Mobile app optimization aims to drive downloads, conversions, engagement and retention of users in your mobile app. For many app developers (until recently - myself included) this term translates to one measurable result: downloads. We direct all our attention to increasing our app's exposure and optimizing acquisition channels to onboard and convert as many users as possible, hoping to that eventually the sheer number of downloads will translate to monetisation and a decent ROI.

You've got it backwards

When mobile app optimization is concerned, you should never start from the beginning. Start at the end, instead. In other words, look at your end goal and figure out why not enough users are reaching it. Downloads are usually not your goal. They're means to an end. Your actual goal is to retain the the users who've already downloaded your app. Retention is the end-game in mobile (and web, and any SaaS for that matter). Great apps retain users and as a result grow, monetize and perhaps even become profitable. If your product is not retaining enough users, no amount of downloads is going to matter, and pouring money into PPC campaigns or Search Ads is only going to drown your business faster.

Dropoffs

Let's take a step back. You're producing a mobile app and your goal is to retain users, i.e. to make them return to your app and find it trustworthy and valuable enough to make a purchase, buy a product or subscribe. Surely, in order to retain users, you must first onboard them. But onboarding isn't enough. In fact, the vast majority of users you onboard will not stick around. They will drop-off.

Pouring time and money into getting more downloads without understanding where and why users are dropping off is like pouring water into a bucket full of holes. You're failing slowly and painfully. The dropoffs are killing your conversions.

An average app offers plenty of drop-off points to users: they download but don't install, install but don't launch, launch but don't sign-up, sign-up but don't finish the tutorial... each step within your user-acquisition funnel offers a drop-off opportunity to the user.

The Funnel

The user-acquisition funnel involves a series of steps which end with a predefined goal, for instance making an in-app purchase. Your mobile app may have more than one funnel. Each funnel will include a few discrete steps ending with a goal. For most apps, there would be at least one funnel broadly resembling this one:

Discover -> Install -> Sign up -> Try -> Purchase

The reason this process is called "funnel" is because, much like an actual funnel from a kitchen or garage, it gets narrower along its length (or progression), allowing less volume to pass through it. The number of users at each step of the process naturally drops. These are the drop-off points. Understanding the drop-off points and fixing them is what "Mobile App Optimization" is really all about.

Metrics

Understanding your funnel and analysing dropoffs requires measurement. If you can't measure conversion, you can't improve it. The term Conversion refers to the event where the user progresses from one step to the next in your funnel. Each step of the funnel defines its own conversion. As a standalone term, "conversion rate" is therefore ambiguous and confusing if it does not also refer to the specific conversion step or metric measured.

Here's how we'd map each step in our generic funnel example to measurable conversion metrics:

Discover (impressions) -> Install (installs) -> Sign up (new users) -> Try (active users) -> Purchase (customers)

In each step, we'd measure the following aspects of the conversion:

• Nominal value - how many unique occurances were measured, e.g. number of impressions our app page received on the AppStore.
• Time - how long does it take users to convert? e.g. from being "active" to making a purchase.
• Rate - the nominal value devided by time, e.g. the daily install rate.
• Growth - the change of rate in time, e.g. sign-up growth rate.
• Cost - the cost of getting X users to convert, e.g. cost per install (CPI), cost per action (CPA) and customer acquisition cost (CAC).

Answers

At 100grams we conducted this optimization process for Psngr app over a period of one year. We began by defining the specific acquisition funnel and metrics and looking closely at our measurements. Sure enough, we identified more than a single "broken" step. In fact, our funnel was utterly broken!

Our next step was to delve into each broken step and understand why users were dropping off. The biggest dropoff point was by far the conversion from active to premium users. Only ±0.5% of all installs were upgrading to premium. This was a disaster. The average rate for normal business operation of a SaaS app would be 3-5% for this metric. Our product was broken. There was only one appropriate way to dig deeper and gain more insight into the product: ask users.

We went all-out collecting direct user feedback:

• Email - we emailed users who had been active and recently dropped off.
• Surveys - we tailored surveys around specific cohorts and asked poeple about their expectations and where the app failed to meet those expectations.
• A/B tests - we tested different premium price-points, varying trial periods and in-app purchase offers.
• Business model - we switched our business model from "Free trial" to Freemium.

Our app was definitely not meeting the expectations of our target audience. One key feature in the app was underperforming, while other highly-desirable features were missing altogether. Features which we deemed beneficial were actually counter-productive for many of our target users. Finally, our "trial" business model was not offering enough time for users to get used to the app and really trust it.

Growth

Embarking on this app optimization campaign saved our app, and our brand. It took almost a year to implement this campaign, and it took several major releases before we started seeing a significant impact on our conversion rates. The impact wasn't immediate. It took months, but gradually the number of new premium upgrades started to grow while the number of downloads remained unchanged. A year into this process we grew our monthly upgrades from about a dozen to over a hundred per month. We've acheived over 1000% growth!

Strategy

Obviously that kind of growth is hard to sustain. Our referece metric was extremely low, which made it easier to grow fast. Now that our reference metric is in the hundreds of premium upgrades per month, growth is much more elusive and difficult to achieve in triple or even double digits. Nonetheless, the apprach is similar: focus on those dropoffs.

What we've found out from implementing this process is that the most effective way to optimize your mobile app is by focusing on the product, not marketing. Spend time and money talking to your users and improving your app's feature offering, stability and reliablility. Invest in a solid infrastructure that can serve your users with little or no lag or hickups. Make sure that your core functionality - what your app promises to deliver - is superb. The rest will follow.