How to Grow Your App Business with Continuous Innovation and Data Science

This is the third installment of our blog series of Data Science and Artificial Intelligence posts by Dr. Colin O’Callaghan, a Data Scientist at 3Advance. If you missed them, check out part one and part two.

“A startup is a human institution designed to create a new product or service under conditions of extreme uncertainty” according to Eric Reese in his book The Lean Startup. Here, Reese hits on a key issue with growing a startup, the uncertainty. How do you grow an app? How do you build an app that users love and spreads virally around the world? It is easy to look at successful startups and think that their product was perfect right from the start. They launched a product and it immediately took off. In other words, they got lucky. 

In reality, most successful startups deal with uncertainty by continuously refining their products and growth strategies with an innovation cycle. An innovation cycle refers to the process of making informed changes to your app, accurately measuring the effect of the changes using data science, and choosing to fully implement the change or not. In this blog we will describe what data you need to store and how to use this data in an innovation cycle to increase your app’s users. 

App Intelligence

First things first, growing an app relies on quality data. “What gets measured gets managed,” as the saying goes. You need data to measure the impact of changes to the app and to identify opportunities. If you have an app out in the wild (or about to be released) then you need to start gathering the data streams listed below. 

Event Tracking

Record when users take certain actions in the app. For example, when a user logs in, makes a purchase, or simply clicks on the menu icon. 

Screen Flows

These are related to event tracking. They are the paths that different user segments take in a session. They can show how much time users are spending (or not spending!) in various parts of the app. They can also show surprising user behavior. For example, you may have some impressive functionality in your app, only to find that users rarely use it according to the screen flows! Knowing how users interact with your app helps you identify a priority list for any future developments.

Conversion Funnel 

This follows from events and screen flows. There may be an action in your app that counts as a conversion for you. This may be to make a purchase or share some content from your app on social media. Whatever action you want your users to do, a conversion funnel outlines each step a user needs to take to get there. Similar to screen flows, this highlights areas where users drop out on the way to completing a conversion.

Deeplinks

These are external links to your app that, when clicked on, open your app in a specific section — for example, the “share via email” option that opens an email app with a prefilled link. Recording these can show you the most effective channels that bring users to your app, helping you increase app users in the future.

Acquisition Channels

What leads users to download your app? Was it through a referral from a friend via WhatsApp, an ad on Twitter, etc.? By recording this data, you can identify your most effective user acquisition channels, letting you allocate resources effectively.

App Store Analytics and Intelligence

Your app store page should not be an afterthought. It is the last page a user sees just before downloading your app. There is data available here, such as how many page views and downloads your app has. You want to maximize the ratio of downloads to page views in order to optimize your app store search rankings.

User Segmentation

This involves the identification of user traits, particularly those who use the app most frequently and those most likely to spend money. This allows for more refined advertising with new user acquisition.

Cohort Retention

This tracks user retention over time. Typically users who installed an app on the same month/week/day are pooled together and whether they revisit the app in following periods is recorded. 

User Growth

This includes new, retained, and churned users. It ties in closely with several other data streams listed here, such as app store analytics and intelligence, acquisition channels, deeplinks, and cohort retention.

Sentiment Tracking

This is when you ask users for their feedback on your app. Those that score the app highly can be retargeted to ask for app store reviews and to invite other users. Very happy users can even help test new features.

There are many services that allow for the real-time capture of this data. Business intelligence (BI) software can load data from several different services into one dashboard, making it easy to check on your analytics regularly. Check out our recent blog on several BI software options on the market. 

Continuous Innovation

When you have a system in place for tracking all of your data, you can start the innovation cycle that will grow your app:

This cycle lets you test all the ideas you have about your app, from marketing to monetization. Design and build new changes to your app, and measure the impact these changes had on key metrics you are tracking with data science. Learn from the analysis of your new data and decide on your next move. Let’s look more closely at the design and measure steps.

When it comes to designing changes to your app, look to your team and your users for ideas. According to Rahul Vohra, the CEO of Superhuman, users who are most happy with your app will give you the best ideas to test. I know, this seems surprising. Your star users already love your app! But they also know what changes would make it that much better. 

The next step is to compile all of the ideas into a list. Then estimate the difficulty of building them and their potential impact. Use the data from the various channels that are outlined above for some guidance on the potential impact. For example, there is little point testing a big change on a screen that is hardly ever visited in screen flows! 

When it comes to measuring the effectiveness of your changes, a great technique is to run A/B tests. Say you want to change the sign-up page to a new layout that you think is more user friendly. How do you know that users will find it more intuitive? You could make the change and keep an eye on sign-ups over the next few days. The problem with this approach is that you haven’t compared like with like. There could be more or fewer new users on your app over that period, just by chance. The correct way to approach the measurement is to split new visitors into two groups at random. The members of one group are shown the new sign-up page, and the members of the other group are shown the old page. This way, you are controlling for random external random factors. You can have confidence in your decisions knowing that they are backed by data.

The most innovative companies experiment quickly and frequently. It’s all about discovering your product/market fit, fine-tuning interactions with your customers, and assessing the assumptions you made when you launched your app. Experiments can and should be performed on every aspect of your app and business processes. If you’d like to learn more about app growth, check out this comprehensive blog post by Andy Carvell.

Your Growth Team

Here at 3Advance, we specialize in designing and building apps with our clients. If you want to take your existing app to the explosive growth stage, get in touch. If you are looking for app developers who can build your MVP and work with you during future innovation cycles, look no further.  

3Advance is an app development company in Washington DC that creates beautiful, simple mobile apps. Get in touch with our team of data scientists, designers, and developers. We’ll work with you to optimize your apps and ultimately help you succeed.