Confirm that the Mac actually slept
Close the lid or choose Sleep, then leave the Mac disconnected from power for a normal, comparable period. Note the battery percentage before and after, whether the computer was warm, and whether it was in the same place with the same accessories. One night can be unusual, so use several ordinary sleep sessions as your baseline.
In Activity Monitor, the Energy pane can show apps that are currently preventing sleep. It also shows Energy Impact and, on laptops, a 12-hour power measure. Those values are clues about activity while awake; they do not prove which process used a particular percentage during sleep.
Look for a repeatable wake pattern
A Mac may wake briefly for system work or because an accessory, network activity, or a connected display changes state. If the machine is warm or its battery falls much more on nights when a device is attached, test that device separately. Disconnect nonessential USB, Bluetooth, and display accessories for one comparable sleep session, then reconnect only after recording the result.
WattGuard can summarize sleep, normal wake, and dark-wake entries from macOS power logs alongside charge changes. Treat a wake reason as a lead to investigate, not a verdict assigning every unit of battery loss to that event.
Use a baseline, change, and recheck
First collect several comparable sessions and use their typical result as a practical baseline. Next make one reversible change, such as quitting a consistently active app, disconnecting a peripheral, or using the Mac on the same network. Recheck with the same sleep duration and compare it with the baseline instead of judging from a single result.
WattGuard builds its sleep baseline from the median of three to fourteen prior sessions. If drain remains high across controlled sessions, check battery condition in System Settings and consider Apple diagnostics or service. Battery condition and sleep behavior are different questions: a healthy battery can still be depleted by activity, while an aging battery can make a normal amount of energy use look larger in percentage terms.
What not to assume
Do not begin by changing hidden power settings or hibernation values from copied commands. Those changes can alter expected behavior and make the original cause harder to compare. The useful next step is evidence: what slept, what woke, what was connected, and whether one controlled change changed the outcome.