Using Precursors for Workforce Planning

There are three parts of all successful workforce plans. It turns out that the three parts are not all equal, however. The first part, which is the workforce forecast, is the most important. It projects company growth and the corresponding change in our talent needs as a result of that growth. In most cases, if the forecasts are inaccurate, the whole workforce plan fails. To say that statistical forecasting the future is “difficult” is certainly an understatement. But forecasting the future growth of a company or an industry is a key element in successful workforce planning. Some organizations have entire staffs that do nothing but forecast future growth patterns. If your firm has a strategic planning unit, you might be able to use their business and economic forecasts for your workforce planning. If you don’t have access to other forecasts ó or if you don’t trust their accuracy and hate statistics ó there is another approach to forecasting that you can take. A Different Approach to Forecasting: From Complex to Simple Before you begin forecasting business upturns and downturns for workforce planning you need to assess you own level of sophistication and access to data. Large HR departments can and should adopt more sophisticated (read: statistical) approaches, while individual managers should look for easier-to-execute approaches. There are three basic approaches to forecasting for workforce planning. They are listed here in order from the most sophisticated to the simplest:

  1. The statistical approach. A small number of large firms use sophisticated statistical models to make forecasts. They use internal and external statistics and, through regression analysis, identify the most likely forecast. This approach is generally well beyond the interest or capability for all but Fortune 200 firms:
  2. Identify precursors, or leading indicators. If you want to be precise but you don’t have the luxury of a forecasting staff there is an alternative approach which is accurate, easy-to-use and cheap. That approach is known as the precursor (or “leading indicator”) approach, which identifies patterns in your industry and firm and then uses these as predictors of when growth will come and, to some degree, how much growth to expect.
  3. “Gut instinct.” Some experienced senior managers just seem to know what will happen next. Some see a variety of new scenarios, while most just see history repeating itself. I know some managers who are uncannily accurate in this type of forecasting. But unless you have tracked your accuracy over time, I don’t recommend you try this approach in a rapidly changing industry.

If you are new to workforce planning or have limited resources, I recommend you avoid the statistical and “gut” approach and go right to the “precursor” approach for at least your first two years. The Four Elements of Workforce Forecasts Workforce forecasts provide managers with “heads up” projections. Both of the realistic recommend approaches (statistical or precursor) have four basic elements that they forecast. They are:

  1. Company growth.This element forecasts or projects the growth rate of the business. If you don’t accurately know how much a business will grow or contract, you are liable to either “over hire” or to end up with a “surplus” of employees. Both are expensive propositions.
  2. Talent (workforce) needs. The second element of workforce forecasts outlines your “needs” (it is sometimes also called demand). It tells you the number and type of employees your company will need in the future to meet the forecasted business growth rate.
  3. Projected vacancies. The vacancy forecast provides recruiters and managers with information about projected vacancies and how many of your current employees will need to be replaced.
  4. Supply of talent. The availability or “supply” element informs managers about how much talent you can realistically get through external recruiting and what talent must be developed internally.

The Precursor Approach: A Simpler Way To Forecast Upturns and Downturns If you hate statistics but are weary of gut projections (both are true for most people) I recommend you utilize the precursor approach to forecasting. The precursor approach is based on the principal that most things in business are interrelated and occur in repeating patterns. Before major events occur, there are minor events, or precursors, which routinely come immediately before these major events. The U.S. government tracks a list of leading economic indicators (LEIs) on a monthly basis to help businesses identify significant economic shifts. Some other general examples of precursors include:

  • There’s always a pre-game show before the Super Bowl.
  • Toy shelves are well stocked two months before a holiday. When toy store shelves become close to bare, a major holiday is coming within weeks.
  • Before long-term employees retire, they call the benefits department to check on their retirement benefits.
  • Before overall industry sales increase after a long slump, certain firms (usually the most profitable) begin building up production capacity.
  • Before most customers begin buying after a downturn, one customer (usually the most well-off) begins making purchases again.
  • Before a major computer company significantly increases its production of computer chips, it buys the latest model chip manufacturing equipment from Applied Materials (a company that manufacturers equipment needed to manufacture computer chips).

Precursors Work Much Better When… There are some situations where precursor forecasting works especially well. These include:

  • Lag firms: If you are a small firm in an industry that “follows” the lead of the larger
  • Region: If you are in a geographic region that lags behind other regions (for example, fashion trends in Kansas might be six months behind New York or L.A.)
  • Slow recovery product: If you are in an industry whose products are often the last to recover in an upturn (e.g., jewelry or luxury cars)
  • Conservative management: If your management team is conservative and traditionally waits until several other first-mover firms act
  • Large capital requirements: If you are in an industry where large capital equipment or land purchases must occur before production can ramp up (e.g., semiconductors or high-rise construction)
  • Approval lead time: If you are in an industry that requires a long lead time for governmental (or other) approvals (for example, pharmaceuticals and medical equipment)

In each of these cases, because you are not “first,” it is easier for HR professionals to judge what number of headcount will be needed and when. Identifying Precursors as Warning Signs The overall goal of precursor forecasting is to identify upcoming industry upturns and downturns early enough to be able to adjust your workforce in time. The process of identifying precursors for forecasting talent needs starts with viewing industry upturns or downturns as historical patterns of connected events. The precursor occurs in the early part of the pattern. When it does occur, it essentially “warns” you about the likely following events. You can begin to identify this connected pattern by looking at events over the last few upturns and downturns. Quite often these patterns and precursors are obvious, so it doesn’t require a statistical analysis to identify them. The precursor’s pattern might not be well known, because few managers have ever taken the time to “step back” and look at the big picture. But the process for identifying precursors goes something like this:

  1. Start by asking several “veterans” of the industry (independently) what the patterns and the precursors are to a business upturn or downturn. Many times there is consistent agreement between these veterans.
  2. Typical precursors to a return to industry growth may include:
    • Increased purchasing of capital equipment by key firms
    • Increased sales by optimistic firms
    • Increased hiring
    • New advertising expenditures
    • More overtime hours
    • Relaxed spending on training and travel
    • Pay increases
  3. You can stop with the “opinions,” but I recommend you go back and check the records of business trends to verify the obvious patterns.
  4. Next, “plot” the increase and decrease in firm growth on a chart for each major upturn and downturn (using a line graph with sales on the side and months on the bottom). Mark every key event (increase in hiring, cut in overtime, rise in advertising by lead firms, etc.) on the chart on the month that it occurred. The chart will then give you a good visual idea of whether there are patterns and precursors for economic upturns or downturns. (For example, increased capital equipment purchases of 10% or more might occur one year before sales turn up by 25%.)
  5. Use the charts to estimate when (i.e., how many months prior) and how large the precursors must be before the “following event” occurs.
  6. The next step is to closely examine the pattern over the last three to ten years to make sure there are no obvious exceptions to the patterns. An exception may be, for example, that the pattern does not hold during a time of war or when oil prices are over $25 a barrel. This is important, because there can be “false indicators,” or times where normal indicators of growth are misleading.
  7. Next, attempt to quantify (usually the minimum level) that each precursor must meet in order to be a true indicator. (For example, capital equipment purchases must rise by 10% in a quarter.)
  8. Combine or group together all of the precursors that you have found to be valid indicators of growth (or recession) into a single index or overall precursor pattern.
  9. Next “test” the overall precursor pattern or index on the historical data from the last significant industry or company upturn or downturn to see how accurate it would have been at forecasting these events.
  10. Run your precursors by the chief planner, the GM, the CFO, etc. to make sure they are in the ballpark.
  11. Finally use the identified precursor patterns to forecast upcoming company or industry upturns or downturn

If you have the time and are good with data, you can normally quantify the precursors to the point where you can both forecast an upturn or downturn and also estimate how much of an increase or decrease in growth you can expect. If you are statistically challenged, another alternative to precise qualification is to instead to identify a “range” of likely growth rates. By developing a plan that prepares for all growth rates within that “reasonable range” you can avoid the need for precise forecasts and cumbersome statistics. Some organizations may choose to combine the two approaches by both estimating a specific growth target and also preparing for a reasonable range of variations from that growth estimate. Although there are some events that don’t have precursors (like earthquakes), most of the ones that must be forecast in business and workforce planning (like changes in sales, production and headcount) do generally have reliable precursors. One final note: because patterns can occasionally change, remember that precursors are designed to give you a heads-up warning. Not every lead indicator needs to “go off” before you need to take a closer look. Treat precursors like smoke detectors. When even one goes off, there might not actually be a “fire,” but you should treat the warning as a “not to be missed” opportunity to examine the situation more closely using your insight and experience to guide you. In the coming weeks, we’ll take a more detailed look at the each of the four elements (company growth, talent needs, projected vacancies, supply of talent) that go into a complete workforce plan.

About Dr John Sullivan

Dr John Sullivan is an internationally known HR thought-leader from the Silicon Valley who specializes in providing bold and high business impact; strategic Talent Management solutions to large corporations.

Check Also

Woman Wearing Blue Top Beside Table Woman Wearing Blue Top Beside Table Interviewing Woman in White Top

Improve Your Interviews With Five Easy Additions (And identify top candidates faster)

Unstructured interviews don’t predict very well, so add a little structure in 5 key areas.  …