Mobility and mobile data are less mature than big data, probably because it's much harder
One of the most attractive crossroads with mobility is big data and analytics. The idea of making decision-enabling data available where and when needed in an easily consumable format. This opportunity lies at the heart of what many corporations today intend to exploit in order to drive productivity or smarter and more timely decisions at all levels of the organization. There are probably limitless permutations of this opportunity but to name a few we are talking about.
- The Sales Rep standing in front of a customer store manager with immediate access to sales data analysis.
- The Key account manager able to deduct yesterday’s downtrend in global sales while travelling to meet with the customer counterparts.
- The Digital/Social manager able to be alerted to and analyse the overnight flood of updates related to key brands while heading to an industry conference.
Now, this strategy often assumes on always-on mobile networks to support the desired capability. In a few cases, this may be an OK assumption, but mostly the solution should take into account that the mobile network may only be available sporadically during the day especially high speed mobile networks (WiFi, 4G, 3G). On the other hand while the big data enabling for a particular initiative may be very hard to resolve in a single instance, once the right data structures and solution has been setup, it’s there for consumption. This is why I’m referring to that resolving for the sporadic data connectivity is much harder (and much less likely to happen quickly) than resolving for a particular big data analytic need.
So what should one look for when going after such initiatives. We need to examine the mobility solution at hand and how it enables the presentation of the big data analysis at the time it’s expected to be consumed. Assuming we are looking at a scenario where mobile network connectivity cannot be guaranteed, then the solution itself needs to contain mechanisms to buffer for this immaturity in the networks and best case shield the user from this fact and present the desired analytic capabilities nevertheless.
This can be done by allowing the solution to intelligently fetch data in the background while connectivity exists and thereafter simulate an online experience whilst the device is off-line. Analysis of real use cases need to be applied when building such a solution as the device will have limitations in terms of storage and compute capacity and will get nowhere near the capabilities of the shiny big data back-end. However by applying logic of frequent usage/consultation, role based data segregation and security and other similar known factors, the solution can be build intelligent enough to bridge the network deficiencies.
For those developing enterprise mobility software for a living, you may be aware that more modern development frameworks and best practices actually enables such behavior very elegantly by providing standardized mechanism for achieve exactly what I’m describing. I will not mention any specific names here, but looking at the top three enterprise mobility development frameworks today, you will surely find these references.
For a more exhaustive analysis of typical problem scenarios, not necessarily limited to mobility/big data, please see mobility blogger Adam Sivell’s ebook Mobile Enterprise Tips and Tricks.