Risk Assessment Examples

Test a software application for production readiness or put into production?

Your company has determined that it is time to roll out a custom software application that you and your team have been developing for over a year. This application has enterprise-wide implications in that if it works well the entire organization will run more efficiently but if there are problems those problems can interfere with a large part of daily operation at varying possible levels of disruption. Your management wants to roll out the application with minimal testing because they are feeling pressure from the executive committee to show results for the ongoing development investment.

Members of your team feel that more testing is needed before the application is ready for production. Your teams analysts have developed some estimates of how much a disruption could cost the company in profits, the likelihood of it happening and the amount of time it would take to correct the error.

They have prepared these estimates and they are included in the table below. Comprehensive testing of the application is estimated to require 60 days and cost around $125,000.

1. How would you evaluate the situation?

2. How would you evaluate the financial aspects of this decision? Show your calculations in determining the financial value of both testing and not testing approaches.

3. What are the direct, tangible factors that influence your decision?

4. What are the intangible ones?

5. How would you determine the uncertainties involved?

6. Would you do the testing or release it to production as is? What would you do?

Use the following information as part of your analysis.

Without Testing	Cost to Comp/Hr	Likelihood	Hrs to Fix	Frequency
Minor	$10,000	80%	3 to 7	>10
Major	$30,000	35%	6 to 15	4 or less
With Testing	Cost to Comp/Hr	Likelihood	Hrs to Fix
Minor	$10,000	30%	3 to 7	5 or less
Major	$40,000	10%	6 to 15	1 or less

Note: There are ranges of possibilities presented in the table. For example, the "Without Testing Minor" incident is estimated to cost $10,000 per hour to fix, with the number of hours required to fix it ranging from between 3 and 7 hours. It is estimated that this could happen more than 10 times and with an 80% likelihood of these estimates being correct. From this information you can calculate a low to high estimate of each scenario. Hint: There is a large difference between the two estimates.