Why I became hyper focused on my Install Rate and what approaches I am taking to fix it.

As I have previously worked very closely with the introduction of players to games, and I try to be very user minded throughout my work, I thought I had taken good, quick steps to address the re-release of my game.

After pivoting my game back into a single player experience to be able to release more updates, I became increasingly engaged with working on my marketplace content.

This was a great experience as I started looking at my product quite differently after a friend helped to cross promote it, and get a better sense about where my time and efforts were better spent.

This is a blog post about Word Ways, my recent undertakings with marketing, and how it has affected my perspectives towards the product and how it's communicated.

1 - Your product isn’t necessarily the problem

Word Ways as an MVP (V1.0), had very few features, but great “rolling retention” metrics considering how little there was to the game. This however would turn out into an issue that Flurry became famous for as developers attempted to know more about specific day retention numbers.

V1.0 - This could mean so much, that ultimately, it means nothing.

Instead of focusing on iterating on my game, I chose to develop a backend infrastructure to support multiplayer and other social features (V1.2) that I deemed were important to create a successful product. This quickly became bigger than myself and I was unable to continue to support it.

More importantly, I never proved my product. So instead, now, I pivoted to a single player experience (V1.3).

V1.2 & V1.3 - First 50 seconds.

Removing every extra feature, and getting some help from a friend, I would go and receive a bunch of users via Chartboost from him, in order to understand how users were enjoying Word Ways.

This time around, I was able to get a clearer picture into the game. After focusing a lot more on just the first 90 seconds and core experience, players were using every feature, and more recently accessing extra game elements such as Powerups.

Bare bones. Session times are up. Still, not enough users to validate anything.

By finally starting to do several updates, and doing a couple rounds of User Acquisition, bigger and more interesting problems emerged.

2 - Marketplace Listing and Install Rates

Before my friend could cross-promote my game, I updated my Marketplace Listing, and generated new advertisement creatives for Chartboost in order to receive some users. Here are those results.

300,000 impressions later, I had a little less than 100 installs.

Although 5% could be seen as a passable CTR number considering many factors, such as the audiences and games being cross promoted, a 0.24% Install Rate is extremely low. This was ultimately reflected on Google Play as well, since they now added a User Acquisition tab where you can get insight into your App's Install performance, including organic play store and web traffic.

Overall performance from different channels.

If I were to cross-promote to other channels, or perhaps do a UA campaign on Facebook, I'd probably find better users and higher install rates, but that's not the point. I'd be simply circumventing the problem and exhausting good users. The effectiveness of the Marketplace Listing wouldn't change.

This became my single most important focus. Engaging users in the first 90 seconds was originally my top priority and so, I started prioritizing my Marketplace Listing like it was my user's first second in the game.

As StoreMaven and Split Metrics bring intelligence services to this key step of the app funnel, new competitors are coming in the market to help solve this problem. If you are bootstrapping like myself, Split Metrics offer a free demo to try out their platform, while new entrants such as app.TestNest.co seem to be offering very similar services, currently, for free.

By simulating different Marketplace Listings, you can drive your own users via FB campaigns and get a natural understanding of how your marketplace store is performing.

Before trying new services and expanding my load, I focused on getting things out the door.

3 - Keep users engaged. Thinking in Brain 1.

One of the most enjoyable and rewarding challenges from this experience has been to detach myself from development and approaching this new challenge. by better understanding these hurdles, and overcoming them, I'll bring a lot more of these learnings back into the game.

A friend of mine uses an expression that is quickly becoming part of my daily lexicon. "Is this Brain 1 or Brain 2?"

Adopted from the studies found in the book Thinking, Fast and Slow, Brain 1 is the processing system in our brain that is fast, responsive, emotional and subconscious. It reacts instantly until it needs to process additional, more complex information.

Brain 1 - User reads "Tap letters to create words" and naturally start solving the puzzle.

Brain 2 starts the moment the person needs to make effort, process information and try to make sense of something. Long sentences, multiple messages, too many points to focus on. Everything in your visuals will draw the attention of the user, make sure to support it with the right Call To Actions.

Brain 2 - LEFT: What am I supposed do in this new kind of game? RIGHT: Is that a Bonus? Why the question marks? No CTA? A lost opportunity to explain your product.

4 - Percentage increments or a new number?

During this process, I would go onto read and research a large number of ASO, SEO and web marketing blogs.

I read as much as I could about adjustments that I could to Icons, Screenshots, Keywords, your App's name, Call To Actions, Copy, Buttons, Ribbons, Colours, Using a phone, Using a hand, Putting people in your apps, fish, dogs, cats cause a 50% increase in conversion, grandmas, a 57% increase! Grandmas are the new cats!

One of the materials that stood out was from the StoreMaven blog, containing a lot of useful information with regards to distinct user behaviours while visiting Marketplace Listings.

Ultimately, the biggest question regarding my marketplace listing was: "How can I make more effective change? Am I optimizing from 1% to 1.1%, or can I perhaps make bigger gains?"

Strong Call to Actions had lowest Click Through Rate, but biggest effects on Install Rates. So did an Icon... Keeping it completely clean had the highest CTR, but led to no Installs. Promoting to that demographic, over 7 days.

This led me to think a lot more about what I could do to re-imagine and better communicate the game. How could I focus on moving the needle more significantly?

Everything pointed to one thing. Less is more.

If my focus is to just increase the effectiveness of my Marketplace Listing, the most "pure" test I could do, is to not influence the user to install with a Call To Action, or with the Logo. Just get them to click through to the store, and let the marketplace listing do the rest.

5 - Re-focusing and re-imagining the message.

It's clear that every part of your store listing is important and there are extensive blogs written about that. As I had recently updated the Icon, and I was being cross-promoted, I focused on my screenshots.

I spent some time sketching different screens and laying out what I wanted to communicate. I cleaned up and recreated the visuals for the core gameplay elements, and proceeded to create, simpler, focused game moments. By bringing it back to the most basic elements, I hope to more clearly communicate what the game is all about.

New Marketplace Screenshots 1 to 3

It's unfortunate to see such dismaying Install Rate numbers, but it's also extremely exciting to be working so close to a critical, visible, and hopefully, addressable problem. By really focusing on cleaning up my screenshots, I hope to be able to engage my users.

After delivering these updates and now focusing on some additional UA, I feel like I'm starting to look at the game with a new set of lenses, especially with regards to where I might be able to make meaningful change.

This has become a crux, and like any other challenge, I hope to be able to see some results.

6 - How might I analyze other products?

I became engaged in looking at other products to practice my analytical eye, and how to more effectively spot opportunities to improve the performance of each visual.

Although Boomlings is an older game and hasn't released updates in quite a while, (still, a much more successful game than mine), I'll use it as an example against Best Fiends as they are both matching puzzle games to look at small details that could lead to a little extra boost to their marketplace engagement and Install Rate.

Boomlings on the Left, Best Fiends on the Right. Where can we improve?

Best Fiends could have perhaps improved by adding more yellow tiles to the right of the action, giving users who are familiar with the genre a feeling of "I know how to score even more!". This would perhaps, be a very minor gain. But, by naturally high-lighting the matching mechanics, Best Fiends clearly communicates what the action in the game is about and makes it engaging.

Boomlings misses the opportunity to communicate how users interact with the game, do you tap two spaces to switch places? Drag to highlight? Move them around the screen? Or Tap a single tile that connects to 4 matching colours?

Highlight the action. A simple tweak to your graphics can help users instantly engage with your product.

Leveling up. Same concept, communicated quite differently.

Again, Boomlings makes some assumptions here about what users are going to understand from the screenshot. While their first screenshot was close to "Brain 1", this one is clearly "Brain 2". Various concepts are being communicated to the player at the same time: collect Gold, Experience, and 3 characters that are too subtle to notice, and with colour too similar to what's nearby.

To perhaps improve this, we should start by first removing the game board from this screen as it's even busier than before, and has little to do with the Call To Action. Second, we should use more screen real estate to feature and visualize these characters evolving and perhaps their new found strengths. In addition, perhaps use the same starburst background as before (BLUE) so these orange characters stand out.

It's hard to continue the exercise with Best Fiends, as they have done a great job. They actually show their main character, and stages 1, 2 & 4 of evolution, skipping the 3rd one to make the visual delta bigger, and are still able to save final evolution from being revealed.

If I had to attempt something, it would first be to switch the Call To Action positions. Placing "Level Up!" at the top where the character is fist pumping, and "Unlock Amazing Powers" by the most powerful stage, at the bottom. This way, the CTA would support the visual message that you are telling the user as they parse the screenshot, also because most people will likely read from top to bottom.

Another option would be to look to the visual language of the game which use black Silhouettes to communicate further evolutions. If this is part of the visual language in the game, and works in creating desire to evolve, and it's already employed throughout the rest of the game, why not take the next step and see if it works in the marketplace? Instead, showing Stages 1, 3, and tease of #5.

Best Fiends teases future stages of evolution with dark Silhouettes throughout the whole product. If an effective tactic, perhaps it could help the "Level Up" marketplace screenshot? Would it become Brain 2?

It's easy to understand why developers wouldn't want to make drastic changes to their marketplace content, especially once they hit a certain level of performance. It's possible that these variations were tried, and they didn't work.

7 - Final Take Aways

The best way to maximize your performance with creatives, will be to offer unique value propositions behind each message. Best Fiends does a great job at this, each screenshot helps to create a lot of awareness and engagement with each feature in the product.

Your App's Funnel has many places for you to tweak, however, your Marketplace Listing is high up there, incredibly accessible and quick to iterate on compared to user acquisition, development, or segmenting user to max LTV.

If you are presenting a lot of information to the user and haven't looked at your screenshots with a Brain 1 analytical eye, perhaps it's time to review where there might be some room for improvement.

This is the first moment that the user is interfacing with your product.

Make the most of it.

About the author

Roberto Alcantara has been wearing many hats for over 11 years in the AAA and mobile game industry. Always curious and always deeply involved, his time has been spent designing, developing and leading teams to get results in AAA and social mobile games. Join him at www.cegames.ca or follow him on Twitter -@iretrograde