It is a lot of work to get started and to figure out what you should do and when. Here I would like to talk about the distinctions I make between triggers that tell you, right now you need qualitative research vs right now you need quantitative research.

Let me break it down:

• Quantitative research, quantitative metrics, numbers, analytics tell you what is going on with your product.
• Qualitative research, talking to people, observing their behavior tell you why those things are happening

So if you want to learn about your product and what is going on with it and how people are using and doing with it, you want to use quantitative research.

If you want to understand your user, their context, why they are doing the things they are doing, and why they are making their choices, you probably want to use qualitative research.

Both work together very nicely but unfortunately not everybody uses them together. Engineers tend to focus on quant while UX designers tend to focus on qual.

I think it is very important to have the qualitative research to generate good hypotheses. It encompasses all questions that we can’t answer quantitatively. And we need to have the qualitative data to understand, say, if the change I made to my product have the behavior change that I expected? Did it cause user behavior to change in the particular way I expected?

So if not enough people are engaging with my product. My hypothesis, which I got from observing people actually onboarding (qualitative research), says that if I improve my onboarding experience in this particular way, it will improve the engagement. Then I can make the changes, and do some quant like AB testing, analytics to know what happened and if it had the impact I expected. In the case it didn’t work, we need to ask why? And observe people use the new experience and try to understand what we got wrong.

To have remarkable results, we need to constantly combine the why and the what and keep that loop going.

But what happens when you don’t have enough customers to generate significant quant data? The great thing about having a small number of users, say 10 or 100, is that you can go and talk to every single one of them, and you can sit right next to them and watch what they are doing, which can give you answers to the what (quant).