It's been quite a while now since Chris posted chapter one of his "Case against Google Maps in IBM Cognos BI" trilogy. Not surprisingly, that series generated a lot of buzz, both pro and contra.
Analyzing data across business dimensions with business intelligence is pretty mainstream these days. Enterprises have been pretty actively procuring and implementing BI for the past 20 years or so, a few longer than that. A lot of us have worked for these big companies that use BI every day. We invest in those same companies, and possibly use their products.
Whenever I listen to a company's quarterly earnings call, I note that almost without fail, the numbers presented are done so based on regions. As in:
"The America's fourth quarter revenue were $xxx million, an increase of yy% over the preceding quarter. Revenue for EMEA was ...."
You get the picture. Revenue, costs, and overall performance always has a geographic component. And within each of an organization's major regions, there are sub-regions... countries, sales territories, service areas, regional markets.
So it makes perfect sense that organizations would want to see their performance on a map, doesn't it? By major region, by country, by... whatever. And it's likely that their data is warehoused exactly that way, so such a represention ought to be simple, right?
So, let's see just how easily one can do that using the Google Maps V3 API. If you look at this documentation about the Google Maps polygon object, you'll see that it takes several lines (40+, including comments) of code to implement a simple triangle on a map.
Imagine doing that for a custom sales territory comprised of a combination of ZIP codes. Or a direct marketing area.
Of course, there are always ready-made KML files that one might find, download (or create!), and use for this purpose. That'll draw the regions.
Now think about hooking those polygons up to BI data so you can thematically shade regions based on IBM Cognos BI measure values. Or defining broadcast events from regions to IBM Cognos report objects so that when users click on them, related data from that area is filtered in those report objects. Neither of those is a trivial task, and take more than a few lines of code. Oh, and don't forget all of the security you've built into your BI, which has to be leveraged in the map. Or the requirement to output map-enabled reports in PDF format.
Contrast that considerable coding effort with the ability in SpotOn Vantage Maps to combine Esri map services and regions with IBM Cognos BI data, with no code, using a drag and drop interface. Then thematically shade those regions based on multiple measures (or attributes) again with no coding. Then broadcast from those regions to supporting lists, crosstabs, and charts: no code.
It's not an open condemnation of using Google Maps or the Google Maps API. It's an affirmation that when you're goal is to
integrate maps with IBM Cognos BI, there are easier and cheaper ways.