At Fairing, we’re optimistic about AppLovin’s potential to drive incremental growth for the e-commerce brands we support. A few months in, we’ve seen over 40,000 orders & $5mm in revenue attributed to AppLovin as the point of brand discovery. There has been a 750% increase in “Game ad” attributed responses since September! The incrementality company, Haus, has started running incrementality tests that also look very promising.
As marketing dollars flood their platform, the question remains: will CPMs rise? Does the current performance have staying power? For now, it’s an exciting time in e-commerce as brands aggressively lean into this new platform for growth.
As more brands onboard their platform, we’ve developed a guide to maximize the value you’re getting from your AppLovin investment and Fairing.
Measuring AppLovin
We continue to see Fairing customers scaling their AppLovin investment, a strong signal of its potential – despite measurement challenges. AppLovin currently relies on a client-side pixel to measure conversions. Pixels often fail to capture conversions accurately, and without a robust e-commerce-focused identity graph like Meta’s, AppLovin is still in the early stages of addressing these measurement gaps. This is a good sign as their targeting functionality should only improve from here.
Currently, AppLovin doesn’t offer a server-side integration (CAPI) which could recover lost conversion data by bypassing client-side limitations, whether that be from pop-up blockers or other browser restrictions. This means their ability to reconcile cross-device and cross-platform attribution is more limited than other players in the space.
Despite these challenges, what we’re seeing from AppLovin in terms of scale is impressive. Their unique in-game ad surfaces coupled with well optimized machine learning for campaign optimization, is delivering performance levels we haven’t seen in quite some time for a new channel. It’s proof that even with iOS 14 limitations, there’s still enough client-side signal to optimize against – we’ll see if this holds long-term.
Optimizing your HDYHAU attribution question
We have a few recommendations for optimizing your attribution survey question setup for mobile games. For starters, if you’re skeptical about AppLovin, we actually recommend not adding it as a response option. This goes against traditional recommendations, but if consumers are actively writing in a response option, it removes any biases created by random submissions. This bias is negated via response option randomization, but again, if you’re skeptical, leaving it out is a valid option.
Another alternative is to add “Game ad” as a response option prior to launching AppLovin to understand your baseline – in our dataset these baselines are typically < 1% of total responses, but it’s still important to establish this and test.
How to work AppLovin into your Response Options
For this, we recommend adding “Game ad” as your response option. This recommendation is based on data we’re seeing in “Other” responses across roughly 5,000 responses in October.
Using a clustering algorithm, we grouped responses into buckets:
- "Direct Game Reference": ["Game", "Games", "Game app", "A game", "Online game", "Playing a game", "Playing games"],
- "Game Ad/Advertisement": ["Game ad", "Game ads", "Game advertisement", "Game advertising", "Game commercial"],
- "In-Game Ad Reference": ["Ad on game", "Ad on a game", "Ad in game", "Ad in a game", "Ad on a game", 'Ad for game', 'Add on game']
The most common response is referencing a “Game” – without mentioning the advertisement. Luckily, and contrary to other social channels, there is no concept of “organic” discovery on mobile games, so clarification questions are not needed to understand what’s paid versus not paid.
The ‘Uncategorized’ cluster predominately consists of consumers referencing specific games, i.e. Solitaire, Wordscape, etc.
Recategorizing Game Responses
If you’re using the open-ended approach, we recommend manually recategorizing those responses to “Game ad” to get a better understanding of the performance compared to other channels. You can do so by clicking “Other” in Analytics and recategorizing the response. For additional information on how to do this, visit our documentation.
New Versus Returning Customer Acquisition
As is our normal recommendation, target your HDYHAU question to new customers only and add another question “What led you to [brand name] today?” targeted at returning customers. You can use the same “Game ad” response option in both questions. This enables an easier extrapolation in Fairing and will also improve completion rates.
Analyzing AppLovin in Fairing
The first step is understanding your general performance across channels. We’re seeing a wide range of results – for some brands, “Game ad” has already surpassed Instagram as a discovery channel – for others, it’s much lower. We’re still looking to determine which brands perform best – whether AppLovin becomes vertical or AOV specific is unknown. To note, at Fairing, we don’t have access to spend data, which could also explain the performance delta between brands.
On the Fairing Analytics page, you can also view AOV and revenue data from AppLovin compared to other channels. We’re seeing AppLovin AOVs come in 3.56% higher than Meta – a great sign that this audience is at parity if not better than Meta, the channel where most Fairing customers are allocating the majority of their spend.
Insights
In Fairing, you can compare your response data with our Attribution Insights. This feature allows you to compare response options against last click UTM parameters/referral source and promo code usage.
In Fairing’s Last Click report, you can filter by the response “Game ad”. In the below example we can see that there is a high correlation between attribution response and last-click – a great sign and most likely an indicator that consumers are seeing an ad and immediately purchasing. This may be driven by the lower (below $100) AOV for this brand or great alignment with audience targeting.
Alternatively, you can analyze your AppLovin response via Last Click by typing “applovin” into the Source/Medium filter. Here we can see that AppLovin is driving a meaningful percentage of new consumers to this brand, while also capturing a meaningful amount of bottom of funnel conversions and consumers who are discovering this brand on Facebook and Podcast ads.
Who is my AppLovin customer?
To understand who these new consumers are, we recommend adding additional demographic questions to your Question Stream – age, gender, and gifting preferences being the most commonly used. With these live, you can then pivot your Game ad responses using the Fairing Comparison View to learn even more.
Below we display a breakdown of customers who responded with Game ad to a HDYHAU question with their response to "Who did you purchase this product for?"
This data is proprietary to your brand and should be leveraged to improve creative and your overall strategy. Does it align with other channels or is AppLovin opening up your brand to a net new audience?
Another useful comparison is gender. In the below example, we can see that AppLovin is driving a female customer roughly two times that of male.
As buyers return to purchase again, you can start using Fairing LTV Analytics to better understand the AppLovin customer lifetime value. Is the performance some brands are seeing driven by the holiday season or are these high quality long-term customers? Only time will tell.
Conclusion
As with any new advertising channel, building a robust measurement framework is critical to understanding both performance and customer behavior. AppLovin is showing incredible potential as an incremental driver of new customers, but its long-term success will depend on how brands measure their return on investment. It’s clear AppLovin is open to bringing in third-party measurement solutions to measure their own success, which is a great indicator in the confidence of their platform.
We’re optimistic about AppLovin’s staying power in e-commerce top-of-funnel discovery, and at Fairing, we’ll continue refining our product to help measure and unlock their full potential.
