A friendly warning: you will not find 300 ROI in the cases below 🙂 But earning money wasn’t our goal this time. We needed to test out the technology/solution and demonstrate it to you. And we did it!
And now let’s dive into the PWA, event optimization, and what we have discovered.
Step 1: hypothesis
So, we assumed that PWA can be used as an alternative to Webview applications. Alfaleads has already had a lot of work done in this direction — we have our own construction tool. Moreover, our PWA can send push notifications to users.
Step 2: development
However, something was missing — we still couldn’t set up event optimization in FB. And now can proudly announce that we have finally managed to implement the event postback and we can work with event optimization.
This is how it works: there is a landing page that redirects users to PWA. This landing page has pixel tracking and powerful cloaking that filters out bot traffic. From the landing page, user goes to the “store” PWA page, which looks just like the well-known Google App store. The user downloads the PWA which has a Webview with the offer inside. Then he registers and makes a deposit. These events are posted back to our tracking platform, and we send them to Facebook from there.
Facebook has a learning mechanism that is trained to find the audience that performs target actions more often. For example, optimization for registration will increase the number of registrations and the CTR. As a result, the number of deposits should rise as well.
Now we only have to prove that the approach works in real life.
Step 3: practical application
In our experiment we took a Russian casino offer and compared the traffic flow in two cases: a) when driving directly to the offer page (with a pre-landing page); b) when driving to a Google Play replica page with a link to PWA.
Why don’t we compare well-known Webview apps and PWA? It is no secret that a Webview app has a lifespan of 3–5 days on average (or none at all). Traffic loss is huge, accounts get banned, etc. For this reason, we turn to alternative schemes.
At the same time, we are studying performance of this approach for a different GEO — Australia. As soon as we have enough data, we will make a case demonstration for you.
Driving traffic directly to the landing page
We have implemented optimization for deposits and registrations on our landing and pre-landing pages. We tested out two options: using our own pre-landing page or choosing one from the advertiser database. Technical issues grew in number dramatically when we used our pre-landers and domain: the clickID was getting lost, the layout of pages was distorted, some devices displayed the pages incorrectly. Multitudes of variables undermine stability. Thus, it was necessary to monitor the process at all times. Therefore, it is much easier to use landing pages provided by the advertiser.
Launch algorithm:
- Activate the offer
- Set up postback
- Generate the link
- Create campaign in the tracker
- Integrate the tracker with Black Page
- Set up cloaking
- Test the technical aspects
- Post the link to incorporate the Facebook pixel ID in it
- Create a campaign on Facebook with “Conversions” objective + “Complete Registration”
Results:
Optimization for deposits did not work well, as the Facebook pixel does not have enough events with only one account. But optimization for registration went successfully: with the decrease in spending the number of registrations remained stable.
Driving traffic to PWA
For our PWA experiment, we implemented only optimization for registration. We noticed that with PWA there were more deposits and the completion rate was higher. Considering the possibility of sending push notifications we get the maximum effect!
Launch algorithm:
- Activate the offer
- Set up postback
- Create a campaign in the tracker separating unique/non-unique users
- Post the link to Google Play to incorporate the Facebook pixel ID in it
- Create a campaign on Facebook with “Conversions” objective + “Complete Registration”
Results:
The results were more impressive than on average for classic Webview applications. For comparison, here is the average data for the applications: inst2reg = 1:10, reg2dep = 1:8.5. However, here you need to consider the installation price — in case of PWA it may be higher than with usual applications!
And here is more statistics on completion rates, which are crucial especially if you work with a high-payout GEO.
Volume
To confirm that the solution is also viable for large volumes of traffic, we attach statistics from our Alfamedia media buying agency.
Conclusions
Based on the results of our experiment, we came to the following set of conclusions:
- Apps are not the only working thing out there.
- Implementing event postback into PWAs grants a sort of a loophole enabling us to use this tech most efficiently.
- Optimization for deposits in our conditions did not provide any advantages: in the iGaming vertical (frequent bans, ad rejection, etc.), Facebook pixel does not have enough events on one account, while sharing the pixel within several Business Managers leads to an immediate ban. During the test, optimization for registration worked well both with PWA and with landing pages.
- Setting up landing pages is a long and complicated process. In contrast, campaign launch with PWA is rather easy and even enjoyable — just compare the algorithm in both cases.
Alfaleads provides affiliates with:
- Free access to the PWA-construction tool
- Adaptable landing page that will be adjusted to the affiliate’s approach
- Tips on creatives for PWA design
- Set up assistance: affiliate manager will simply issue a link for you, so you don’t need to dive into the guides and browse through tech issues
To repeat (and, of course, improve!) our case and start working with PWA, contact your affiliate manager — he will give you access to the construction tool and supply you with everything you need to start the journey. We provide a ready-made solution for PWA campaigns — you just need to choose the right offer!
Want to learn more about PWA? Find more information here!
Comments