You Can (and You Should) Test for Organic Installs: How to Use Paid Traffic as a Proxy
They’re almost mythical. Spoken about with an air of reverence and an undefinable, unboxable magic.
Organics are the holy grail of installs and their perceived mystery is part of their appeal. They’re the hard-to-quantify reward for a job well done. (In this case, ‘job’ could be referring to the product itself, the brand, or any and all marketing efforts.)
They’re supposed to just… happen.
The concept of organic traffic originally began to materialize when paid web-based advertising started, mostly in the form of initial paid search results on the very first search engines.
Necessity is the mother of all invention and in order to distinguish between traffic coming from a paid funnel and other traffic sources a new term was needed. ‘Organic’ became a catch-all term for traffic that didn’t come from a direct click on a paid ad. It is what it is not.
Because of this, marketers have viewed (and treated) organics as a completely separate entity as if there are two separate entities, a paid side of things and an organic side. But this is a self-fulfilling prophecy and not the most helpful one either.
While there are different methodologies to improve the performance of each segment, the distinction has become so ingrained in our minds that we forget that on the other side of both paid and organic traffic is still an actual human. Traffic, at its core, is just another way of categorizing humans and utilizing that generalization in order to optimize funnels.
Organic installs are defined by two distinct journeys.
The app stores divide organics into two: search installs, and browse /explore installs.
- Browse/ Explore – this includes any install that originated from a tap on either a top charts/category chart listing or a featuring listing (in the various placements they can appear, e.g. Today tab, Games tab, etc.)
- Search – this includes any install that originated from a tap on a search results page listing. This can be either:
- Branded Search – a search for your brand name or for your competitors’ brand name (that you rank for).
- Intent Search – a search for any keyword (that you rank for) that describes a function or a specific problem a user is looking to solve. Any keyword search that is looking at a type of an app without knowing exactly the brand they’re after such as “yoga”, “music”, “racing games” etc.
Marketing, remember, is all about communicating the right message to the right audience at the right time. And the key to cracking that message is to never forget there’s a person at the other end of that ‘audience’. Focus on the human and you’ll discover that we’re a fickle bunch. We’re moody and we’re impressionable in equal measure. This means we should focus more on identifying and targeting prospects based on their shared mindset.
As testing on the app store involves sending dedicated traffic to an app store replica sandbox environment through a paid UA campaign (often via Facebook, but through other channels as well), the question is simple: Can you use paid traffic as a proxy for organics?
The short answer is yes; the more interesting question, though, is how.
How can ‘mindset’ help test for organics?
Many believe that the medium from where users arrive at the store is the most important variable for predicting their behavior; however, after sampling hundreds of millions of installs from various sources, we saw something different. It’s the mindset of these users that plays a much more important role in determining their behavior.
Users with the same mindset coming into your app store page through an ad or through an organic search in the store will behave in a similar manner.
If someone is boredly scrolling through Facebook, just looking for something to catch their eye, they exhibit the same mindset of a browse user in the app stores. They’re there, bored, and looking for something of interest. They are both open to discovering new things in a similar way. Make sure you’re the thing they stop to discover.
Organic traffic sources and their characteristics
Browse traffic is, simply put, the broadest audience you’ll encounter. It consists of all App Store and Google Play active users who browse through the different tabs and sections of the stores (today tab, games tab, top charts, etc.).
Although browse traffic tends to have the lowest conversion rate (as it is the broadest audience) it can be a valuable source of installs. If you focus too much on the most targeted audiences, you risk leaving growth on the table. There are many high LTV users waiting in less-targeted, adjacent audiences ripe for taking.
Figuring out how browse traffic responds to the messaging and creatives that you’re putting in the store is key for maximizing growth and making sure you’re not alienating these users.
The chances that any stock-standard average browse user who stumbles upon your product page is actually interested in your specific app are low. They didn’t come into the store specifically to search for your product, they were just window shopping and landed up in front of your store and took a peak at your window.
Because of this random happenstance, the characteristics of this group of organics are very broad, encapsulating all users in a certain country that visit the app stores.
Let’s pretend you have a yoga workout and meditation app. You may already acquire users through paid media sources such as Facebook and would simply utilize the same targeting as your mobile app install campaigns in campaigns dedicated to testing purposes. Your users may also download the app through organic methods, in which case you can target the types of users who usually download your app through Facebook-specific targeting. You can target users by demographic, location, interests, competitor keywords, and dedicated lookalike (LAL) audiences.
Target browse users on Facebook by staying broad (like them), such as U.S. iOS traffic male and female 18-45 and optimize towards Reach as opposed to Conversions. This will result in sending browse-like traffic to your test as a sample, giving you valuable, actionable insights into how to improve conversion rates for this audience.
Search traffic is of much higher quality than browse traffic because their intent is much higher. They’re on a mission. They know what they want and to get that install you just need to provide it. Branded search users are looking for a specific brand, intent search users are looking for a specific solution.
Target branded search users on Facebook by creating branded-search-like traffic by targeting people who like your fan page/s. Or target a similar audience through LAL traffic or competitor targeting.
Target intent (and paid) search users on Facebook by using interest-targeting. This way you target users who have already expressed interest in your category or offering. By targeting an audience with an interest in yoga, the sample that will enter the test will closely resemble the organic search users who search for yoga in the app stores too.
Once you’ve identified your chosen target, it’s important to know what you want to hit them with. Make sure your tests are based on the most informative and useful hypotheses, relevant for those specific targets, based on competitive research, guided by best practices, and based on a strong understanding of your app and its most resonating messages/functionality. Your yoga app test should have clear hypotheses and designs that clearly demonstrate them. Which messaging resonates the most with your audience? Is demonstrating the overall benefit with a holistic lifestyle-centered approach the right one? Should you focus on UX and actual functionality? Should you spice up your icon? Is that the asset letting you down?
Remember, targeting traffic is only step one. Make sure the next steps in the testing process are just as thought through. Whichever hypothesis you choose make sure your test accurately tracks users’ behavior once they’re on your product page. This is where your most vital data lives and where you can derive the most benefit. After all, if you’re looking to connect with your users in the long run, it’s as vital to understand their behavior as it is to find a once-off solution for a creative update. The more you know, the more informed your actions will be going forward across the broader marketing funnel and in the post -install experience.
We’ve conducted hundreds of thousands of tests and based on these, one thing is glaringly onbvious: as emotional creatures, how we act depends more on our mood than on our interests. Understanding that mindset will enable you to improve your messaging and creatives to increase organic installs no matter where they’re coming from. After all, the potential for uplift exists everywhere.