Below is the link explaining my the table and the situation:
When comparing the two situations, one where the query is run using the column-store index and the other, where the query is run on a simple heap. When I compare the two results, I observe that even though the query with the column-store performs better than the other case, when simply run on a heap. But, the query using the column-store index involves a physical read(1) while the original one doesn't.
Both the queries have the same execution plan. Also, I run the queries in both the situations in warm and cold buffers. In a cold buffer, the original query requires 4 physical reads, while in a warn buffer, it required 0 physical reads. The behavior of the query using the column-store indexes however remains the same. Is there any particular reason behind this?