Last week I wrote some thoughts about using the Mac OS X Dashboard for business intelligence. I still think that is a neat idea, but so much has changed in a very short period of time.
The main thing that changed is obviously Google AppEngine. Like with every major tech story these days there has been a mass rush to judgement. The fact of the matter is that it is too soon to tell. Google is a somewhat untrusted (or at least untested) vendor when it comes to mission critical services; Google has also earned more credibility than anyone else when it comes to web applications. So, the jury is still out on whether or not AppEngine will the great or mediocre. One thing we do know is that the addition of Google AppEngine makes Google as a platform more viable. And isn’t becoming the platform the real goal for most tech companies?
Spreadsheets are still the BI tool of choice for most people. Still, Excel has some serious deficiencies for BI. Spreadsheets are typically Shadow Systems by the nature of the way Excel interacts with source of data. (Shadow systems are separate databases outside of the central data repository). This makes spreadsheets cumbersome. You either have to re-import a CSV or tell Excel to update data from an external source (i.e. an ODBC data source) in order for spreadsheets to stay up to date (or option #3 of update it manually). once you share a spreadsheet you now have multiple Shadow Systems.
I am excited by Google docs because they are easy to share and thus reduce the number of Shadow Systems. Unfortunately Google Spreadsheets don’t have the ability to update from a remote data source. This means that to update a Google Spreadsheet you either have to export a csv from your data source and re-import it or you can use the Google API to push data. This is still cumbersome, but the data update can be automated. One big problem with using Google Spreadsheets is that there are limitations on the number of rows you can have. Without being able to query a remote data source you would have to continue pushing data in.
With the beta release of AppEngine developers now have access to a database to use when developing on the Google Platform. It is not a relational database and I don’t expect it will be replacing any mission critical databases. However, I hear that BigTable is fast. Damn fast. I am wondering if it would be a good data warehouse environment. It would require some creative data modeling and there are limited querying functions, but it is fast.
If (and this may be a big if) Google allows BigTable queries from Google Spreadsheets then many of the problems with Excel would be overcome. The spreadsheet would be easy to share without creating Shadow Systems. An added bonus would be that multiple people can collaborate on a spreadsheet at the same time.
BigTable is not designed as a reporting database. It may work as one, but that is not the stated intent. What would be cool is an OLAP database as a part of Google Apps and possible Google AppEngine. I have heard that this may be coming. I am hoping it is sooner rather than later.
