Use
By choosing Proposals ® Proposals from Queries, you can select created queries for the selected InfoCube, and view proposals for a minimum as well as a maximum aggregate for each InfoCube. The name of the proposed aggregate is derived from ‘MIN’ or ‘MAX’ and sequence number.

Use this function when you are optimizing an InfoCube for the first time. If you have executed queries before, use the other options for optimizing, because the number of times a query has been executed, and individual navigational steps are also taken into account here.
The minimal aggregate ‘MIN’ corresponds to the smallest aggregate possible, which contains only the data that is needed for the first time you drilldown on a query.

For this reason, set the query to hierarchical reading or reading on demand.
The drilldown defined in the BEx Analyzer is accelerated by you activating the minimal aggregate.
The maximum aggregate ‘MAX’ represents an aggregate that supports every navigational step of a query. This comes from the theory that the relevant query is drilled down over all characteristics (all free characteristics that is). Even if this happens less often in practice, this aggregate can be used by all the possible navigational steps of this query. But you must still check whether this aggregate really is of any use. These aggregates are often the same size as the InfoCube, as far as volume is concerned. In this case, do not use the aggregates.
If the components, meaning the characteristics and attributes of a newly defined aggregate that are not summarized, are identical to an aggregate that is already proposed, they will not be added to the list. Instead, the number of calls is increased by one.
In general it is the case that the higher the number of calls of an aggregate, the more useful the aggregate is. This means that the higher the number of calls of an aggregate, the higher the number of queries in which it can be used.
You can modify or delete the proposed aggregates.
See also:
Proposals from BW Statistics Optimizing Proposed Aggregates