Testing Mobile App Acquisition Strategies: What Bidding Model to Choose? [Case Study]
This post is also available in:
PT 
ES
There are a lot of factors that will determine whether or not your mobile app acquisition strategy will be a failure or a success.
Yes, it all comes down to how many downloads you have at the end of the day, but it can be a long way to the point where you find the numbers satisfying.
One of the decisions you will need to make is what bidding model to choose for app acquisition, and testing different options for top offers is the crucial part if you ask us.
Here is the case study with one of the clients we have worked with for years. It proves exactly what we are talking about to be right, and gives a lot of insights into how the process of testing goes on here at PropellerAds.
A Mobile App Acquisition Campaign: eCommerce & Multiple Ad Formats
First, here is everything you need to know about user acquisition for mobile application – what is it, how to make it work, what are the necessary settings, targeting, what your creatives should look like, etc:
And now the story – this is how it goes.
| Geo: Brazil, Colombia, Indonesia, India |
| Offer type / Vertical: eCommerce, Social Networks |
| Bidding model: CPC, CPA Goal, SmartCPC |
| OS: Android |
| User activity: All |
| Ad formats: Classic Push ads, In-Page Push ads, Onclick Popunder ads |
| Case study time frame: November 2022 – January 2023 |
The advertiser behind the case runs campaigns regularly on different types of mobile applications – some of them are eCommerce, some are in a short video domain, and some are multipurpose apps intended for ordering food, and booking hotels and flights.
And he usually runs it with geos like BR, IN, ID, SA, MX, PK, and PH.
In most cases, the strategy was based on him buying push traffic on CPC and CPA Goal as those were performing the best (October 2022).
But then, in November last year, we presented a new bidding model called SmartCPC, and it was a game changer.
Now, these are the results our client faced before we started using SmartCPC. Take a close look at these, and later on see for yourself how the numbers grew.
| Bidding model | Impressions | Clicks | Conversions |
| CPC | 2 662 397 780 | 2 506 190 | 14 229 |
| CPA Goal + SmartCPA | 169 551 270 | 1 436 599 | 16 519 |
| Campaign | Impressions | Clicks | Conversions |
| Push | 2 830 184 502 | 3 937 201 | 25 851 |
| Onclick | 1 694 318 | 0 | 4 867 |
| Interstitial | 70 230 | 5 588 | 30 |
Why Opt for SmartCPC for Mobile App Acquisition?
Let me just explain in a couple of sentences what this bidding model does and why it is so special.
First and foremost, it is auto-optimized, unlike classic CPC which demands manual work. It allows you to apply automated zone management – you won’t need to monitor zone performance, and you won’t have to bother about what bit would better suit a particular zone.
Read more about it here:
This is what will save a lot of your time, but also money – with SmartCPC you will be able to purchase 20%-30% cheaper traffic.
The smart algorithm is helping with simplifying the bid selection process. It allows you to select the top bid you are willing to pay for a click, and avoid overspending on the zones where traffic price is below your top CPC bid.
November 2022: Changing the Situation in 9 Days
When we released the new SmartCPC bidding model for Classic Push and In-Page Push our Account Strategist suggested this advertiser to test it out. We believed this model had to suit perfectly the offers he was running.
The very first move was to test the top offers that perform great with CPC and CPA Goal. We chose a well-known short video app for Brazil and one quite similar for Colombia.
For the first app, the expected CPA (cost per action) was $1.2. During the first week of launches, we got around 95 conversions daily with an average price of $1.24. But every day the price went down a bit, which means SmartCPC was obviously doing its job.
| Day | Impressions | Clicks | Conversions | CTR | CPC | CPA |
| 22/11/2022 | 13 543 342 | 28 497 | 126 | 0.210 | 0.005 | $1.18 |
| 21/11/2022 | 18 018 620 | 31 398 | 143 | 0.174 | 0.005 | $1.14 |
| 20/11/2022 | 13 474 217 | 22 967 | 111 | 0.17 | 0.005 | $1.09 |
| 19/11/2022 | 11 394 246 | 21 495 | 88 | 0.189 | 0.005 | $1.32 |
| 18/11/2022 | 10 514 233 | 20 181 | 72 | 0.192 | 0.006 | $1.54 |
| 17/11/2022 | 12 343 135 | 20 091 | 83 | 0.163 | 0.005 | $1.32 |
| 16/11/2022 | 7 205 522 | 11 274 | 49 | 0.156 | 0.006 | $1.29 |
Yet everything else went up – in just one week the number of impressions increased by 10.000 and more, the number of clicks tripled, and we went from 50 to almost 150 conversions. Logically, the revenue had significant growth as well.
With app #2 expected price was $0.6 and we reached that goal as well, getting around 20 conversions daily with the average CPA being $0.59.
| Impressions | Clicks | Conversions | CTR | CPC | CPA |
| 11 098 389 | 22 560 | 160 | 0.203 | 0.004 | $0.59 |
To conclude: The first launches based on the SmartCPC model showed us great results and we decided to go further with it.
Our next move was to launch a similar app campaign in Indonesia. The app #3. And it was such a touchdown!
The expected CPA was again $0.6, and once again we made it real. We were getting around 500 conversions daily.
| Day | Impressions | Clicks | Conversions | CTR | CPC | CPA |
| 30/11/2022 | 19 886 034 | 69 736 | 476 | 0.351 | 0.004 | $0.56 |
| 29/11/2022 | 23 320 374 | 88 587 | 541 | 0.38 | 0.004 | $0.63 |
| 28/11/2022 | 34 065 163 | 139 414 | 740 | 0.409 | 0.004 | $0.69 |
| 27/11/2022 | 12 997 403 | 58 178 | 421 | 0.448 | 0.003 | $0.47 |
| 26/11/2022 | 19 919 362 | 82 416 | 591 | 0.414 | 0.004 | $0.50 |
| 25/11/2022 | 31 040 497 | 88 622 | 569 | 0.286 | 0.004 | $0.58 |
| 24/11/2022 | 27 919 884 | 84 359 | 532 | 0.302 | 0.004 | $0.60 |
| 23/11/2022 | 16 836 427 | 74 234 | 410 | 0.441 | 0.004 | $0.77 |
| 22/11/2022 | 7 550 786 | 21 301 | 149 | 0.282 | 0.004 | $0.57 |
In only 9 days time period the situation changed drastically, and our partner was well aware of it since his revenue and profits went up three times.
December 2022: Around 550 App Conversions a Day
At the beginning of December, we added app #4 to our SmartCPC. It was an eCommerce app for ordering food, booking hotels and flights, etc., and we launched the mobile app acquisition campaign in India with the expected CPA being $0.17.
This time we were just waiting for good results. And we got to around 550 conversions a day on Push traffic.
Everything was working as smooth as butter.
| Impressions | Clicks | Conversions | CPM | CTR | CPC | CPA |
| 427 500 200 | 1 067 854 | 17 078 | 0.007 | 0.25 | 0.003 | $0.17 |
January 2023: Reaching 159k Conversions
Building on this success, in January 2023 we adopted additional strategies to further enhance campaign performance.
These included introducing new creatives and testing different prelanders, both created by the advertiser and sourced from top performers in the market. We also implemented manual blacklists to improve overall campaign efficiency and keep CPA at a certain level.
Here is the table showcasing what was achieved from the beginning of January till the end of April 2023:
| Bidding model | Impressions | Clicks | Conversions |
| SmartCPC | 3 082 160 502 | 12 074 928 | 159 865 |
Spoiler: The results were so good that our client invested extra $37κ in his campaign.
Conclusion
The market conditions change in a blink of an eye, and what was working yesterday doesn’t mean will bring results in days to come. So we suggest two things:
#1 Testing, testing, testing!
#2 Trying out our new fresh features we are working so hard to deliver so your business can bloom.
In this case, implementation of our new SmartCPC model proved to be highly successful in optimizing the performance of the mobile app acquisition campaigns. By leveraging this model, we were able to generate significant revenue gains in a relatively short period.
But moving forward, we will continue to explore new avenues for improvement and deliver even greater value to our advertising partner.
