How YouTube Influencer Marketing impacts app downloads
Are you wondering how well YouTube influencer marketing campaigns work? In this article, we’ll examine the impact of paid promotions in YouTube videos on app downloads.
YouTube influencers promoting your app: a winning strategy?
You’ve probably been through it: you’re watching the latest video from your favourite YouTuber when, suddenly, an ad pops up. Not the kind of external ad that stops your video, but a “paid promotion” made by the YouTuber him or herself! In an entertaining way, they tell you how great an app is, show the gameplay and some screenshots, offer extra coins or a temporary bonus, and, finally, all you have to do is download the app by clicking the link in the video description. Surely this is much more convincing than the usual banner shown at the bottom of the video. As top YouTubers can reach millions of people (think PewDiePie with his 107M subscribers), it’s tempting for marketers to go down this road, but is it really worth it? Let’s find out.
Ad for Hustle Castle by Amixem at 00:44
Empirical analysis: our methodology
In this analysis, we specifically focused on France. We took into account 773 videos published by some of the most popular French YouTubers, and filtered them to keep only those that contained a paid promotion for an app. We were left with 26 relevant videos from the YouTubers Amixem (6.63M subscribers), CYRILmp4 (4.58M subscribers) and MATH (1.47M subscribers). Then we compared downloads, both on iOS and Android, of each promoted app before and after the video publication. We wanted to determine whether an increase in downloads was correlated with the video’s total views or the YouTuber’s number of subscribers.
About AppTweak’s download estimates
We relied on AppTweak’s download estimates to build this analysis. If you want to know more about our algorithm, have a look at this article:
App Download Estimates explained by our Data Scientists
Impact of YouTube influencers’ campaigns on app downloads
Top campaigns increased downloads between 20% and 200%
Promotions having a positive impact on downloads
The video that had the best impact was a promotion for Hustle Castle on June 13th, 2020. Estimated downloads grew by 134% on the day the video was published and grew by 329% on average over the following week. A video usually has the biggest impact on downloads one day after its release. Then downloads gradually decrease during the following 4-5 days to return to their baseline.
But results differ greatly from campaign to campaign
However, a peak in downloads after a paid promotion is not the universal rule. Actually, some videos did not have a significant impact on downloads. Either the effect was too small to increase weekly downloads, or there was no visible impact at all. In short, we could not find a correlation between the number of YouTube video views and downloads, which makes influencer marketing on YouTube a risky strategy. It is also worth noting that, while AppTweak is confident in its download estimates, other factors might play an important role in downloads: Were other Apple Search Ads (ASA) campaigns conducted at the same time? Did seasonality, or any other relevant parameter, impact downloads? Finally, it should be noted that while we did not study retention for this article, it constitutes yet another factor to consider when assessing the real potential of YouTube influencer marketing.
Promotions having no clear impact on downloads
Results are not consistent for each creator
Creators attain very different results for different apps, and the number of views each video generates does not necessarily correlate with an increase in downloads. For example, for Hustle Castle (green circle), there were 25,000 extra weekly downloads after Cyril’s video compared to the previous week, but when Math published a video promoting Hustle Castle, the average number of downloads decreased the following week, meaning its impact was very low. On the other hand, for Raid Shadow Legends (pink circle), one of Math’s videos generated more downloads than Cyril’s, despite receiving fewer views.
In the chart below, each point represents a video. On the Y-axis, you can see the difference between apps’ downloads in the week following the video versus downloads the week before. Videos that promote the same app are aligned vertically, while their size accounts for the number of views in early September. The name of the creator is indicated inside the circle.
Each circle represents a video and its color stands for the app promoted in it. Downloads for promoted apps did not systematically increase during the week after video publication.
More views or more subscribers does not mean more downloads
We did not find a correlation between the number of views per video and the sum of downloads one week after promotion. Most videos in our subset had between 1 million and 6 million views in early September. While YouTubers with more subscribers tend to have more views, it does not necessarily result in any additional app downloads.
Each point represents a video and its color stands for the app promoted in it. Videos that got more views (bigger size and to the right) did not bring more downloads one week after publication to the app they were promoting.
YouTubers each had positive and negative results regarding extra app downloads after their videos, and videos of the most famous YouTubers did not necessarily bring more downloads.
Each “line” stands for a YouTuber, each dot for a video he made and numbers on the X-axis represent the YouTuber’s subscribers (in millions).
YouTube campaigns have no impact on keyword search volume
We found that branded keywords’ volume history did not correlate with paid app promotions on YouTube. We had initially hypothesised that traffic for a branded term would increase after the release of a video promoting the app, but this was not confirmed by the data. We assume here that people directly clicked the link provided by the YouTuber, which would then not affect search traffic, or that people would not massively search for the app during the few days after its publication.
Beware of cannibalization and external factors
In some cases, downloads surge directly after the video but will drop a few days later. Sometimes, this is due to external factors such as other promotions being stopped, or the app might have lost some category or keyword rankings. Or, from time to time, a phenomenon called “cannibalization” may occur. It indicates that a paid promotion will have a positive effect for a few days, but then downloads will decrease to a point below the baseline; users that downloaded the app, thanks to the video, on day 1 might have downloaded the app on day 6 anyway, without the video. Therefore, to fully understand a video’s impact, both extra downloads and potential drops must be considered. For example, Math published a video promoting the app One Punch Man on August 7th, 2020. As we can see in the chart, this brought 5,000 extra downloads over 5 days, but downloads subsequently continued to decrease, resulting in a loss of 3,000 downloads compared to their prior level of 4,000 downloads per day. It’s surprising to see such a decline a few days after a campaign. This can be explained by different external factors, and also demonstrates the eventual cannibalization effect.
Cannibalization – Downloads of One Punch Man just after a paid promotion on YouTube.
Promotions having a positive impact on downloads, followed by a subsequent decrease.
We began this article wondering how effective influencer marketing was on YouTube, and how it impacts app downloads. As we’ve seen, it’s a risky strategy; extra downloads in the week following a video highly vary from one campaign to another. In our examples, there was no correlation between downloads and a YouTuber’s number of subscribers or the number of views per video. However, there may be long-term effects that can still benefit your app. You should also pay attention to cannibalization effects that might reduce the impact of your paid promotions.
Should you choose this strategy, you will need to do your homework beforehand: study the influencer’s audience demographics, make sure your strategy is relevant to your app users, and evaluate your KPIs. Also, don’t only bet on the biggest influencers. It will cost you more and, as mentioned earlier, they won’t necessarily be more effective than smaller ones. In fact, this was confirmed by Darius Moravcik from journal app Reflectly at the App Promotion Summit London in 2019. They ran various tests between big influencers and micro-influencers at scale for the same budget and time duration. They found that micro-campaigns at scale provided 334% more installs than big ones. Therefore, you’ll need to evaluate different options to assess what works best in your particular case, but it might be worth giving micro-influencers a shot. And finally, if you want to know how to build an influencer campaign, have a look at this great guide published by Phiture.
Did you like this article? Would you like a similar analysis for another country? Would you like to talk about it? Don’t hesitate to contact me.