Kat’s verdictPush when the upgrade is useful and the extra cost fits your plan. Wait when the reward would persuade you to spend resources needed for something better.

Count the whole upgrade

Illustrative example only, not a current reward table. B means billion; M means million.

Suppose your food counter is at 4.8B and the next target is 5B. The gap is 200M food. An upgrade costing 200M of each resource would close it, but the total bill is 800M.

Hypothetical resource budget
ResourceExtra costNew rewards onlyNet spent
Food200M50M150M
Lumber200M50M150M
Stone200M50M150M
Ore200M50M150M
Total800M200M600M

If the newly unlocked rewards returned 50M of each resource, that would refund 200M of the 800M cost: 25%. You still spend 600M net, and you still get the upgrade.

The chest is contributing. It is not paying for the building.

Avoid the double count

Compare stopping now with completing the planned action. Count every new milestone it actually unlocks across the eligible categories. Exclude rewards already earned, even if you have not claimed them.

Keep stamina, materials and speedups separate. Count selectable resource chests only once. Food returned and ore returned are not interchangeable when food is blocking your next upgrade.

When the answer changes

Check the active event before spending: eligible resource or speedup categories, remaining time, progress and reward details. Do not assume a rotation or that every speedup type counts.

Background reading: Theria’s Consuming Return guide and its resource calculator. These explain event planning; they do not verify today’s rewards. The example above is invented to show the calculation. Use your own event screen for a real decision.

Plan the upgrade.
Count the extra cost.
Let the chest justify itself.