Expanding Engagement

Hamilton is one of the most popular musicals in history, and Hamilton fans are some of the most ecstatic people you will meet.

The marketing team at Hamilton came to our company, Very Good Ventures, to create a way for fans to interact with the musical outside of the stage.

When I started working with the team, they had developed many features for the musical, including:

• A news section for all updates around the musical
• A lottery feature where users could enter for a chance to win free or discounted tickets in their city.
• A treasury section where users could purchase in app currency and purchase stickers, filters, and songs
• A trivia section that quizzed you on facts from the musical
• A karaoke section that allowed users to make videos singing the different songs of the musical.
• A camera with different filters based on Hamilton characters.

This app was Very Good Ventures’ first major success, and was the first to put Flutter on the map. We had over 7 million monthly active users and around 100 thousand daily active users. However, the engagement on the app was minimal and pertained mostly to a single feature.

This is the story of our efforts to improve engagement of the fans.

The Problem

About ninety percent of users on the app came to the Hamilton App for a single feature, and then left. That feature was the lottery. Hamilton tickets cost around 400 dollars, but many people with less of a budget wanted to see the musical. Users would get regular notifications when a new lottery opened, and they could enter for a chance to get tickets for the show.

Of the seven million monthly active users, there were a maximum of around one hundred fifty thousand users on any other part of the app, or about two percent. My role was to investigate the cause of this, and create some solutions to bring people to other parts of the app.

My Role

My job as Senior Technical Project Manager was to:

• Discuss the goals of the app with the Hamiton marketing team
• Research and understand the analytics pertaining to those goals
• Produce designs and specs for the team
• Oversee and manage the delivery for each feature

Hypothesis

It made sense that the lottery was the most popular feature, as the amount of people that love a movie enough to play video games and post on socials about the movie will always be significantly less than the amount of people who simply want to see the movie. 

One possible factor may have been that not enough users had prompts to interact with the features other than the lottery. My theory was that the hamburger menu of the app made users less likely to explore other features of the app after entering the lottery. While hamburger menus have proven to be incredibly helpful tools for containing large amounts of menu items, many designers argue that they sacrifice discoverability.

Of course there was also the possibility that 98% of users simply weren’t interested in the other features of the app, but this was an opportunity to find out.

Design Plan

The big picture for improving engagement would be to have UI after the user signals finishes using each feature that would link them other features of the app.

Flutter provides developers with a built in way of creating intrinsically reusable widgets, so it wouldn’t be too difficult to create a widget to preview each feature of the app. These widgets could then be placed sequentially in an “explore” section under the completed lottery screen and any other feature we would like.

Below is a wireframe of what I envisioned for the section.

With the trivia feature, we could take a single question, and make a widget to preview it.

For the karaoke feature, we could simply take two of the song listings, and make them a widget.

The full screen would look something like this...

Implementation

To find a simple way to test our hypothesis on a small scale before making the entire system, we started by building a widget that could be used one time within only one feature. We started by making a related articles feature within the news section.

Users could also swipe left and right on the articles to view the different articles chronologically.

Additionally, we added a simple button below the lottery screen that took users to the trivia screen, just to see if we got activity from the lottery.

We logged analytics events for swipes, related article taps, and returns back to the menu, to see if users actually engaged with the new features.

The Result

Unfortunately, as soon as the features were released, we saw a sharp decline in usage across every feature of the app… almost 100%. Our conjecture was that this decline was due to the date in history: March, 2020… when you might remember what happened.

Poor Broadway!

Reflections

With the given situation, it was not the easiest task to effectively evaluate the improvements brought about by the related articles feature; however, they did get some activity. This provided some sway toward the discoverability argument. 

That being said, this only illustrated an appetite for more content within a single feature. We would have had to provide links to different features of the app, to test if cross-feature promotion would actually be effective. This was no longer possible due to budget constraints.

I was quite disappointed that we never got the chance to see the project to completion, but I was happy that we made some progress and minimally validated our hypothesis. Progress over perfection!