It is like accounting for the usual holiday rush before deciding whether a shop has become busier than normal.
Go deeper
Employment, prices and other economic measures often follow seasonal patterns. Holiday shopping, school schedules, harvests and usual weather changes can affect the numbers at similar times each year. Statistical agencies estimate these recurring influences so a monthly comparison can reveal movements beyond the normal seasonal pattern.
Adjusted and unadjusted figures answer different questions. The unadjusted series records the measured level before this adjustment; the adjusted series helps compare neighboring periods. Seasonal adjustment does not remove every unusual event or all statistical noise. Its estimates can change as more observations reveal how seasonal patterns are evolving.
Make it concrete
Hypothetical example: retail employment falls after holiday workers leave. If the fall is smaller than the usual January decline, the seasonally adjusted series could rise even while the unadjusted number falls.
When you read a headline
Compare figures with the same adjustment status. A seasonally adjusted increase does not necessarily mean the raw headcount rose, and one month's change may still be noisy.