Manzana Insurance
The Fruitvale branch is not faring well, they are behind competitor Golden Gate in new policies(375 to 326 in favor of Golden Gate), endorsements(300 to 206 in favor of Golden Gate), renewals (1,400 to 1,063 in favor of Golden Gate), turnaround time (2 days to 6 days in favor of Golden Gate), and renewal loss rate (15% to 47% in favor of Golden Gate) in the most recent quarter. The number of renewals, and endorsements at the Fruitvale branch are lower than they were for the same quarter the previous year, and the turnaround time, number of late renewals, and renewal loss rate have increased. The Fruitvale branch has performed especially poorly during the 1991.
The main cause of the problems is that the Fruitvale branch processes RERUNS last because they are seen as less profitable than new policies. When RERUNS aren’t provided by the due date, it displeases customers and causes them not to choose to renew the policy with Manzana. In a market where prices are essentially identical, customers value the service they receive and how fast their requests can be processed. Therefore, since the Fruitvale branch is falling behind in this regard, the renewal loss rate is at a record high. Additionally, if a renewed policy is late, the insurance agent is likely to suggest another agency to customers, thus adding to the low retention rate. To make matters worse, only 15% of RAPs become RUNs, so they aren’t acquiring a large number of new customers.
An increase in new policies written has been contributed to Fruitvale’s inability to process RERUNs on time. In the most recent quarter the Fruitvale branch wrote 326 new policies, which is up from 278 in the same quarter from the previous year. Since new policies are processed first, the increase in new policies written causes every other request to be pushed back. This has an especially large effect on RERUNs since it is the last request processed. The firm claims that they follow FIFO when processing requests, except for the branch orders the importance of the different type of requests. For example, new policies are processed first when they come in, even if there are several RERUNS already waiting to be processed because RUNs and RAPs take precedence over RERUNs. This means that the branch really does not follow the FIFO process at all. Using this flawed system, the Fruitvale branch has increased their backlog since 1989.
Another cause of their problems is that the distribution clerks do not receive RERUNs until the day before they were due. The idea behind this process is to use the most up to date information in the RERUNs. However, with the current FIFO policy the company has, if there is an influx of RUNs and RAPs the workers are unlikely to process the RERUNs the day they are generated, making them overdue.
Additionally, Fruitvale has lost many customers because agents were not recommending customers to Fruitvale. Agents often call insurers to ask for their TAT, the number of days between when a RUN is received and when the final policy is issued. Since Fruitvale had a slow turnaround time, 5 days as opposed to Golden Gate’s promise of only one day, many agents would not bring customers to this firm.
Manzana Insurance has two major problems with the way that they calculate turnaround time in exhibit 3.This flawed understanding of how to calculate turnaround time is negatively influencing certain process decisions. For our analysis of this exhibit, we are adapting Manzana’s FIFO rule to include a longest time rule to deal with the backlog. Our suggestions included in this exhibit are only for analyzing the issues with Manzana’s calculation of turnaround time. First and foremost, Manzana multiplies the 95th percentile of the standard completion time per request by the average number of requests each worker will do for each process (requests/worker). This is incorrect because each request will only have one worker on it. For example, in exhibit 3 in the case, the 4 distribution clerks have one RUN to process. The 95% SCT is 128.1 minutes. Manzana multiplies this by .25 to get a total time of 32 minutes. This assumes that all four clerks are working on it simultaneously, but in reality one clerk will be working on it for the standard time while the other clerks do a different request. Calculating the turnaround time using one worker per contract will increase the total times to complete each operating step, as the labor cannot be divided up amongst a single request. The method depicted in exhibit 3 also does not take into account that the clerks are all working at the same time. Using Manzana’s method of calculating time, the distribution clerks will be working for 248.7 minutes, while our method using the longest task rule will take 262.3 minutes. (Exhibit A)
The other major issue with the method that Manzana uses to calculate turnaround time is adding each individual operating step’s total throughput days together. This method of calculation ignores the fact that each operating step can be working at the same time. Also, each individual request can be passed to the next team as it is processed instead of waiting for every request that the team has to be finished. Each operating step has work to do as the week begins, and they will receive work as it is processed by the preceding team. Therefore there is a certain amount of overlap in the times that each team requires to process a request. This problem is most evident in the underwriting team because the underwriting team is the bottleneck of Manzana’s operating flow. Exhibits A and B, attached to this write-up, demonstrates this. The basic principle of these exhibits is that the distribution clerks will be able to finish the work on their desk before the underwriting team finishes the backlog originating on their desk. Therefore, as the underwriters finish that backlog, they can immediately start the work the distribution clerks finished without stopping. Therefore, the throughput days from the distribution clerks do not add to the total throughput days.
Although we used the 95% SCT in these calculations, as we were measuring turnaround time using the same method the company did, they would be able to better calculate the turnaround time if they used the mean SCT. The mean represents the average time a worker can complete these tasks in and using the 95% SCT is making estimates that are too high.
The first suggestion that we would make if we were Bill Peppen is to eliminate the FIFO method of processing requests that the company claims to use. In its place, we would institute a ranking system that prioritizes RERUNS. This system would schedule RERUNS out into the future, and release them to the DC’s with a 2 day lead time so that the different employees would be able to balance their schedules around the idea of knowing when they will have RERUNS to process. This system will help to get rid of the high variance in employee’s schedules that is currently causing both idle time and time with an overloaded work load. We believe that Manzana’s most important competitive priority is reliability and quick customer service. The case states, “As competition in the insurance business had increased, insurance rates and commission schedules had become nearly identical among competitors. As a result, agents had begun to steer their clients to particular companies on the basis of service.” Therefore, a company must be quick in its turnaround time to hold onto its customers and maintain market share. The biggest issue with the Fruitvale branch’s turnaround time was that the underwriters continuously pushed RERUNS to the bottom of the pile. As shown in exhibit 6, in two and a half years, there were no RAPs or RUNS late, but anywhere between 15 and 47% of RERUNS were processed late. Furthermore, RERUNS are a constant source of revenue throughout the years, whereas a new policy with a 50% chance of being dropped might only bring in revenue for 1 year. For these reasons, RERUNs should have the highest priority. In the new system, RUNS will be second priority behind RERUNS. Although the volume of RUNS is not nearly as high as RERUNS or RAPS, RUNS are the most important source of revenue per contract. However, on pure volume, RERUNS are still the highest grossing source of revenue. According to exhibit 5, RERUNS brought in three times as much revenue as RUNS. The third priority in the new system will be RAINs. RAINs are a more consistent source of revenue than RAPS, which are the fourth priority. Because only 15% of RAPs are converted into RUNS, Fruitvale can more efficiently use its time by prioritizing the other processes. According to Tom Jacob’s memo to John Lombard, the Fruitvale branch receives 17 RERUNS and 22 other requests(RUNS, RAPs, and RAINs combined). Using requests processed numbers from the first quarter of 1991 in exhibit 6, we were able to determine that 13% of other requests are RUNs, 68% are RAPs, and 19% are RAINs. This amounts to an average of 3 new RUNs per day, 15 RAPs and 4 RAINs. We used these numbers as the basis for our new projected schedule that is seen in exhibit C attached to this write-up. In exhibit C, we calculated our projected schedule by using the mean processing times to get an estimate with the new requests determined above. Based on a 7.5 hour work day, there are 450 minutes for each employee to work during the week. As stated in exhibit C, the distribution clerk with the highest utilization of his 450 minute work capacity is working 424 minutes during the day. The underwriting team with the highest utilization is working 372.3 minutes during the day. The rater with the highest utilization is working 355.9 minutes during the day, and the policy writer with the highest utilization is working 275.3 minutes during the day. All of these totals are possible work day capacities, which will allow for request that take longer than the mean processing time and allows Fruitvale to reduce its current backlog. The excess capacity allows for more effective scheduling of RERUNS because it provides the ability to move RERUNS around to prevent a backlog if there is a lot due at a certain date.