# Monitoring recovery after a product harm crisis.

Management has a responsibility and an interest in monitoring
operations. In the specific case of a product harm crisis, management is
especially interested in monitoring performance measures relating to recovery. Recovery rates indicate the success of management strategies
in dealing with a crisis. The extent to which goals and timetables are
being adhered determines changes in management strategies. The
determination of performance measures is critical in evaluating
management's effectiveness in dealing with the post-crisis
situation.

Performance measures

There are several types of performance measures that may be of interest to managers. In the case of product recall followed by an introduction of a replacement, market share could be the performance measure of concern. Recapturing pre-crisis market share according to some timetable is probably in management's greatest interest. In the case of product recall without replacement, overall company or SBU sales could be the performance measure of concern. Here, management is interested in regaining overall pre-crisis company or SBU sales; in addition, management is interested in minimizing the negative effects of the crisis involving one brand on the rest of the product line. In the situation of no recall, management would be interested in both of the above measures in addition to the company's stock price. Of course, stock price could be a performance measure under all scenarios. Moreover, goals and timetables are appropriate for all scenarios.

In the product recall case with replacement, if market share fell from 30 percent to zero, an appropriate goal could be to return to a 25 percent market share by the ninth month after the replacement. In the case of product recall without replacement, if overall company sales fell from $250 million to $220 million, then a relevant goal might be to recoup half of the lost sales within a year after the crisis. Finally, in the case of no recall, if the company's stock price is the performance of concern, a return to the pre-crisis stock price within three months might be a suitable goal.

A model for monitoring recovery

Whatever the performance measure, it is a function of time and other parameters; as such, it is a special case of transient analysis. In general, the performance measure at time t after the crisis ([P.sub.t]) is:

[P.sub.t] = f(t, ...) [1]

If the function f exists, and if

[g.sub.t] = f(t.sub.g) [2]

then,

[t.sub.g] = [f.sup.-1]([g.sub.t]) [3]

where:

[g.sub.t] = the goal at time [t.sub.g]

In many applications using transient analysis, the exponential function is used as a first approximation. Proceeding in this manner, of using exponential recovery, we have:

Pt = H - (H-L)[e.sup.-at] [4]

where:

H = the pre-crisis level of the performance measure

L = the immediate post-crisis level of the performance measure

a = the exponential parameter

Several viewpoints on the use of this model are now presented.

Let us assume that the pre-crisis market share value (H) was 30 percent; the post-crisis market share value (L) after recalling the product from, say, the New England states was 10 percent. Three months after the recall, the market share ([P.sub.3]) went up to 15 percent. From [4] the exponential parameter a is given by

a = [[sup -1]1n[(H-L)/(H-([P.sub.t)] [5]

Using the above values, a is determined to be 0.095894.

Let us suppose that the company wishes to have a market share of 25 percent by the end of the ninth month after the crisis. Using [5], with t=9 and P=25 percent, the value of a is determined to be 0.154. Since this value is greater than the value of "a," previously determined (0.095894), it is concluded that the company would miss its goal and timetable if the exponential parameter based upon the three-month data continued. It is possible to then determine the extent of the "miss."

From[4], the value of market share based upon the exponential parameter equaling the value of 0.095894 and a nine-month period is equal to 21.56 percent. From [5], the time to achieve the goal of 25 percent market share can be calculated; the result is 14.4565 months. Thus it is apparent that after nine months the shortfall in market share is 3.44 percent (25 - 21.56 percent), and the time to achieve the goal of 25 percent market share is exceeded by more than 5 months (14.4565 - 9).

For possible interest, the case of a "halfway" recovery is presented. In this situation, the goal is (H+L)/2 and the time is noted [t.sub.g] From[5], the value of the exponential parameter ([a.sub.g]) is determined to be 0.693/t

In addition to the exponential recovery model, other models exist, e.g., the logistic model (see Naert and Leeflang 1978 for an extensive presentation of several possible recovery models).

Indices for implementation

The output from the monitoring process includes actual performance measures and the time of their occurrence ([A.sub.t]). These values when used with management's goals and timetables ([G.sub.t]), can "flag" situations that require management actions. Three indices are presented; they are a ratio index (I), a modified ratio index (II) and a subtractive index (III).

The first index under consideration is a basic ratio index ([A.sub.t]/[G.sub.t]), noted I. For example, if ([A.sub.t]/[G.sub.t] > 1, then the recovery is better than planned by the goals and timetable schedule. This is an illustration of a ratio index. If this ratio index is less than 1, then the recovery is not proceeding according to plan. However, if the ratio index is 0.98, most managers would consider this good enough; if the ratio was as low as 0.75, this would be a sure indication that action should be taken.

This basic ratio index I can be modified to take into account the starting point (L) after the crisis; this is noted index II:

[A.sub.t] - L)/([G.sub.t] - L)

In the case that H = 30 percent, L = 10 percent, and [A.sub.t] = 15 percent, index I equals 15/30 or 50 percent, and index II equals 25 percent. From the viewpoint of index I, the company has achieved half of the scheduled target market share. However, from the viewpoint of index II, the market share achievement is only 25 percent of the scheduled recovery. This is because only 5 percent of the full recovery of 20 percent has been achieved.

A third index III is subtractive; viz., [G.sub.t]-[A.sub.t]

In the above example, index III would equal 15 percent shortfall. The drawback of this index is that it is negative when the company is doing better than its scheduled recovery.

The three indices presented here are not meant to be exhaustive; others can be used. However, any index selected should be used consistently. Action values may be dependent upon where in the time frame of the recovery period the indices were developed. For example, if index II was equal to 25 percent in month 2, it would not be as serious as if it had occurred in month 8. If index I were equal to 98 percent, it probably would not be time sensitive.

Summary

This paper presented an approach to monitoring recovery after a product harm crisis. Suggestions for performance measures were given; they included market share, sales and company stock price. A model for monitoring recovery was developed. The specific case of exponential recovery rate was used. Three indices for implementation were suggested; these were I) a ratio index, II) a modified ratio index and (III) a subtractive index. A numerical example has been included.

For further reading

Naert, Philippe and Peter Leeflang, Building Implementable Marketing Models, Boston: Martinus Nijhoff Social Sciences Division, Leiden, 1978.

Performance measures

There are several types of performance measures that may be of interest to managers. In the case of product recall followed by an introduction of a replacement, market share could be the performance measure of concern. Recapturing pre-crisis market share according to some timetable is probably in management's greatest interest. In the case of product recall without replacement, overall company or SBU sales could be the performance measure of concern. Here, management is interested in regaining overall pre-crisis company or SBU sales; in addition, management is interested in minimizing the negative effects of the crisis involving one brand on the rest of the product line. In the situation of no recall, management would be interested in both of the above measures in addition to the company's stock price. Of course, stock price could be a performance measure under all scenarios. Moreover, goals and timetables are appropriate for all scenarios.

In the product recall case with replacement, if market share fell from 30 percent to zero, an appropriate goal could be to return to a 25 percent market share by the ninth month after the replacement. In the case of product recall without replacement, if overall company sales fell from $250 million to $220 million, then a relevant goal might be to recoup half of the lost sales within a year after the crisis. Finally, in the case of no recall, if the company's stock price is the performance of concern, a return to the pre-crisis stock price within three months might be a suitable goal.

A model for monitoring recovery

Whatever the performance measure, it is a function of time and other parameters; as such, it is a special case of transient analysis. In general, the performance measure at time t after the crisis ([P.sub.t]) is:

[P.sub.t] = f(t, ...) [1]

If the function f exists, and if

[g.sub.t] = f(t.sub.g) [2]

then,

[t.sub.g] = [f.sup.-1]([g.sub.t]) [3]

where:

[g.sub.t] = the goal at time [t.sub.g]

In many applications using transient analysis, the exponential function is used as a first approximation. Proceeding in this manner, of using exponential recovery, we have:

Pt = H - (H-L)[e.sup.-at] [4]

where:

H = the pre-crisis level of the performance measure

L = the immediate post-crisis level of the performance measure

a = the exponential parameter

Several viewpoints on the use of this model are now presented.

Let us assume that the pre-crisis market share value (H) was 30 percent; the post-crisis market share value (L) after recalling the product from, say, the New England states was 10 percent. Three months after the recall, the market share ([P.sub.3]) went up to 15 percent. From [4] the exponential parameter a is given by

a = [[sup -1]1n[(H-L)/(H-([P.sub.t)] [5]

Using the above values, a is determined to be 0.095894.

Let us suppose that the company wishes to have a market share of 25 percent by the end of the ninth month after the crisis. Using [5], with t=9 and P=25 percent, the value of a is determined to be 0.154. Since this value is greater than the value of "a," previously determined (0.095894), it is concluded that the company would miss its goal and timetable if the exponential parameter based upon the three-month data continued. It is possible to then determine the extent of the "miss."

From[4], the value of market share based upon the exponential parameter equaling the value of 0.095894 and a nine-month period is equal to 21.56 percent. From [5], the time to achieve the goal of 25 percent market share can be calculated; the result is 14.4565 months. Thus it is apparent that after nine months the shortfall in market share is 3.44 percent (25 - 21.56 percent), and the time to achieve the goal of 25 percent market share is exceeded by more than 5 months (14.4565 - 9).

For possible interest, the case of a "halfway" recovery is presented. In this situation, the goal is (H+L)/2 and the time is noted [t.sub.g] From[5], the value of the exponential parameter ([a.sub.g]) is determined to be 0.693/t

In addition to the exponential recovery model, other models exist, e.g., the logistic model (see Naert and Leeflang 1978 for an extensive presentation of several possible recovery models).

Indices for implementation

The output from the monitoring process includes actual performance measures and the time of their occurrence ([A.sub.t]). These values when used with management's goals and timetables ([G.sub.t]), can "flag" situations that require management actions. Three indices are presented; they are a ratio index (I), a modified ratio index (II) and a subtractive index (III).

The first index under consideration is a basic ratio index ([A.sub.t]/[G.sub.t]), noted I. For example, if ([A.sub.t]/[G.sub.t] > 1, then the recovery is better than planned by the goals and timetable schedule. This is an illustration of a ratio index. If this ratio index is less than 1, then the recovery is not proceeding according to plan. However, if the ratio index is 0.98, most managers would consider this good enough; if the ratio was as low as 0.75, this would be a sure indication that action should be taken.

This basic ratio index I can be modified to take into account the starting point (L) after the crisis; this is noted index II:

[A.sub.t] - L)/([G.sub.t] - L)

In the case that H = 30 percent, L = 10 percent, and [A.sub.t] = 15 percent, index I equals 15/30 or 50 percent, and index II equals 25 percent. From the viewpoint of index I, the company has achieved half of the scheduled target market share. However, from the viewpoint of index II, the market share achievement is only 25 percent of the scheduled recovery. This is because only 5 percent of the full recovery of 20 percent has been achieved.

A third index III is subtractive; viz., [G.sub.t]-[A.sub.t]

In the above example, index III would equal 15 percent shortfall. The drawback of this index is that it is negative when the company is doing better than its scheduled recovery.

The three indices presented here are not meant to be exhaustive; others can be used. However, any index selected should be used consistently. Action values may be dependent upon where in the time frame of the recovery period the indices were developed. For example, if index II was equal to 25 percent in month 2, it would not be as serious as if it had occurred in month 8. If index I were equal to 98 percent, it probably would not be time sensitive.

Summary

This paper presented an approach to monitoring recovery after a product harm crisis. Suggestions for performance measures were given; they included market share, sales and company stock price. A model for monitoring recovery was developed. The specific case of exponential recovery rate was used. Three indices for implementation were suggested; these were I) a ratio index, II) a modified ratio index and (III) a subtractive index. A numerical example has been included.

For further reading

Naert, Philippe and Peter Leeflang, Building Implementable Marketing Models, Boston: Martinus Nijhoff Social Sciences Division, Leiden, 1978.

Printer friendly Cite/link Email Feedback | |

Title Annotation: | Crises in Management |
---|---|

Author: | Kabak, Irwin W.; Siomkos, George J. |

Publication: | Industrial Management |

Date: | May 1, 1992 |

Words: | 1317 |

Previous Article: | Design for Manufacturability: a survival strategy for the American manufacturing industry. |

Next Article: | Examine costs of system quality via perfect process analysis. |

Topics: |