Michael Fitzgerald of Computerworld offers an article to the effect that mobile business intelligence is hot. He cites just about every vendor except Microstrategy as seeing or indeed pushing this trend — and that probably just means Microstrategy didn’t return his call quickly enough, as they’re betting heavily on the mobile BI trend themselves.
In essence, mobile BI seems to be about small, portable dashboards. Now, I’ve been critical of dashboard technology for years, because of myriad ways in which it fails to live up to the potential of decision support. Some (not all) of those criticisms are being addressed by more recent dashboard technology developments. But with one exception, those criticisms are of little direct relevance to the mobile case.
What’s going on in mobile BI is not so much general decision support as it is quick information retrieval and navigation. Use cases include:
- Getting information about your prospect right before a sales call
- Checking on your customer’s order status during a sales call
- Checking up on a logistics or maintenance issue, in areas such as:
- Airline (re)scheduling
- Truck/warehouse dispatching
- Medical procedure availability
- Heavy equipment maintenance and repair
- Comparing against benchmarks the performance of something at your current location before you drive away, where “something” could be:
- Customer purchases
- Machine uptime
- Floor cleanliness
Get the UI right on a handheld device is of course a big challenge, due to the small form factor. But any analysis being done in these cases is very simple,* and the same can be said for query execution. So a number of the usual issues in BI and analytic data management fall by the wayside. Other than the small — ahem — matter of squeezing things onto iPhones and the like, what’s going on in mobile BI is pretty straightforward stuff.
*I’m sure sophisticated and flexible handheld analysis is a nice-to-have. But I expect use cases in which it is a must-have will remain scarce for at least the next several years.
But there’s also something that’s evidently not going on, which highlights one of the most perplexing areas of non-progress in all of software — alerting. Alerting technology is clunky and inflexible. Hence it is plagued with intrusive false positives. Hence people don’t want to use it much. And so they miss out on some of the greatest potential advantages of analytic and mobile technologies alike.
I last wrote up my vision for alerting technology 11 years ago. It’s probably time for me to do so again. But you know what? When I do, very little will have moved from the “Should do” to the “Already done” ledger in the intervening decade.
Edit: Here’s some of where I think alerting technology needs to go.