My team is working on improving performance for our web application.

We have a general idea that we want to:

  • Lower the number of "bad" API calls by focusing on driving down the number of requests that take longer than X seconds (focusing on particular endpoints).
  • Improve overall performance as measured by e.g. Apxex.

Unfortunately, it is difficult to estimate how long each performance optimization will take as we're not sure of what's causing some of the individual bottlenecks.

How can we take the information we have now and turn it into relaistic quarterly goals?

  • What are you going to use these goals for?
    – Erik
    Commented Jul 19, 2018 at 21:09
  • I think that's a good question. I think the general purpose is to make sure we're making progress on important objectives. Commented Jul 20, 2018 at 17:00
  • I don't have control over when we set the goals. Commented Jul 20, 2018 at 17:08
  • Never set a KPI unless you know what you're measuring and why. You want to identify a metric that improves some measurable quality in a clearly defined way.
    – Todd A. Jacobs
    Commented Jul 20, 2018 at 22:11

2 Answers 2


I recommend holding one or two rounds of diagnosis before finalizing the quarterly goals.

So basically you can dedicate two or four weeks to just diagnose the source of the performance issues you want to enhance then you may end up with a list of the things you need to do or the fact you need more research/troubleshooting to know more.

After you have a the reasons list, then it'll be more realistic to set quarterly goals. So the initial goals can be something like:

  1. Troubleshoot the source of the issues (time boxed)
  2. Define the solutions if possible for the cause of the problems or what needs more research
  3. Set the quarterly goals and estimate the effort/time needed to finish them
  • Accepting this because I feel like this is pretty reasonable. I had to come up with goals on a shorter timeline so I'll add a shorter answer to share what I did in case anyone else came runs into a similar situation. Commented Jul 20, 2018 at 17:07

Since I couldn't accurately forecast the amount of work involved in making the performance improvements I focused my goals primarily around the process of making performance improvements.

So, even if we can't control (or know) the amount of work needed to make improvements right now if we accomplish our goal we should be able to make these performance improvements more quickly due to having a better process.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.