OneLink campaign on Meta for app purchases

[elitchka]

Hello,

So I have the following situation:
- the advertiser has an iOS app (no android, and web LP is only there to funnel users from web to app) - this is not a sub app per se but it's profitability is highly dependant on repeat purchases
- we're testing a campaign with this web LP as a landing page vs a web campaign where we use OneLink - exact same creatives, targeting, objective, etc.
- there's no integration between Appsflyer and Meta - we only use the OneLink for the user journey
- we're tracking the key in-app conversions via CAPI - but not a typical server to server-side pixel one but we get conversions straight from the production databases and pass them directly to ad networks (this was a deliberate choice for tracking accuracy).
- we’re optimizing the campaigns towards these specific CAPI events
- the app doesn't currently have an ATT prompt implemented - they're about to launch that - so there's no consent collected at the moment

What I'm seeing from the performance of the OneLink campaign:
- the click to registration rate is much higher than for the regular web URL campaign (better cost per registration too)
- the conversion rate from registration to first purchase is similar to the web one’s
- but the number of all purchases and the average purchase value is considerably lower than for the web LP one - which is not what we were expecting.

How would you interpret what is happening here?

I know there's probably a lot to unpack here but I've not worked with this particular tracking context before so I don't know what are the implications of these choices.

 

[davamosk]

you are basically having an app journey vs a direct purchase in a LP which is way more immediate as the user has as shorter flow so it makes total sense that the purchases are higher in the LP if your app onboarding is not optimized or converting as good as the LP for the purchase.

I would:

• Analyze the messaging/angle of that LP vs your app onboarding and the connection of these with the creatives you're running. I'd be on the web, there's a better connection between the ad and the post-selling in the LP.
• Analyze difference in age, gender and placement distribution
• Pricing?

Still, I miss lot of data here so it's hard to say the driven.

 

[elitchka]

The web LP - that's only a page with download the app button and provides a very short and brief context. There is literally no way the user can convert on web, you can't even create an account - you have to download the app to be able to do anything. That is what's making this all the more baffling.
That pretty much bare LP features the same information as the screenshots on the App Store - so the context for a new user is the same.
The age, gender and placement distribution is roughly the same.
In terms of pricing, there is problem with the entry-level prices so to speak that are very very low....which is partially how we end of with a lot of these repeat purchases but extremely poor revenue.

 

[davamosk]

then maybe you are using different attribution windows on the web LP campaign that inflates the numbers? I would never do such a setup in any case. If you have to go over the app, I would go with app campaigns or a web2app properly setup with appsflyer and the facebook web integration.