Analysts and decision-makers use what-if analysis to assess the effects of hypothetical scenarios on historical data. Current On-Line Analytical Processing (OLAP) systems support what-if analysis only by physically replicating the data warehouse and modifying it according to the scenario. This process may take hours, hence limiting the applicability of OLAP.To eliminate this inefficiency, we built and OLAP toolkit, called Sesame, that exploits the following two opportunities. First, typically a small part of the modified data is needed to answer the hypothetical query. For example, the analyst, might be interested in results of the new strategy only during certain periods of time, and maybe 90% of the modifications were not needed. Second, data warehouses tend to have pre-computed materialized views, to help answer popular queries.