In the context of business analytics mobile analytics is declining in popularity

The promise of mobile business intelligence has been with us for a long time: anyone remember Cognos Go! for the Blackberry in 2007? At the time, it was a novelty, or at best, an inconsequential feature to fulfil a niche requirement. Since that time however, you have probably noticed that mobile devices have became smart and ubiquitous. This graph from The Economist  (story here) illustrates the trend: tablet usage came out nowhere around 2012, mobile is on the rise and the desktop is in steep decline. Even if these trends are levelling off, this has been a huge shift in the past 5 years. A quick look around any office will give anyone anecdotal evidence of how these trends have played out in the workplace. Business people are bringing their consumer habits to work with them, encouraged even further by the growing popularity of BYOD policies.

To understand how much of an impact the shift to mobile has had on our behaviour, consider Facebook: it was not always obvious that Facebook would be the dominant global social network. A key turning point was the decision by Mark Zuckerburg circa 2011 to reorganize around a mobile-first strategy. In 2012, VP of Product Chris Cox was quoted, “no one can present a new product concept without a mobile mock-up”. Zuckerburg and his executive team were quick to recognize the trend illustrated in the above graph. They had the foresight to see how important a mobile interface was going to be in maintaining high user engagement across the globe.

Many other consumer tech companies also embraced “mobile first” strategies but the Facebook story is significant for three reasons:

1. Facebook was already a successful company, with a huge user base happy to use the desktop web browser,
2. It was not clear how a mobile audience could be profitable, as there was uncertainty around how their current advertising based business model would work
3. They continued to iterate on their mobile first strategy by breaking apart the monolithic Facebook app into smaller pieces, creating tailored user experiences optimized for mobile devices. Gigaom characterized it as “The Great Unbundling — the process of taking discrete parts of the monolithic social network and breaking them down into individual apps, such as Instagram.”

The Facebook desktop browser web app has everything: the news feed, messaging, photo tagging, groups, etc. Instead of creating a mini version of the desktop Facebook app with a comprehensive set of features,  they crafted user experiences around a limited feature set, specifically designed for smartphones.  Facebook Messenger, Instagram and Paper are all separate from the core Facebook mobile app. The theory is that a single app cannot be all things to all people on a mobile device, where screen real estate is limited and user interaction occurs in short bursts.

A similar two step shift in thinking needs to happen in order for business intelligence and analytics to become pervasive in an organization: one, make the mobile BI user experience a project priority, and two, not packing everything into a miniature web portal. The mobile interface needs to be intuitive and engaging enough that users turn to it instinctively.  A report or data visualization that doesn’t render correctly on the smartphone will be dismissed as broken, not as something to be revisited later when I get back to my desk.

There are many considerations when making mobile a key component of a larger BI project:

• What are the mobile use cases? Where will business users be located? What will they be doing?
• What is the point of entry to the mobile BI or analytics portal? A mobile app, an email link or a something else?
• How much data will be pushed out to users versus pulled down on demand?
• What are the network bandwidth constraints? How much will this vary by location?
• How can email and text message be used?
• What is the preferred method of sharing and collaborating from the mobile device?
• How will data security be enforced?
• What are the impacts on the system architecture and hardware requirements?
• How should dashboard layouts be modified to accommodate smaller screen space and reduced attention span?
• How will different mobile devices be accommodated? What is the impact of the myriad of hardware and mobile operating systems out there?
• What level of self-service and interactivity can be offered?
• How can mobile usage be audited and measured?
• What are the key measures of success for mobile BI?

All the major BI vendors have mobile options but the feature set varies. Many claim to have a write-once, render-anywhere development environment but the results are mixed at best. Developing for mobile requires thinking differently about the layout and user interface, as the number of pixels available vary greatly from one device to another. Care must be taken to not produce a user interface than only someone with perfect eyesight and tiny fingers would love. The report or dashboard has to be reduced to the essentials that will influence the decisions and actions at hand, in the context of where the user is located at that moment. Mobile, by definition, means that the user could be anywhere, in a customer boardroom, on the shop floor, or in an airport lounge waiting for a flight.

The organization that prioritizes mobile users on new BI projects will see adoption rates soar. However, there is no magic mobile sauce that can be applied to an existing BI solution. We interact differently with our smartphones, the expectations are different, and herein lies the opportunity for delivering real value: a personalized experience with an intuitive user interface, delivering context sensitive information at the right moment. As the Facebook example implies, it is not just enough to have a functional mobile experience, but an experience designed for mobile.