What research method is appropiate for this case?

Question

We have a mobile app, with 3 million of downloads and 1.5 million of active users. This is a young product with the main market overseas. I just arrived to the company and I realized that they haven't done any user research before at all and they are creating more features in order to keep people using the app, since they leave after downloading it. These features then are not following any strong insight.

I think it's advisable to know our users obviously, so everyone is in the same page about them and we can create more meaningful features in the future and improve what we have in the app right now, that's what I think at least.

I can only think of two; one-on-one interviews and online surveys. From my understanding, interviews offer a deep understanding on needs and behaviours, but seems they fit more when a product is on its discovery stage, not live for 2 years already. As for surveys, they show how the people is using the app today, who they are, what they want, but it doesn't provide the whys and deeper understanding.

What kind of research we should go with? I guess the answer will largely depends on budget, time and our goals, also considering the fact that I'm the only one who takes care about ux at the company. Still, I would like to know the options so I can offer the one appropiate for our case.

Answer 1

User research is quantitative/qualitative and attitudinal/behavioural . The decision of when to use which UX research method depends on your product stage as well as the factors listed by you. As you already have a significant amount of active users, my first question would be - What app analytics do you already have about the app usage? This falls in the product assessment part and is more quantitative in nature, what we call as Mobile app analytics.

Mobile app analytics usually have the following components:

• Behavior Reports - Information about app usage and the types of action a user takes.This could include screen views, interactions within a screen, and what crashes and exceptions they encounter.
• Screen Tracking
• Event Tracking
• Custom Metrics

For more details, you can take this free course by Google on Mobile App Analytics.

For your app, I would also recommend doing a social sentiment analysis. This can give you some insights into:

• What your users are saying right now
• What they like/don't like and what features they are looking forward to
• How your customer base is changing over time in response to your offerings

Along with the data from app analytics and social sentiment analysis, you can use the following User Research methods before and after launching new features. I am short-listing them from an article shared by NN/g :

Concept Testing (Before): A researcher shares an approximation of a product or service that captures the key essence (the value proposition) of a new concept/feature or product in order to determine if it meets the needs of the target audience; it can be done one-on-one or with larger numbers of participants, and either in person or online. Your mobile app analytics can help you with identifying the right set of users for concept testing.

Desirability Studies (Before): Participants are offered different visual-design alternatives and are expected to associate each alternative with a set of attributes selected from a closed list.

True Intent Studies and Customer Feedback (After): Can be covered in mobile app analytics.

Hope this will help you in conducting user research.

Answer 2

(Just adding to Adit's plan) You have a lot of users, and that's gold . If there's also a strong community behind the app or game you are working on, or you would want to build one, products like UserVoice (https://www.uservoice.com/) or the free tool http://doorbell.io will help you collect feedback and ideas. Basically, your features could be driven by user's opinion if you make it flow.

I agree that analytics + simple preference surveys is something inexpensive and useful, but you might need more evidence. Use your registered users for more studies, rather than personas. Your original personas are prone to change after you analyze your real user's behavior.