You might have a problem and a desire to solve that problem but where do you go next. Imagine being in a situation where your museum app is opened regularly but then no other features are accessed, as assessed through analytics. You know that you need to figure out why this is happening. What is your first step?
User testing, such as task analysis, can help you understand where challenges are going. To use your money wisely, you should test with the demographics that mimic those who are already using your app. Right now, you are hoping to figure out why the people who are using your app are having problems. Of course, the challenges with the app might also be turning off those who are not even logging in. But, leave that challenge aside right now.
So, start with the types of people who are using your app. Think of the ways that you can categorize them. What age are they? What gender? Education level? Salary level? Are they familiar with technology? Are they museum visitors? Members? After making this snapshot of users, then you will need to create a screener that helps you creating a testing sample that mimics your audience. You might even create a faceted matrix to help you get the right mix of participants.
After that, then you begin thinking about the scenario and tasks, you would want to assess during task analysis. You will need to try to think of something that is not so prescriptive as to miss challenges and not something that is so broad as hide trends. Try to think of actual scenarios in your institution. Once you have created your scenario, say, you are a new visitor to the museum looking for ivory sculptures and you have downloaded the app onto your phone, then you need to create a list of tasks. You want to develop tasks based on items that you have already seen. Your tasks should help you explore the ways that users employ all facets that you are exploring.
Finally, you will want to make sure to use this task analysis exactly the same with each participant. In the end, hopefully, you will be able to see trends between each of the users problems. You might find that everyone is having trouble with the login screen. Or you might find people in a particular demographic have a hard time seeing the exit buttons.
In the end, task analysis is quite useful, because you are creating a systematized way of observing how a number of people use the same digital product. It allows you to see where there are challenges in order to make improvements.