Hi again! How are you? Ready to move forward? We have to talk about SLI now.
I know SLIs can be a bit confusing, so let me help you with a visual example. First, I want to repeat this: SLI is a metric of service success from the client's point of view. The definition of "success" can be different for each one of them - you will have to negotiate it in each and every case. We have yet to do this for the Taxi product...
For example, the classic SLI is error rate: when instead of something useful, the service responds with an error. This is very important to the client and fits the definition of SLI. To show the number of errors on the chart, you first need to select the time intervals on which we will count the errors: for example, 13 errors occurred in the segment from 22:05 to 22:10, and in the next 5 minutes we had 25 errors, and so on. An example is the second graph in the picture below (Http status codes). Green statuses are good, red and orange statuses are bad.
Such a graph can almost certainly be found on the dashboards of your tech team, but what is its disadvantage for a product manager?