As a Roblox developer, it is currently too hard to find an accurate estimation of ad impressions when advertising around weekends. This is because the estimation is currently based on data from the previous day, which fails to consider the increase of active users in weekends and holidays. As a result, the impressions given can sometimes be misleading and therefore lead to significantly fewer impressions than estimated.
For example, if I choose to put up a 500R$ ad on Monday, I might be given an estimation of 60,000 impressions due to the data from Sunday. However, as a large chunk of players are back at school for the week, the amount of impressions might be around 35,000 instead.
To solve this, I think it would be more beneficial if estimations are instead based on impressions from the previous week(s). This should also apply to any known dates where Roblox usually see an influx of active users, except the data could be based on previous years and/or previous weekends.
This would allow advertisers to see a more accurate estimation of how well their ad will perform, which will give them the opportunity to make more a informed decision before bidding on their ad, allowing them to make the most out of their bid.
There’s a lot developers can do with the data given to us there. The system suggested seems to be over-complicated for Roblox to make, especially year to year Holiday impression estimates. I think there are better ways to go about the ultimate problem here, which is making the most out of an ad bid.
I’ve made a tutorial on this, and to be honest, I am usually very busy to even follow this 100% of the time and I mainly go off of gut nowadays based on previous experiences, but the fact is the economic principle is there, if I were to get super intense about my data. Without anything Roblox had to do first.
However, here’s a quote directly related to this request within my tutorial