Probably one of the most popular terms in mobile business is MVP. Numerous evangelists, advocates, consultants and mentors work with startups and help them apply the Lean Startup ethos. They push them towards building an MVP and going further with incremental updates. The main goal of this activity is to find a product market match and understand financial characteristics. To put things simply, these are preliminary steps to build a real business plan and scale the business. If you scale an unprofitable business model, then you scale losses. There are a lot of materials on this topic, but there is not so much information on how to actually do that, or how to get actionable data. This is not a problem, though. I’ll show you how to make it using Google Analytics, which is both simple and free. To make it more fun and more interesting I’ll use PhotoSuerte app as a case study.
There are many areas of mobile business which require optimizations:
- Landing page optimizations – to convert better (get more downloads)
- Web site SEO optimizations – to get more quality traffic to your landing page
- App Store optimizations (ASO) – to get more downloads from app page visitors
- Mobile App optimizations – to offer the best user experience (and get more referrals, word of mouth, in app purchases, etc.) – this is our focus in this article
I suggest everyone optimize their mobile app User Experience based on use cases instead of simply optimizing overall UX or concrete screens. At the end of the day, you need to make the whole service work well rather than a single screen or one element. You should have at least one scenario for user behavior in your app when you develop it. Odds are, the actual behavior will differ a lot and your task is to help users to get into the scenario or adapt the product up to the new behavior (you may treat that as a pivot).
Start by listing all the major usage scenarios which you think users should accomplish to make the product work for them and to get value. Think big. In most cases there are some common processes like on boarding, purchase, etc. Here is the list of usage scenarios for PhotoSuerte:
- Onboarding – getting a newbie into ready to start user
- Main Scenario – Chatting – the continuous process of getting photos, replying and waiting for the replies
- Interruption – push notification – when a user gets reminded about a photo received
- Purchase – do we sell well enough?
PhotoSuerte is the app which could remind you of the Rando app in some way. You send a photo and it is distributed to another user of the service at random. The recipient could reply to you and you’ll have a chat. You don’t know anything about your peer except their location. So, you make friends all over the world and have fun.
Onboarding process optimization
Onboarding is the process of getting a new user onto the service. You need to get them aware of the product, its features and most common techniques. Also, it is very important to get the user ready for using the app, e.g. informing them whether something should be configured beforehand. This is usually done during this onboarding process. For example, you might need the user to sign up, or to import some content.
The onboarding experience is very important. It helps you to decrease bounce rate. Early versions of PhotoSuerte were used only by less than 10% of users, others were bouncing being unable to figure out the value. We spent several months getting the onboarding process right. Now almost 80% of newcomers use the service. Here is the list of the most important lessons learned. The list is made in the form of mistakes we made and explored afterwards.
1) We don’t need a first start tutorial. Our app interface is intuitive. Wrong! It is intuitive for you after you spend weeks and months working on it. To prove it, just check new unaware users’ feedback, check what new users do in your app. We didn’t have a starting tutorial in the PhotoSuerte app, our UX was terrible, many users didn’t understand what should they do.
2) Nice icons are intuitive. Well, in many cases, users are afraid of unknown icons and touch them only if they don’t have another choice. Named buttons usually work better. There is a limited set of well known icons.
3) We use just well known gestures. There are no well known gestures! If you have gestures you should think how to explain and show those gestures to your users.
4) Ask for users permission only when it is really needed. This is a controversial one. If your app needs a set of proper permissions to work then you need to either request them beforehand or right when they’re needed. The choice doesn’t matter when there is only one permission to request, but if you have three of them then it could be a sequence of bumpers on the user’s road. Consider this: you allow your user to start using your app right after the download. For some actions, he needs to be signed in (for example). He makes a move and a sign in popup appears. Signed in. Next he makes a new photo. But he needs to allow camera access first. The popup appears. Next. The photo is going to be saved into the phone media library. But he needs to allow access to his camera roll. Another popup appears. Most likely a user would like to get a push notification when he gets a reply. Another popup. What a mess?! Finally, we decided to have a tutorial with several steps explaining what the app does and permissions buttons in place for the next button. Works like a charm.
Here are some techniques for you to analyze during the onboarding process in Google Analytics. I like using Behavior > Behavior Flow report. The key here is to choose only the «New Users» segment in order to analyze first use data only. Also, you should analyze only one app version at a time in order to analyze first time use after particular improvements. Delete the app from your phone, install it again. Start the app. Go through the onboarding process via all major paths which you see in the Google Analytics report. Ask yourself this set of questions on every screen:
- Is this persuasive? Am I eager to know what’s next?
- Is it intuitive – what should I do next?
- How could I decrease a drop-off rate?
- Are the most important messages standing out over the noise (graphics)?
- Are you excited after getting through the tutorial?
- Do you feel that you life will be at least a little bit better now with this new cool app?
Here is the summary for the Google Analytics segment I use to analyze the onboarding process in the Together app:
- Technology: Operating System = iOS (we have Android as well)
- Conditions: App Version = 2.0 (latest)
- Sequences: Active User Type = New Visitor
This segment provides data only for the latest version newcomers. Hence in Behavior Flow we will see their step by step walk through the onboarding process.
At the end of onboarding process, a user should be engaged enough and start using the service, i.e. start a couple of new chats, make photos. The percentage of engaged users after the onboarding process is the good measurable goal for your work on the UX improvement. You may configure the goal in Google Analytics and analyze improvements. I.e. check how many users complete the goal during the onboarding process (conversion rate).
You could also configure the «churned users» segment in Google Analytics. Those will be the users who launched your app only once and were not persuaded to use the app, i.e. churned.
Here are some resources to help you with onboarding optimization. Look at how others do it. This is very useful for your own inspiration.
Afterthoughts on onboarding
I think you just need to make your onboarding scenario really smooth, intuitive, engaging. This will make a great basis for the rest of mobile app experience. Treat onboarding as small talk in a verbal conversation.
Main Scenario Optimization
This is the process when a user gets the main value from your product. The more effective this process, the more likely that the user will recommend your app to their friends, or buy something in your app. There could be a good analogy. Treat your app like a game with indefinite number of levels. After that, you need to analyze the users’ gaming performance.
The most common approach is to have goals defined. The goals represent different tasks you design for users to complete. In PhotoSuerte the main scenario is to start new chats and chat with existing interlocutors.
- How much time does the user spend in the app?
- Is the process intuitive?
- Are you hanging onto the user?
- What’s the Goal Completion rate? How could you improve it?
You may find the behavior flow report in segment «Sessions with transactions» very useful for the optimization. It helps you to analyze what behavior leads to conversion, and what you should catalyze.
The Behavior Flow report on Google Analytics will help you to figure out how your users perform. It can also be very useful to have your own analytics to understand the users’ behavior better. Custom built CMS reports can be a great option. A main scenario could be very specific and not so easily tracked with Google Analytics. In that case, you could build some custom database queries to analyze users performance.
Interruption Optimization
The reason to optimize the process of user interruption is to get users as smoothly as possible into your app. Users aren’t likely to spend their whole lives in your app, nor even a whole day. So, you need to get them back and hook them there. From a technical point of view, you need to have a segment of returning users starting on the screen where you land users after a push notification. Check behavior flow for this segment and analyze goals completion.
Here is the list of questions to keep asking yourself:
- How quickly does the app load?
- How quickly could the user interact with the app?
- Is the push message clear and engaging?
In our case, we realized that it wasn’t convenient when users were getting a new photo and waited too long for it to be loaded (with all other updates). We spent some time optimizing the experience, and special back end API was added.
Purchase Optimization
Probably the obvious process where you need to understand what is the best way to lay out the values for a user and when the best time is to suggest a purchase (if you have freemium model). Use Goals and eCommerce sections of Google Analytics reports. If you have several ways to a purchase then you could figure out which one works better.
So, here are two questions to keep asking:
- Are users appealed to purchase something at the right time?
- Is the message good enough?
Also, the In App Purchase screen in your app could be the best pretender for A/B testing and a Google Tag manager feature.
That’s all for today. You are welcome to offer your own recipes for using Google Analytics in mobile apps optimization. If you find the article useful, please share it.