Analytics simplified and cubed.
PowerOLAP bills itself as a straightforward, fast, and affordable OLAP
alternative (darned if I can find pricing information on their Web site,
though; some VAR sites suggest you're looking at upwards of $1,000 per
seat). I took a look at their 3.0 release (they've announced 4.0, but
3.0 is still what you can download to test out). As with most applications,
there are things to like and things to dislike here.
On the plus side, the PowerOLAP engine does seem to be reasonably fast
and flexible. It's also well-integrated with Excel, and once you've constructed
cubes, business users should have no problem manipulating the data.
However, I'm less a fan of the initial setup process. Apparently, your
administrative users or developers must define cubes by inputting the
metadata into the PowerOLAP interface, rather than by working interactively
with a relational database. There is a facility for importing metadata
from text files, but this is less convenient than the sort of automatic
extraction and tight integration with data sources that you'll get from
some other OLAP products.
The 4.0 release adds a number of intriguing features, including OLE DB
for OLAP compatabilitywhich should make it much easier to define
cubes in a product such as SQL Server Analysis Services and then work
with them through PowerOLAP. There are also strong real-time ties to various
accounting and ERP systems and various performance improvements. The PowerOLAP
connectivity story is also strong, including automatic links between the
native relational data and cubes, synchronization between servers, and
a Web Services interface to allow OLAP analysis in the browser.
Mike Gunderloy, MCSE, MCSD, MCDBA, is a former MCP columnist and the author of numerous development books.