Why do mobile gaming companies use fake ads? One more point
Breaking down the fake ads strategy
A few days ago, Mobile Dev Memo’s Eric Seufert published an article titled “Why do mobile gaming companies use fake ads”. You can find it here. The TLDR version of the article is:
- Mobile game companies have started using ads that falsely portray the gameplay of their app, sometimes in a significant way.
- As Facebook and Google act more like ‘black boxes’ of optimization, UA teams running ads can experiment with a wide variety of creatives and let the platform optimize those that seem to work. This also means that these UA teams can’t really know why these ads are successful, but just that they do, in fact, work.
- Fake ads are a type of false advertisement that carries significant risks such as being banned from running these ads on the platforms (FB + Google) and even potentially getting fined by the FTC for false advertisement.
Whatever the cons, these ads seem to work since they’re still running – the companies behind them keep on paying for them, which means they must generate some sort of positive return on ad spend (ROAS) according to their measurement.
The one key area that the article didn’t cover is why these ads are working, a question that’s definitely worth some deep thinking and analysis. So let’s break down the problem.
Why would a company run fake ads? The ad must work but what does it mean?
In mobile user acquisition, an ad has two ways to add value in terms of installs and revenues (both in-app purchases and ad-driven revenues):
1) Direct contribution – the ad would directly generate app installs. Some of these installs will convert to revenue-generating users (whether through in-app purchases or ads), which is mainly measured on attribution platforms that map ads to the installs they’d generated.
2) Indirect contribution – the ad would generate an indirect effect that will increase the app’s visibility — both for branded search traffic who then go looking for the brand they were exposed to in the ad, and as increased visibility through non-branded keyword rankings and top/category chart ranking (as first time install volume grow and push the app higher in ranking).
Let’s break down the different effects in which the ad could generate value.
Possible audience expansion
Hypothesis 1: a company that engages in fake ad strategy is actually significantly expanding its potential audience.
Facebook, Google, and the networks optimize to serve ads to the people who will be the most likely to install them. When the ad creatives clearly portray a certain type of game it naturally reaches a certain audience. This is something successful companies try to do with or without fake ads.
Let’s take Playrix’s Homescapes example:
An Ad Creative that ran for Homescape on Facebook on April 2nd, 2019
It’s reasonable to assume Facebook will optimize this ad to be served to a match-3 audience. That audience is inherently different than an audience that’ll look to play the game portrayed in the ad they ran later that year:
An Ad Creative that ran for Homescape on Facebook on October 18rd, 2019
After saturating its targeted match-3 audience, it’s possible that Homescapes managed to tap into a much bigger growth opportunity by appealing to a different audience, interested in a more story-based genre. If Facebook/Google algorithms worked to optimize for installs and not downstream in-app events, they started to serve this ad to that audience as they understood it yields strong install rates.
Let’s now analyze possible effects of such an ad, and what gets it to work for Homescape.
Effect A: They drive users that actually convert to players and paying users
You might (reasonably) assume that users who are likely to play a game that looks like the story-telling ad above aren’t going to want to play a more ‘vanilla’ match-3 game.
That said, Ishai Smadja, one of the industry’s top player researchers has published a fascinating analysis last year into game audience-affinity. Smadja basically mapped which other games players that play one specific game would also play by analyzing massive game audience-affinity data from App Annie.
Playrix has a similar game to Homescapes called Gardenscapes. Let’s look at Gardenscapes audience affinity:
In the words of Smadja:
“…Gardenscapes would fit into the Switcher genre However, its players originate mostly from audiences of genres who incorporate a light narrative and customization, and it is not as connected to the vanilla Switchers’ audience; further reinforcing the importance of the meta layer.”
So we see that Gardenscape and Homescape, both story-driven match-3, or “Switcher” games manage to attract Audiences that are farther from the basic vanilla Switcher audience (like Candy Crush).
But — if Homescapes UA team kept letting Facebook run ads with vanilla Match-3 creatives that accurately portray the main gameplay, they would only keep getting that same vanilla Switcher audience. To break out and signal to Facebook to serve its ad to a different audience (Resource Management, Hidden Object, Story Adventure) who would also enjoy the game but need to discover it, they needed to somehow tell the optimization algorithms to focus on that audience. They did so by creating creatives for such types of games.
If you focus on what these fake-ad creatives are about, you’ll see that they’re a combo of all three of these genres. There’s a bit of Resource Management, finding objects and using them to interact with other objects (that could relate to Hidden Object) and what seems like a very rich Story Adventure game.
All of this means that there’s a (not insignficant) probability, that once a player that responded positively to this seemingly misleading ad installs, opens, and starts playing the game, they’ll enjoy it and continue.
This means the ad itself could very well be generating direct value through high LTV users. But that comes at the cost of annoying the Vanila Switcher audience. We’ll get back to that later on.
Effect B: Increasing reach into new audiences increases the number of new installs, thereby increasing rankings
The fact that this ad is attracting a lot of new installs from these different and larger audiences tells both the App Store and Google Play algorithms the game is relevant.
They then start to reward it with higher rankings on both the Top and Category Charts, as well as for keyword rankings.
Top Chart and Category Ranking for iOS Homescapes in the US
Top Chart and Category Ranking for Google Play Homescapes in the US
Credit: Mobile Action
It’s clear to see that when this ad strategy began (around the beginning of the year), rankings have been going up.
That said, we can see that the effect has been getting weaker in the past 2-3 months. This could very well be the work of Effect C (below).
These higher rankings lead to significantly more organic visibility and to even more installs that didn’t go through the journey of viewing the “fake” ad. This organic growth alone could theoretically be what makes the ad work.
The ranking growth on Google Play also gives us a great hint that the fake ads aren’t actually driving poor downstream user behavior. As opposed to the App Store, the Google Play Store has publicly declared that they take into account retention, churn and other in-app metrics when they calculate rankings. If these ads would have increased these negative metrics, there is a low probability we would see rankings continuing to increase and maintaining their position.
Effect C: Creating dissonance isn’t great for app ratings and reviews, and that damages conversion rates
The “ugly” side of this strategy relates to the fact that there is definitely a group of users that get annoyed from the fact they expected one game then installed a game that looks and feels different.
This can be seen by explicit backlash on Homescapes app store page reviews:
Credit: Mobile Action
You can see how the negative reviews significantly grew from the beginning of the year when the “fake” ads started to run.
Now this effect isn’t as damaging as it could be because we also see a surge in positive reviews – Homescapes is a fun game after all, and that new audience coming in probably were happy to express their enjoyment once they got into the gameplay.
From StoreMaven’s data we do know that only about 15%-20% of users are exposed to the reviews, and only a fraction (1%-4%) interact with the review widget (which is how you read more reviews)
But still, these negative reviews have potential to hurt overall conversion rates over time, especially as they accumulate. (Not to mention the damage to the product managers egos after seeing their baby so harshly reviewed).
Additionally, If the overall rating goes lower than 4.0, conversion rates will begin dropping as well.
Source: Appfollow’s research into star ratings and potential installs. This correlates well with our aggregated rating test data.
Tying it all up — is it worth it?
So now we know that a “fake” ad can actually drive real (read: quality) installs, especially when the game has a strong affinity with different audiences that the ad attracts. That’s not going to be the case for every game, so trying to replicate this strategy is risky.
At the end of the day, false advertisement isn’t something that Facebook, Google, or Apple are going to endorse and although they may profit greatly from this strategy, they could also choose to crack down on it any day. As more and more users voice how upset they are from this strategy (Effect C) the platforms will be more likely to take note and act.
Plus there is a large, corporate-wide risk in running fake ads as it is illegal in many countries and large fines may apply.
But looking at the bigger picture here, this strategy gives us a few key takeaways:
- As UA platforms become more automated, it’s hard to break out of your core audience into new ones that could very well be likely to play and enjoy your game. There are certainly new and creative ways to do it.
- Understanding the organic lift associated with attaining significant volumes of new installs is an inseparable part of understanding the performance of your ads.
- Eventually, creating a growing group of users that speaks negatively about your brand all over the internet and on your app store page will hurt conversion rates.
Basically? Fake it til you make it isn’t your best marketing strategy. But showing off a more hidden side of you just might just be the way to grow.