Typically transactional replication was used for load balancing and read only copy of database for reporting purposes.

– It is a failover solution
– any subscriber becomes a publisher
– data movement is bi-directional
– the status is known that replication is done so that it doesnt repeat again and again.
– flexible as both server can be both read or read-write simultaneously

Mostly DB Servers are read-intensive, so have more read-only nodes so overall I/O thoroughput can be increased.

Above also works as fail-over solution because if either of them fail the application can be easily pointed to the node acting as main DB.