Developing a Mobile Business Intelligence Capability – Key Considerations
Disclaimer: I am an employee of Pfizer. The statements or opinions expressed are my own and do not necessarily represent those of Pfizer.
This post outlines a general technology direction. Pfizer has no obligation to pursue any approaches outlined in this post or to develop or use any functionality mentioned in this blog. The technology strategy and possible future developments are subject to change and may be changed at any time for any reason without notice.
The business need for faster, more intuitive and self serve access to business intelligence (BI) continues to grow. With the exponential pace of mobile device adoption, delivering on the promise of pervasive and real time BI is more urgent. New approaches to store and compute vast amounts of data have removed the cost barriers of delivering on this opportunity. The app stores have commoditized and reduced the cost of Mobile BI application development.
As disruptive as it may seem, the fundamentals of enabling an organization with Mobile BI are not all that different from rolling out a new business intelligence solution.
Mobile BI – What it is and what it’s not:
Delivery of analysis and reporting content to smart phones and tablet computers is loosely referred to as Mobile Business Intelligence. From a technology perspective, Mobile BI modules are usually extension of the enterprise reporting platforms that convert the BI content to work with the navigation construct and screen real estate of a Mobile device. Mobile BI is another distribution channel for your BI content. It builds on your current investments in data repositories, guided analysis, reports and data visualization. Usually, it’s not a ground up development of a new BI solution. The limited screen size is more suited towards simple visualizations and displaying aggregate values like key performance indicators. The focus is on distributing key pieces of information required for real time actions.
Garbage in garbage out, still applies. An iPad will not fix a BI solution that has performance, data quality or data latency issues, although it will make it look prettier.
Why do we need a strategy for Mobile BI?
Smart phones sales have outpaced personal computers. Tablets are geared to overtake desktop sales in the next couple of years. These trends will make mobile devices the primary source for creating and consuming digital content. The mobile device market is still emerging with new devices coming to market continuously. From a development and IT roadmap perspective, Android and iOS remain the dominant players in the Mobile operating system space. Blackberry still has a large install base in the enterprise space and analysts have really high hopes from Microsoft’s latest release of Windows OS. In short, a clear leader is yet to emerge. In this ever changing ecosystem, Business Intelligence platform vendors are trying to best meet the need for a Mobile BI product. Each vendor will have varying levels of interoperability within and external to their own BI offerings. An organization may have multiple mobile enablement efforts specific to a line of business. It’s important for data management strategists to develop a Mobile BI roadmap that factors in these internal and external drivers.
Key areas of consideration for practitioners and organizations developing a Mobile BI strategy:
· Lines of Business (LoB) systems continue to move into a software-as-a-service (SaaS) model. It’s realistic to assume that an organization will work with multiple SaaS vendors. Each vendor will have its own standards and offerings around Mobility in general and Mobile BI in particular. Consider the impact from multiple Mobile BI delivery methods. Some vendors may have an feature rich app, others may offer a app that’s a refined way to navigate through their web interface and yet another may have an HTML5 implementation.
· Big Data storage and compute has a fairly robust set of offerings, both form a product and system integration and development perspective. However, few have the big data analysis and insight distribution figured out. Predictive intelligence is of little use unless we shorten the amount of time it takes to be acted upon. Mobile BI is a key enabler for reducing the ‘action latency’.
2. Mobile Standards
· Watch the developer community closely. The make or break for a mobile operating system has hinged on the strength of its developer community. The same holds true for Mobile BI platforms as well. The more the developers, more the number of innovative apps available on that platform. Mobile BI Vendors will need to ensure that their products offer a robust set of application programming interfaces (API) and licensing structures to make it worthwhile of an app developer’s time. Understand the Mobility Strategy of the vendor in detail.
· BYOD - BI on Bring Your Own Device doesn’t mean any and all variants of devices should be supported. Organizations should standardize on one or two devices and operating systems. For multinational companies, the standardization choice should take into account in-country and geographical preferences for a device/OS combination.
· Flash - Keep an eye on development standards. A position by a technology vendor with a big clout can move development standards rapidly and radically. Adobe Flash is on its way out with the momentum shifting to HTML5.
· No matter how sleek your Mobile BI app, the network coverage and connection speed will be crucial to the overall user experience. Talk to your Infrastructure guy about your organizations Bandwidth woes and be prepared for an hour long monologue. It’s not unlimited and its costly. The growth of mobile devices, mobile workforce and real time updates on everything are making bandwidth a precious commodity. Be sensitive to how much of your Mobile solution needs to be bandwidth bound. The content design should adequately factor in slow connection speeds and choppy coverage.
· Connectivity aside, device management and monitoring will need to factor into the overall Mobile strategy. Suffices to say that a Mobile BI strategy needs to be lock step with the overall Infrastructure strategy around mobility.
· A Mobile BI data protection strategy should go beyond identity and access management (authentication, authorization, and entitlements). Think about remote swipe of in-app data. This is separate from the remote swipe of the full device which is useful for a lost or stolen device. Ability of remote swipe within an app used to remove the ‘reports’ stored within the Mobile BI application remotely. Usually this is triggered when a BYOD user changes roles or leaves the company. Remote uninstall should be considered for license management for seat based app licensing.
· Many choices of in-app integrated cloud storage are available. Consider your companies policies around cloud storage of enterprise data. Evaluate if the Mobile BI app of choice enables management of in-app capabilities
· Factor in the ability of the BI platform API/SDK to enable User/Group Management, Single Sign On, Object Level Security, Row Level Security for the Mobile BI components.
· Who is your audience? Think about who can benefit. Mobile delivery of business intelligence is not limited to real time data on cost, revenue, margins and distribution channels for sales or finance or the executive tier. The real value of BI is when it’s pervasive in an organization. Mobility makes it possible to push BI to the edge of an organization. Arming the teams that work directly with customers, vendors, partner is where the business case is. As it is with any BI solution, a Mobile BI implementation should be geared towards serving the needs of the targeted community of users. Expectations of an executive tier user will be different from a sales representative or a scientist.
· User requirements for Mobile BI solutions are usually around:
a. Self-server access to information in on-premise or Cloud LoB systems (CRM, ERP etc)
b. Access to reports and metrics optimized for context sensitive mobile navigation
c. Offline report access to stay current with customer information
d. Alerts – Subscribe to alerts and be notified of data changes and refreshed reports
e. Over-the-air installation – Deploy the software at convenience using a wireless connection
f. Ability to collaborate form within the Mobile BI app
g. Should be intuitive with little to no training requirements
· Content design considerations
a. Mobile BI is primarily graphical in nature. Developers need to be strict followers of rules around visual display of data. The graphing choice should invoke thinking and lead the eye to seek comparisons. The navigation should be intuitive and context driven. Considerations should be made for the screen real estate. Cramming a lot of information in a small space may make it hard to read.
b. Consider a standard template based approach. BI content aimed for Mobile distribution requires a tighter partnership with business. For this, a developer will need to know in advance the various navigation permutations a business user may take.
The next post will be on key considerations for Big Data Analytics. Stay tuned…
@MayankMisra | http://www.linkedin.com/in/mayankmisra