Search Unity

  1. Megacity Metro Demo now available. Download now.
    Dismiss Notice
  2. Unity support for visionOS is now available. Learn more in our blog post.
    Dismiss Notice

Details on Unity Monetization SDK 3.0 User Classification

Discussion in 'Unity Ads & User Acquisition' started by ferretnt, Oct 23, 2018.

  1. ferretnt

    ferretnt

    Joined:
    Apr 10, 2012
    Posts:
    412
    Hi.

    Can Unity expand on the very high-level description from their blog of the User-specific monetization with personalized placements? From the blog, it says this:

    "For example, player A enters your game. She does not typically make in-app purchases..."

    The definition of calculating "she does not typically make in-app purchases", is quite important, so can Unity please expand:

    - After install, what is the expected timeframe before a user is classified?
    - What variables are fed into the ML for "does not typically make in-app purchases" - is this based on your apps, or is unity tracking user IAP across all Unity IAP-based apps so that we can benefit from that dataset?

    We already do something similar, using Firebase Predictions and segments. This works, but typically we have to wait for a long time after install (multiple days) before a user is classified so that they can be targeted appropriately.

    The joint ads/IAP approach sounds great, but we need a bit more detail before changing our current implementation.
     
  2. mikaisomaa

    mikaisomaa

    Unity Technologies

    Joined:
    Sep 14, 2015
    Posts:
    365
    Hi,

    Personalized placements and LTV valuation gathers data from the user's IAP and Ads activity and unless opted out from personalized data collection - this data already exists when the player enters your game for the first time (assuming the player has already been in other games with Monetization 3.0).

    The gathered data comes from other apps using Ads + Unity IAP - your monetisation will benefit from data gathered from many other games.

    Additional benefit here is the lack of need for segments - our ML algorithm treats every player uniquely.

    The previous model optimised for revenue on the spot for each ad, whereas the new model optimises for lifetime value of each user.

    I hope this helps.
     
    Rutger and ferretnt like this.