Workforce planning leads to a new approach to talent management. Using simulations based on different assumptions as a way to create hiring forecasts can help organizations better understand -- and plan for -- their talent needs.
A new approach to talent management begins with workforce planning.
As I discussed in my last column, the new problem that talent management faces stems from uncertainty, the virtual impossibility of knowing what business demand and talent needs will be years into the future.
In the period before the 1980s, talent management took a bureaucratic and long-term form because business demands were highly certain and predictable. When those demands become uncertain, the talent practices have to change as well, and those changes begin with workforce planning.
Recent surveys suggest that up to two-thirds of U.S. employers do no workforce planning of any kind, no forecasting of how many employees they will need and what roles they need to be in. (A generation or so before, roughly 90 percent of U.S. employers not only did workforce planning but had a dedicated "manpower planning" function.)
Failing to do workforce planning in practice means organizations have to rely almost entirely on outside hiring because internal development requires some planning to execute.
Relying entirely on outside hiring has lots of drawbacks -- including undercutting morale and retention of current employees and, more recently, facing a tight labor market where costs are rising and supply is uncertain.
From Forecasts to Simulations
If outside hiring has reached its limits and the traditional planning approach doesn't work, what's the solution? It begins with acknowledging the risks that our forecasts will be wrong.
One approach that leading companies such as Dow Chemical and Capital One have taken to this problem has been to move away from efforts to generate a specific forecast of future demand and instead moving to develop simulations of future demand.
The difference is that simulations generate lots of estimates based on the different assumptions that operating managers might hold: "Tell us your best guess about your business plans and we'll tell you how many and what type of talent you need. Now give us your second best guess and we'll generate a new estimate."
The idea is to get at what statisticians refer to as robustness -- what is the range of likely outcomes for talent demand? A hidden benefit from this approach is that HR gets to influence the business plans of the operating managers: "Your strategy will require this many outside hires and internal promotions. Are you sure you can handle that?"
A "Deep Bench" or "Inventory?"
The next step in forecasting, and one that few companies yet have tackled, comes directly from supply-chain management: What happens when our best estimate is wrong, as it surely will be?
The odds that we will be able to predict the number of new people we need exactly is about zero. So what does it cost us if we are wrong?
We can be wrong in two ways: too little talent, which can get in the way of getting business done and too much talent or surplus employees, leading to layoffs.
Fortunately, calculating these "mismatch" costs is straightforward: What would we have to do if we have too little talent or too much, and how much would that cost us?
A generation ago, the big concern was falling short on talent needs because there was no just-in-time alternative. "Going long" on talent was less of a concern because excess talent could be parked on a bench where it would wait until needed.
Now the situation is reversed.
There are lots of alternatives to find just-in-time talent -- outside hiring, temps, contractors, etc. -- as long as the amount of such talent is relatively small. Going long means maintaining a "deep bench," which is another term for inventory.
The reason that an inventory of talent is so expensive now is because of attrition. The best way to lose valuable employees is to give them skills and then ask them to sit on a bench and wait to use them, turning down the calls from headhunters along the way.
Good business practice requires minimizing the sum of both of these costs. We do that by moving in the direction of the risk that has the lower cost. For most employers now, that means undershooting our best estimate of demand to avoid the risk of going long.
If our best guess is that we will need 100 additional middle managers next year, we might want to plan on developing only 90 to avoid going long. There is a good chance that we will have to hire a few middle managers on the outside market if we fall short.
If our estimate of demand is really uncertain -- we could need as few as 80 and as many as 120 -- we might need to develop only 80 internally to avoid the risk of having surplus talent. Then if we fall short, as we likely will, we make up the gap with outside hiring, contract work or temps.
A more realistic and cost-effective approach to management talent begins by addressing, not ignoring, the uncertain environment in which we operate.
(Editor's Note: Peter's next column takes the next step, describing how employers should organize talent management.)
Peter Cappelli is the George W. Taylor Professor of Management and director of the Center for Human Resources at the Wharton School of Business. www.talentondemand.org.