You will alternative options for different types, going through all of your system objectives. Some of these alternatives may change the original assignments of responsibilities amongst the system components. For example, you might text out notifications rather than emailing them. Some of your alternatives will resolve different conflicts where those conflicts often arise because you're trying to balance all of your nonfunctional requirements like security and usability, and scalability. Stakeholders need to at least somewhat agree on the evaluation criteria for resolution of conflicts and risks. Much of this discussion, much of this negotiation, is going to be about the contribution to the nonfunctional requirements. You then compare options according to the criteria and pick out your best options using the qualitative or the quantitative reasoning techniques that we've discussed. And now we're going to go into a bit more detail. You've already read and heard about qualitative evaluation. You can rank things based on very positively, positively, negatively, very negatively and you're usually using these ratings to describe your nonfunctional requirements. As a reminder, those nonfunctional requirements are your ities, security, scalability, usability, etc. In the quantitative analysis, you're pretty much doing the same thing, but with numbers. In order to analyze risk, build a weighted matrix, and given your qualitative or your quantitative analysis, translate. Estimate the score of each option on each evaluation criterion and then weight it by some relative importance. Next, select the option with the highest overall score based on all of the criteria. Each criterion is assigned some kind of weight where that weight represents the significance relative to the others. One good way to do this is to put it on a 0 to 100 scale or a 0 to 1 scale. Everything should add up to 100, showing its importance. Then within your matrix, a matrix cell is associated with a balance of the option and the criteria. This gives an estimated score percentage of the option with respect to the criteria with respect to its overall significance. A score here, a score of x, means that the option satisfies the criterion in x percent of the cases. In the bottom of your weighted matrix, the last line gives the total scores of each option as a weighted summation of its scores, where that weighted summation is with respect to the various criteria. In the example that's shown right here, the e agenda is estimated to satisfy minimal inconvenience. Minimal inconvenience requirement is covered perfectly, but there are fairly poor estimates for getting a reliable response.