Great difference between what people say they do and what they actually do
Observation is one of qualitative user research techniques. By observing people using your app you might get very important insight, that couldn't be obtained otherwise.
Observation could be split into two kinds:
- Natural conditions aka "real life"
- Laboratory conditions aka "lab based"
How would you do observation?
First what needs to be said - you need to inform a person about the research. Spying on people is creepy and unethical. The only exception might be observing people in public places, such as streets, coffee shops, cinemas, supermarkets and so on. But even in those conditions you better inform people.Let's imagine the situation: you're working on corporate software and currently struggle to define your next major release. So you decided to observe how some of your current customers actually use your product.
You call a couple of customers and ask if you could visit them and spend a day observing their employees working with your tool. Of course, you'd want to explain the purpose of the research - "it's all for your benefit dear customer". Hopefully you've got few "yes".
One day you arrive in your client office, you're being introduced to an employee who you'll be observing. Most likely that it would be a different person to the one you've talked over the phone with (buying vs user personas). Make sure you explain to the user why you here, what exactly you'll be doing and how long it would take. You'll need to do your best in empathy to setup friendly, informal atmosphere. The user will need to clearly understand that it's not him or her that being judged - it's software. When you have your user onboard you may start the research.
There are a couple of ways how observation could be executed. In the most passive option: you could just sit and watch user working, you could make some notes and ask few occasional questions.
Alternatively, you could take more active approach: asking the user to explain why he or she does things in a certain way. It might uncover some reasoning that is beyond your product.
Another way to perform observation research is to role-play a bit: imagine that you are a fresh employee in your client company and that more senior employee (your user) needs to educate you on your job. Not only on the tool itself but on everything that comes with it. Entire process. This might reveal unknown user problems.
How do you sum up observation?
The best way is to record entire observation on camera. This will enable you to analyze it again and maybe spot details that you missed. If the video is not an option - do some pictures and a lot of notes. If you can - bring another person with you, who can take notes and maybe notice what you haven't noticed.
When writing the report from observation you will need to summarize findings and prioritize them. Remember to mention how the knowledge you obtained related to your initial goal. In the example above, did observation you've performed helped you to define next update for your corporate software?
Learn more about observation
Read more about Observation,