Every database is unique in relation to its subject matter(data). I have found out that relational databases are more efficient to deal with than flat file based databases like excel. I have also experienced that the size of the database does not affect its input output performance. What really affect performance is how the database is modeled and the presence or absence of database objects like indexes, referential integrity, well tuned SQL queries and the number of database users. Transactional databases may be slower than reporting databases because of the constant input of data from different data sources. We can talk about normalization and online analytical processing( OLAP) or online transactional process(OLTP). How does normalization affect these two concepts? We can also talk about our experience with database users, DBMS, security, data sources and so on depending on what the interview wants to know.