Manzana Insurance

Manzana Insurance is the second largest Insurance Company in the Property Insurance space in California. The case relates to one particular branch, the Fruitvale branch, which has been losing market share to its competitor Golden Gate. This branch deals with the commercial property insurance and has three underwriter teams that cater to 3 geographic territories. On an average, the Fruitvale branch receives around 39 requests per day (22 requests for new insurance and 17 renewals). The underwriting teams are supported by distribution clerks, the rating team, and the policy writers. ‘Requests for New Policies-RUNs’ are considered to be the most profitable business, followed by ‘Requests for Price-RAPs’. ‘Requests for Additional Insurance-RAINs’ are considered third and ‘Requests for Renewal-RERUNs’ are placed last. The employees process requests on a FIFO basis within these prioritized categories of requests.

The Problem:

– High turnaround time (TAT) – Loss of renewals business to competitors

– Worsening finances over the years; the branch reported a loss of US$ 174,000 and US$ 121,000 during the first two quarters of 1991

– Backlog of Policies

– Improper load balancing among employees – resulting in tight schedules and idle time

– Pressure due to competitor Golden Gate’s assurance of a single day turn around time (TAT)

The Objective:

To identify bottlenecks in Manzana’s operations using process flow analysis.

The Analysis:

Manzana is a service company and hence, it has a constraint of not being able to store its products for supply unlike a manufacturing company. In this case the turnaround time or TAT defines the service quality. Manzana has a problem of not being able to renew its insurance policies on time as a result of which, many clients choose to go elsewhere. Further, in a service industry the peak demand creates backlogs and these, when allowed to be carried forward, can affect the future services of the company. In this case Manzana has bottlenecks in its underwriting department. The VP of underwriting operations calculates the capacity of the departments based on the mean time required to process the requests and claims that the Fruitvale branch has enough capacity to handle the requests.

However the problem with such an argument is that it does not consider the fact that some requests have deadlines before which they must be completed. Hence one cannot calculate capacities based on mean time to service requests but instead calculate the capacities based on the 95% SCT (Standard Completion Time). An SCT of 95% means that only 5% of the requests will not take more than that time to be completed. So we have recalculated the capacities of each of the departments based on the 95% SCT.

We have been provided data pertaining to the processing requests during 6 months in 1991. During these 6 months, the firm would have had 120 working days (1month – 20 working days). We can thus calculate the average number of requests of each type per day.

Number of requests during the last 6 months

Territory 1 Territory 2 Territory 3 Total

Agents 23 26 27 76

Policies in force 1151 1393 1402 3946

RUNs originating as RUNs 162 100 88 350

RAPs converted to RUNs 112 79 83 274

Total RUNS 274 179 171 624

Total RAPs 761 513 524 1798

RAPs not converted to RUNs 649 434 441 1524

RAINs 196 125 130 451

RERUNs 636 840 605 2081

Renewals lost 403 227 296 926

Average Requests per day

Territory 1 Territory 2 Territory 3 Total

RUNs originating as RUNs 1.35 0.83 0.73 2.92

RAPs converted to RUNs 0.93 0.66 0.69 2.28

Total RUNS 2.28 1.49 1.43 5.20

Total RAPs 6.34 4.28 4.37 14.98

RAPs not converted to RUNs 5.41 3.62 3.68 12.70

RAINs 1.63 1.04 1.08 3.76

RERUNs 5.30 7.00 5.04 17.34

Renewal lost 3.36 1.89 2.47 7.72

Total Requests per day 14.63 13.15 11.23 39.00

Now we can analyze the bottleneck by calculating the number of hours the various departments would have to spend to handle the requests that they receive on any given day. (Consider the 95% SCT for any given task)

Distribution Underwriting Rating Policy

RUNs originating as RUNs 128.1(mins/req) *

2.92(req/day) 107.2(mins/req)

2.92(req/day) 112.3(mins/req)

2.92(req/day) 89.3(mins/req)

* 2.92(req/day)

RAPs converted to RUNs 107.8 * 2.28 87.5 * 2.28 88.7 * 2.28 89.3 * 2.28

RAPs not converted to RUNs 107.8*12.70 87.5 * 12.70 88.7 * 12.70 0 * 12.70

RAINs 68.1*3.76 49.4 * 3.76 89.4 * 3.76 72.1 * 3.76

RERUNs 43.2 * 17.34 62.8 * 17.34 92.2 * 17.34 67* 17.34

Total Time (in minutes) 2993.93 2898 3591 1897

Total Time (in hours) 49.90 48.31 59.86 31.62

Available (assuming 7.5 hrs of work a day) 30 22.5 60 37.5

Hence, the Underwriting team is clearly unable to handle the necessary volumes and is becoming the bottleneck. Further, the distribution team too faces a problem as it has to handle 49.9 hours of work, whereas it has only 30 hours of capacity.

The Calculations for the capacity of a department can be done as follows:

The weighted average processing time can be calculated by using the 95% SCT times for various tasks taking the average number requests per day of each of the tasks as the weights.

The Weighted average processing time for Distribution =

(128.1 min * 2.92 + 107.8 * 2.28 + 107.8 * 12.70 + 68.1 * 3.76 + 43.2 *17.34)

/ (2.92 + 2.28+ 12.70 + 3.76 + 17.34) = 76.77 Minutes/ Request

Hence the number of requests that can be processed per hour = 0.781 by one person. As there are 4 distribution clerks, they can together process 3.126 requests per hour. In a similar manner, we can calculate the capacities for the other divisions. The following process flow diagram shows the capacities:

Note: For underwriting the capacity is 2.422 requests for 3 teams

Here the Underwriting department is the bottleneck as it can process only 2.422 requests per hour. So in 7.5 hours, it can process only 18.165 requests, whereas the demand rate is 39 requests.

Demand Rate per day Processing Capacity Increase in

Capacity required

Distribution 39 Requests 23.447 requests 15.553

Underwriting 39 Requests 18.165 requests 20.835

Ratings 39 Requests 39.09 requests ——–

Policy Writers 39 Requests 46.251 requests ——–

Note: if we use the mean time of processing to do the calculations (instead of 95% SCT), it would throw up a different picture as given below:

Demand Rate per day Processing Capacity

Distribution 39 Requests 43.90 requests

Underwriting 39 Requests 47.53 requests

Ratings 39 Requests 51.13 requests

Policy Writers 39 Requests 41.05 requests

E.g. For distribution, weighted average processing time for request = 41 minutes. So in one hour it can process 1.46 requests per clerk. In a typical day of 7.5 hours with 4 clerks, we can assess the capacity as 43.9 requests/ day.

Thus we must realize that while it is possible to handle all the requests at leisure with the current capacity, Fruitvale cannot meet deadlines with this capacity. The branch thus witnesses an increase in backlogs and the occurrence of late RERUNs.

If we split the Underwriting Operation requirement among the three territories, we get the following data:

Territory 1 Territory 2 Territory 3 Total

RUNs originating as RUNs 107.2 107.2 107.2

RAPs converted to RUNs 87.5 87.5 87.5

RAPs not converted to RUNs 87.5 87.5 87.5

RAINs 49.4 49.4 49.4

RERUNs 62.8 62.8 62.8

Total Time (in minutes) 1113 954.5 830.8 2898.427

Total Time (in hours) 18.55 15.91 13.85 48.31

Available 7.5 7.5 7.5 22.5

Thus clearly, territory 1 is the worst affected by the bottleneck.

Conclusion

To eliminate this bottleneck we need to add an underwriting team.

The additional Underwriting teams required to handle the load has been calculated as follows:

In one hour 3 teams can process only 2.422 requests or 0.8073 requests per team per hour. To process the balance 20.835 requests (of the 39 requests received daily), we need 20.835/ 0.8073 i.e., 25.80 team hours. As each team works for 7.5 hours each day, 3 Underwriting teams have to be added and the 6 teams together would be able to handle the daily demand of 39 requests.

If 3 Underwriting teams are added, the bottleneck would then shift to Distribution.

No of Distribution clerks needed:

No of requests that can be processed per hour = 0.781 by one person. To process the balance 15.553 requests, we need 15.553/ 0.781 = 19.91 man-hours. As each Distribution clerk works for 7.5 hours each day, we need to add at least 2 (19.91/ 7.5= 2.65) Distribution Clerks to handle the daily demand of 39 requests.