Figuring out what test automation’s to tackle first doesn’t have to be a herculean effort.

The assumptions here are:

  • You are getting started in test automation.
  • You are using sprints
  • You are currently testing manually.
  • You want to find out the biggest bang for the buck without wasting lots of cycles.
  • You have less than 5 testers (more and you should reduce the card ratios)
  • You know how much time you can dedicate to testing

Ingredients:

  • Testers
  • 10 blue index cards X number of testers
  • 3 Green index cards X number of testers
  • 3 Red index cards X number of testers
  • Sharpies

Process

  1. Ask the testers to write down their most important tests (test they are doing or wish they had time to do). One per index cards.
  2. Sort, count duplicates and post the list
  3. Give each tester 3 Green and 3 Red index cards. Using the list:
    1. On the red the test that covers the greatest risk to the application
    2. On the green the tests that gives them the most confidence the application is working
  4. Sort, count duplicates and post the list. Using the list:
    1. Estimate the cost to perform the manual test for the list
    2. Estimate the cost to automate the test on the list
  5. Make some decisions – but use the data. Using the list:
    1. High frequency, high risk, low automation cost, high manual cost
    2. Low frequency, high risk, low automation cost, high manual cost
    3. High frequency, low risk, low automation cost, high manual cost
    4. ….

Stop when you don’t have any more time available in the sprint.

Adjust the process to suit your team – this is just a guide and has worked well in the past.

Remember understanding what to automate is as important as the automation itself. That “what” starts with the vision for your testing team.