It will always depend on your architecture: servers involved, replication type (Trans, merge, peer-to-peer etc), workstation involves and so on. Push will put more pressure in the publisher server and pull on the subscriber server or workstation. Pedro R. Lopez

Fair enough, in this case you can take help of log shipping (manual) to perform updates between central repository and client's end (laptop). Also replication will help you to update only required tables. If the client's database is READONLY then simply you could schedule a manual job to backup...