Not only is the process of normalisation valuable for increasing data quality and simplifying the process of modifying data, but it actually makes the database perform much faster. To prove the point, Peter takes a large unnormalised database and subjects it to successive stages of normalisation.