After the fact: was it a setting, or was the geometry always too much to ask?
The Overhang and Bridge Risk Scan on this site helps before you print; this tool helps after, once a specific overhang or bridge has already come out sagging and you need to figure out why. The two broad explanations, a settings problem (cooling, speed, layer height) or the geometry genuinely exceeding what any settings tuning can fix, call for different next steps, and this triage separates them from what you observed.
How the triage works
An unconfirmed cooling fan (you have not actually checked it was spinning at that section) weighs heavily toward insufficient cooling, since a silently stalled fan is a surprisingly common and easy-to-miss cause. A recent print-speed increase weighs toward speed exceeding what the feature can handle. Genuinely unsupported geometry (no support material used at all) weighs toward the feature needing an actual support strategy or a reorientation, not just a settings tweak.
Worked example
A sagging overhang, cooling fan not confirmed working, no recent speed change, no supports used: the tool ranks insufficient cooling as the top hypothesis given the unconfirmed fan state, and recommends physically checking the fan is spinning as expected during that section before assuming the geometry itself is the problem.
A common mistake
A frequent error is assuming a fan that spins is a fan that is actually delivering airflow to the print. A partially clogged duct, a fan running at reduced power, or ducting that shifted out of alignment can all look fine at a glance while barely cooling the part. If cooling is the top hypothesis, check actual airflow at the nozzle, not just that the fan blade is turning.
Limitations
This tool ranks plausible causes from the symptoms you described. It cannot see your printer or measure actual fan output, so treat the ranked list as a place to start testing, not a confirmed root cause.