I noticed that when running ads, the ads are not shown consistently, which results in my game having very few people half of the time and a lot of people half of the time.
This wouldn’t usually be an issue for most games, but the game I’m creating works best when there are a lot of people playing (it’s like a battle royale style game), so it’s really inconvenient to have the player count go up and down constantly.
In my opinion I think the old system where the sponsored games were shown consistently worked best for almost all cases, because the amount you bid would directly translate to how many times your ad would be seen, rather than the new system where you pay based on who clicks & plays your game, which I believe would result in less impressions for games with a higher CTR? I’m not sure about that that’s still confusing to me
I am experiencing the same issue. The ad system seems to use a “winner take all” bidding system where at certain points in time, it just decides that your ad should get a majority of the exposure out of a group of other concurrent ads, then revokes that exposure shortly after. This causes thrashing of the CCU which is not desirable
What I’ve found is that the average like ratio is consistently lower when running the new ads than with the old ads. This factor combined with the fact that the algorithm essentially shadowbans games that are running the beta ads (after the ads finish) due to instability across the board (thrashing CCU, drop in player retention, lower qualified PTR, lower average session time) means that it is absolutely impossible to achieve organic growth by utilizing these ads.
If you have a game with some small existing consistent playerbase and you want to promote it, I’d suggest against running these ads as they can negatively impact your game in the long run. It gives these ephemeral CCU spikes which probably look good from the standpoint of the algorithm (the AI algorithm is probably optimizing for maximum lifetime CCU across the duration of the ad instead of consistency), which actively harm your game.
This isn’t to criticize any of the Roblox team members working on this feature however. I believe it is a well intentioned feature that has the potential to be effective, but it just has some things that need to be worked out, some prioritized metrics that need to be reviewed and switched around, some weights changed, etc.
I have the same issue when I only run 1 or 2 ads,. I found if you run several (with different settings) it’ll give a more even amount of players over the same period.
okay, not a great experience trying this new system… 8 credits spent over 2 days for a grand total of 93 plays when supposedly optimizing for “lowest cost per play”. Was paying 1/8th that under the classic system, which was still far more per play than the original system. Paused my campaign because this looks like a horrible value that will not make sense to continue.
