A Checklist for Predicting Corporate Disasters — Is Your Firm the Next BP?

You might feel far removed from the current oil disaster taking place in the Gulf of Mexico because your firm is not involved in oil drilling or sophisticated mechanical operations. However, assuming that there are not important lessons to be learned from BP’s handling of the issue would be shortsighted. Any organization experiencing a “critical incident” needs to look beyond equipment failures and natural disturbances to determine if human factors or people-management practices were an underlying or contributing cause. Apart from establishing accountability, investigations play a more strategic role in helping identify indicators of impending disaster that could serve as a warning moving forward.

If people management practices play a role in creating situations whereby a disaster is more likely, then it stands to reason that people management metrics (long collected, rarely leveraged) could provide vital insight into disaster risk.

Put more simply, HR functions have a strategic responsibility to determine if measures of overtime, absenteeism, temporary labor use, training hours attended, engagement, etc., can be used to predict potential business problems.

Do People-Management Factors Cause Failures?

In recent years, numerous scandals and disasters have demonstrated the far-reaching impact of corporate policy, and more importantly corporate practice. While firms like Enron, Bear Stearns, and “systems” best practice icon Toyota had well-documented processes, policies, and procedures, they also all engaged in management practices that led people to make bad decisions that ultimately cost the organizations billions.

Given the spotlight on bad management, it is somewhat disturbing that so few HR leaders are acknowledging their role and stepping forward with solutions to help predict and prevent disasters. Yes, most HR leaders react quickly to assist when disasters occur, but rarely if ever do they step forward to suggest that management systems may have played a major role in causing the disaster, and accept responsibility. Even fewer HR leaders attempt to proactively predict upcoming disasters.

If you were the new general manager of an NBA franchise and were presented with a set of people-metrics that revealed that your average team member is 5’9”, 40% of starters are on injured reserve, team salaries are the lowest in the league, and that scoring attempts per game result in success less than 9% of the time, would you really be surprised that the team lost every game? Would it be news to you that unless significant “talent-management” changes were made, that you would continue to lose?

While the investigation into the BP disaster continues on, like other recent disasters, it’s already clear that human factors played a major contributing role. Employees surviving admit damaging the blowout preventer prior to the incident and indicate that management routinely dismissed warnings and documented procedures to hasten making the well productive. A one-time incident BP could blame on a bad hire, but BP’s safety track record is one of the worst in the industry, indicating that a culture (established through management practice) of tolerance for sidestepping is present.

“Failure Prediction Checklist”

With the list of corporate scandals and disasters growing, it is becoming clear that people-management plays a major role in crafting situations that allow disasters to occur. What follows is a long list of people-management factors that, when modeled, can serve as predictive indicators. Not all factors will contribute to critical incidents in every industry, so investigate the linkage in your own organization and devise an appropriate set of warnings.

Short-Term Indicators (Weekly/Monthly Monitoring)

Absenteeism % — a shorthanded team or one augmented with temporary underskilled labor, particularly in maintenance and safety roles, will result in increased errors rates and activity postponement. Significant absenteeism may be symptomatic of deeper problems.
Overtime utilization – like absenteeism, chronic fatigue and stress may contribute to increased errors rates and activity postponement.
Temporary labor % — when temporary labor makes up a high percentage of a critical work team, their unfamiliarity with processes and standards may give rise to the foundation of a critical incident.
Vacancy rate — a high vacancy rate could be indicative of important work being postponed, ignored, or reallocated to underskilled labor.
Performance/engagement/satisfaction rates — low or dramatically decreasing 360°, engagement, or employee satisfaction survey results may be an indicator of future problems with an individual or a team.
Productivity rates — sudden bursts or declines in workforce productivity may be indicative of policies being violated, process steps being side-stepped, or warning measures being ignored.
Cost-containment rates — cost-containment rates that exceed planned operational buffers may increase use of shortcuts and inferior materials. Oversight is needed to ensure that short-term savings do not result in long-term costly disasters.
Whistleblowing incidents — having no whistleblowing reports can be an indication that employees lack interest, have experienced retribution in the past, are unaware of how to report issues. Regardless, both too few and an abundance of reports may be an indicator of serious problems.
Accident rates — not only are they costly, but they may be an indicator that procedures are not being followed. Unresolved safety issues, the routine violation of safety rules, or lax safety training may all be a predictor of larger safety issues.
Error rates — although errors are not entirely a human factor, a high error rate may indicate that procedures are not being followed or that the training is lax. Delaying or not responding at all to major errors may be the most critical indicator.

Longer-Term Indicators (Quarterly/Annual Monitoring)

Turnover/transfer rates — managers with high employee turnover rates or high employee “transfer out” rates may be weak managers prone to higher rates of major problems.
Complaint levels — when complaint or grievance levels increase, it may be a predictor of upcoming major problems.
Training — an untrained team, an unevenly trained team, or one with “dated” training can dramatically increase short-term error rates and eventually lead to major disasters. Lax training of supervisors and managers may be a separate but more serious problem.
Staffing levels — whenever the ratio of employees to work falls below a certain point, fatigue and stress may eventually lead to major errors.
Management span of control — when the ratio of managers to employees exceeds a certain level, the likelihood of poor supervision increases dramatically. In the other direction, the likelihood of the micromanagement of employees also increases.
Performance management issues — an overall high percentage or an increase in the percentage of individuals on a performance management plan may be an indication of bad hiring or poor supervision. In contrast, having an extremely low percentage may also be an indication that management is lax and that it is unwilling to acknowledge and punish poor performance.
Excess accumulated vacation — an excess of accumulated vacation may mean that workers are overworked.
Managers fail to meet goals — a manager with a track record of failing to meet goals may continue that pattern unless proactive action has been taken. New managers unfamiliar with practices may cause more errors, as may a long-term manager that has grown complacent.
A history of problems — the mere existence of a pattern of problems in the past may be the best indicator of future problems in a team or business unit. (It has been widely reported that BP racked up more than 760 safety violations, while Exxon received only one in the same time period.)
Poor internal handoffs — whenever complicated interdependent work is handed-off between different internal functions or teams, the likelihood of an error increases dramatically. Poor or nonexistent coordination and integration mechanisms are an indication of continuous problems.
Disability rates — high short-term or permanent disability rates may be an indication of both current and future problems, if they are the result of high stress and poor working conditions.
Recruiting metrics — a slow time to fill and a weak quality of hire may be precursors to future problems.
Pay rates — when average pay rates are significantly below the prevailing norm, the likelihood of weak hires and unmotivated workers increases.
Incentives — whenever bonuses or incentives exceed 25% of base pay, there is a significant chance that the high incentive will drive employees and managers to violate rules or ethical standards.
Outsourcing % — miscommunication and poor coordination are always issues when critical work is outsourced. When a significant percentage of critical or sophisticated work is assigned to vendors, the handoff process may also be a source of major problems.
New procedures — whenever new or significantly revised processes or procedures are put into place, the odds of a significant error increases tremendously.
New technology — whenever major new technology is implemented, the likelihood of a significant failure increases dramatically.
No forecasting algorithm — the lack of a statistical process that forecasts and provides alerts by itself may be an indication of upcoming problems.
No if-then scenarios — the lack of a routine process for testing managers and employees on how they will handle likely worst-case scenarios may be a predictor that solutions will be slow in coming when problems do occur.
A rigid company culture — although there may be many benefits from having an established company culture, a rigid one may also be an indication of inflexibility and overconfidence that problems will not occur.

Review past major incidents in your own organization to determine which factors are indicative of increased risk of incident to come. Rank and weight each, then set a warning threshold based on safe operations. Be conservative as too many warnings are better than no warnings.

Future Developments

While simple analysis and a checklist are a decent start at being prepared, much more robust systems are needed that can analyze thousands of pieces of data in real time to determine where risks exist. The practice of analyzing risk isn’t new; many corporations already have a risk management function, but the focus of risk assessment related to people analytics is relatively new and rare. Risk modeling involves gathering massive volumes of data and applying sophisticated statistical models, the latter activity being one that few in HR are good at. As Dan Hilbert, founder of OrcaEyes and a pioneer in the space of risk modeling related to workforce factors, can attest, few in HR “get it.” When demoing his suite of business intelligence solutions, he often finds that while finance and operations professionals get it immediately, many in HR are overwhelmed and dismissive. The demand for robust workforce risk modeling is most certainly present in organizations, but the interest to fill that demand isn’t coming from HR.

Final Thoughts

“All the things that they told us could never happen, happened.”

This was the bottom line observation of Mike Williams, the chief electronics technician who was on board the ill-fated Deepwater Horizon drilling rig at the time of the explosion. Take heed of his observation and develop a process to more accurately predict upcoming problems in your organization. Start by realizing that there are precursors or early warning signs to most problems. The key is to develop a process for identifying these precursors and using them to trigger action early enough to prevent minor problems from becoming major problems. Finally develop a checklist of these factors and periodically use the checklist to identify potential problem areas.

Note: What additional people-management factors could serve as precursors and thus should be added to the checklist? Add yours using the comments function.