’s travels around the globe, which have been extensive over the past year, the innovations in talent management practices in large corporations have happened outside of the US. HR teams in the US, reflecting the corporations they serve, run too fast and lean driven by the gods of EBITDA, market cap, and free cash flow. This leaves little time or tolerance for reflecting on ‘what if’ or ‘if then’. Things have to be really broken in a way that clearly cuts into quarterly numbers or “why fix it?”
Tom’s Take | Tom has not packed on Kevin’s international air miles, but took in events consistent with his trend. One example- a large global bank chartered in the UK looks to transition its ‘vetting’ process (a.k.a. candidate screening) from the local HR function to the corporate ‘fraud and risk’ function reporting to the CFO. You see, the procurement leads noticed the many millions of dollars spent annually on complex assessment testing services provided by a comparably global test publisher. These services delivered comprehensive assessment reports on candidates– so comprehensive that most hiring managers couldn’t understand them. They knew that HR required them, so they dutifully insisted that candidates take the inventory. Then they would file the reports in the appropriate circular file and hire the candidates they liked based on their qualifications and interview performance. The reports, with their broad array of personality, motivation, interpersonal skill, and mental ability dimensions left them confused. The fraud and risk people wanted a clear number on fiduciary trust- the component that spoke to the reason for ‘vetting’ in the first place.
Kevin Trend | Automation will replace routine processes for white collar (clean hands) jobs even faster than it did for blue collar (dirty hands) jobs. Machines that perform repetitive physical tasks have already replaced millions of jobs that will never come back. Now, thinking machines replace repetitive mental tasks and are already taking on creative design tasks as well. Thinking machines with access to big data analytics may well create and deploy online services and solutions faster than people can. And Kevin sees this change as accelerating as we get deeper into it. All of this calls for HR to play a more constructive role in helping todays leaders and talent get better prepared for this coming wave of change than we were for the last big trend- Globalization.
Tom’s Take | Knowing how hard it is to deliver tangible science-constrained optimal value from analytics, I am not as concerned about the accelerating rate of automation as some are. When I drill down on people using the phrase ‘artificial intelligence’, I often find a lot more that is artificial than intelligent. But the wall of change is coming at some point. And with HR running so thin to fill openings, count heads, take temperatures, and file the paperwork, it’s going to be a challenge. The pleasant thought that almost everyone currently delivering routinized work can be ‘re-purposed’ to step up to the technically-creative-work plate may be wishful thinking. Then what? More angry, marginalized, citizens and refugees? More Brexits? People used to earning sustainable incomes with a comfortable retirement aren’t going to be happy with lower-paying service jobs that require ‘a cheerful, positive disposition that routinely exceeds customer expectations’. Drawing on the talent analytics that now shows up in so many Google searches and HR job titles, HR can model and project the future talent impacts of changing role requirements. HR can advise on the size of the future workforce, who can make the jump to light speed, and where those who may need to ‘right-size’ their career can land. It’s not nearly as much fun as finding and shaping high potentials, but it will be gravely needed. To make that happen, we will need HR professionals who understand enough about analytics to intelligently commission the work and explain the results to executives, who need a compelling narrative before they will give up their leadership prerogative to evidence-based decision processes.
Kevin Trend | HR professionals progress from a being a pair of hands & ears, through processing paperwork, past operating HRM systems (ATS, HRIS, payroll, LMS, 360° feedback, engagement surveys) to finding and communicating pivotal predictive analytics. Only then can operations executives properly prioritize human capital decisions among the other capital allocations they make and significantly raise decision hit rates. HR Professionals who get stuck at the low end of the HR value chain will be outsourced. Those who make it to the high end will earn their seat in the boardroom not by telling executives what to do (never a great career move), but by providing executives with options and consequences around the types of analytics to present, and then letting the numbers/graphics do the talking.
Predictive Analytics go beyond merely tallying up cost per hire, time to fill, revenue per employee, or retention by job class. Those are Descriptive Analytics. Predictive Analytics are more forward facing- pointing towards, for example, where the firm will face recruiting shortages in mission-critical roles. Predictive Analytics pinpoint who to hire to achieve specific strategic and tactical objectives or who will leave and the consequences for operational effectiveness.
Tom’s Take | Tom Becker, who also addressed us at FOT 2016, brought up a recent large international survey by Deloitte of CEOs. It found that of the top ten things that keep CEOs up at night, the usual talent management suspects of engagement, retention, and hiring made it in. So did talent analytics. But while the other three were down the list on urgency, CEOs also rated their organizations as reasonably ready to handle the challenge. Talent Analytics was rated as most urgent on the list, but CEOs rated their organization’s readiness to tackle it as non-existent. Why?
Being able to deliver on Talent Analytics takes more than a title. It takes expertise, integrity, and courage. HR professionals don’t get a ton of training on quantitative analytics, and probably didn’t get into the profession to exercise their stats chops. While being valuable doesn’t require being able to prove the selection utility equation or invert large matrices, it helps to know the differences between expected values, means, averages, and the first moment of distributions (psst! there aren’t any). Not everything can be discovered in 90 seconds with a Google search. Good Moocs on the topic may be one answer. We need more HR professionals with a solid grounding in both analytics and psychological measurement.
It also takes integrity, given the forces all around us that push for the fast, easy solution to difficult, complex problems. Every investment in analytics should be as simple and cost effective as possible, but not simpler. It takes integrity to draw the line when pushed to do it faster or with less data than the analysis needs to have a decent chance of finding a positive result, given that one exists.
Finally, it takes courage. Jack Kennedy said, “Too often we hold fast to the cliches of our forebears. We subject all facts to a prefabricated set of interpretations. We enjoy the comfort of opinion without the discomfort of thought” (or evidence). Leaders like what they like, and they are used to getting it. HR professionals offering the comfort of truth to replace the comfort of opinion need to involve operational executives in making the decision to seek an objective answer with a direct connection to improved business outcomes.