When I was working on international Stack Overflows, my workday mostly consisted of planning and thinking through events that we regularly held in our communities: contests, competitions, offline meetups, etc. So when the folks at SolPlex asked me to help them with their community, my top priority became planning and running events for their community as well.
Reflecting on our collaboration, we decided to answer one very interesting question:
Do the number of active users and the number of messages increase on event days?
To answer this question, we looked back at data from 58 activities (AMAs, game nights, promo code giveaways, etc.) held from October 2025 to July 2026. The results of the analysis are below.
As always, huge thanks to Pavel for the technical side of this analysis.
Results
TL;DR Yes, in most cases, the number of messages and the number of unique active users increases on event days.
There’s one important caveat. It’s impossible to give a definitive answer to this kind of question. Since we can only live in one reality, we can’t run an A/B test and measure the impact: if an event has already happened, there’s no way to know what would have been if one hadn’t run the event that day. At the same time, we can still provide some estimates. For example, we can do the following:
- Look at the activity level on event days.
- For each of those days, simulate a what-if-no-event scenario and estimate the resulting activity level.
- Compare the actual activity with the estimations from the statistical model.
The what-if simulation can be fairly simple. For instance, you can look at the average number of actions for a given day of the week, adjusted for trend. Or you can use more advanced statistical models like Facebook’s Prophet or Google’s Causal Impact.
For this research we decided to use a simple regression model. Here’s what we got.
-
Game Nights. On game night days, the number of messages was on average 4.1 times higher than simulated, and the number of active users was 3 times higher.
-
Contests. On contest days, the number of messages was 2.6 times higher than simulated, and the number of active users was 1.9 times higher.
-
AMAs. On AMA days, the number of messages was on average 1.1 times higher than simulated, and the number of active users was 1.2 times higher.
On the day of releasing the video about the game, we saw 7.9 times more messages and 4 times more active users. On bounty task days, we recorded 5.1 times more messages and 4.4 times more users. On promo code giveaway days, there were 1.3 times more messages and 2.4 times more active users.
Conclusions
Game nights and contests showed the best results across a large sample of events. The video about the game generated huge interest among players, as did bounty tasks, but there weren’t many of those events, so it’s hard to make strong claims. There were also events with a negative effect. For example, on days when contest results were announced, instead of a spike in activity we saw a slight decline.
The mere fact of running an event, of course, does not guarantee a surge in activity. Events need to be run wisely and at the right time. If you need help figuring out which events are right for your community, we’d be happy to assist!