go to the store at 3 am
A Roblox experience owned by Pixl-ish Studios.
HorrorENhorrorobstacle coursemultiple endings
go to the store at 3 am is a horror experience created on 2023-11-06. It has accumulated 3,494,279 lifetime visits and holds 6,890 favorites. The title maintains a 79% like percentage. Currently, zero users are playing. The game features nineteen distinct endings and was developed by PixlDev.
Description (written by the developer; not a statement by this register)
Your parents have given you the task of going to the store for beans. Besides this task you also need to take out the trash There are various obstacles that are in your way! Can you complete the tasks in time or will you die trying? There are 19 endings, will you try to beat them all? - Good - Bad - Spoiled Beans - Ran Over - Ran Over (MrBean) - Shot - BearTrap - Arrested - HouseFire - Deliverant - Forest Fire - Escape Simulation - Heaven - Hecc - Stole FireTuck - Granny Good - Granny Bad - Sewer Good - Sewer Bad Built and scripted by PixlDev (Pixlish_Studios) You can send ending suggestions in my group. Please like and favorite if you enjoyed! "Go to the store at 3 am" V3.0
Last updated by the developer 30 Jul 2026, 19:06 UTC.
Live players over time
Growth
Growth history is still building. We hold one measurement for this entry. A trend needs at least two, taken days apart — so this will fill in on its own rather than being estimated. We do not publish numbers we have not measured.
| Measured (UTC) | Players | Change |
|---|---|---|
| 10 Aug 2026, 23:23 | 0 | first reading |
Concurrent players at each of our measurements (insert-on-change).
Monetization — game passes
Game passes for sale and their prices, measured from the Roblox passes API. Developer products bought inside the game aren’t publicly listable.
Badge funnel
How many players have ever earned each badge — a proxy for how far players get before they stop. Steep drop-offs mark where a game loses people. Win rate is the past-day share of players who earned it; categories are assigned by a language model from each badge’s own name and description.