Johannesburg, 20 May 2011
The ultimate goal of mobile business intelligence (BI) is to provide useful information to decision-makers at any place and at any time. That is quite a loaded statement, as the successful deployment of mobile BI has a number of serious implications and challenges that need to be addressed.
The below provides an outline of such challenges that decision-makers should be aware of when looking into mobile BI, as well as giving insight into ways in which these can be addressed to ensure the successful implementation of mobile BI.
Suitability:
Mobile applications, BI included, are not always applicable to all organisations - it depends greatly on the nature of the industry and the culture, as well as the "information maturity" of the organisation.
For example, a life insurer can be as technologically advanced as possible; however, with monthly premium payments and the odd once-off high-valued claim that has to be seriously investigated and approved before pay-out can be made, mobile BI hardly makes any business sense. Additionally, on the other hand, organisations in the fast-changing retail or telecommunications industries may seem ideal candidates for deploying mobile BI, however, if their approach to information management and their deployment of conventional BI are still in disarray, mobile BI would only serve to create more confusion and spread the chaos even further.
For mobile BI to be applicable, the organisation`s data integration, measure standardisation and reporting/dashboarding portal solutions must already be well bedded down. A centralised BI function obviously helps; however, its existence in itself does not mean that the organisation is ready for mobile deployment and distributed reaction on the deployed information. There must ideally be an executive interest and drive to mobilise the workforce and distribute the active exploitation of information.
Single view:
The confusions around "one version of the truth" become a lot more complex and far reaching with distributed information delivery and utilisation.
To ensure that each user is reading the same report, a single point of origin for all reports and publishing should be used; this will ensure that the remote BI is much more manageable. Collaborative BI analysis, which has been produced by various departments, will give the necessary enterprise insight, but it is important that all this data must flow through the single point if it is to be consumed correctly.
This is where the challenge of device independence also comes in - not only do users have their favourite smartphones and other handheld devices, but some users may also be switching regularly between these devices. For example, it is quite possible for a single executive to work on a PC in the office, check a few key performance indicators (KPIs) on his/her phone between meetings, investigate the days` activities on his/her iPad throughout the day, and then initiate a few actions with international operations from his/her laptop at home. Across all these platforms, he/she wants to run similar applications, which gives him/her exactly the same view on the same data, only updated in near real-time.
Therefore, standardisation and information governance, which should include master measure management, is crucial to have in place before embarking on mobile BI deployment in any organisation.
Display challenge:
Another challenge with mobile BI is to provide a satisfactory experience on the relatively small screen space, relatively immature mobile browsers, and with relatively slow data transmission speeds. With a more limited display capability, the amount of information dished up to the respective users should be carefully considered. The level of detail has to match the level of responsibility of the user. Especially on small smartphones, the amount of drilling down and drilling around should be minimal. The volume of data displayed must also be kept to a minimum to enable the user to immediately focus on what needs attention.
Although visually pleasing displays may be initially much more preferable, users will very quickly lose interest if they need to go through more steps to get to the information they need, or if it doesn`t provide the necessary value. Displaying information effectively on a big screen is a complex task - displaying it meaningfully on a much smaller screen is an order of magnitude more complex.
There is a big difference between merely deploying an existing BI solution to a mobile platform versus developing and deploying a mobile BI solution. The first will more than likely result in navigation difficulties, display challenges, and frustrated users. A mobile BI solution should display only the relevant data, at the appropriate level of summarisation, with suitable navigation options, using the facilities appropriate to the mobile platform.
Real-time workflow:
Mobile BI implies a much more real-time aspect than often initially perceived. It does not help to provide decision-makers with information wherever they are, but the information is so old, that they cannot act to influence it.
Of practical importance is "active reporting" technology that combines data and interactive capabilities into a single HTML. This is the foundation for "active dashboards" that pack a lot of information into a very small space. An associated payload of data enables users to display charts and tables using the zoom function of the browser.
Users can perform analysis and forward their findings without any network or device constraints.
In the retail world, for example, if check-out queues are running long, or if a certain product line is suddenly selling out fast, the deployment of more cashiers and the re-ordering and despatching of more stock must happen almost instantaneously. Should this not happen, the window for corrective action would have been missed, and the result would be disgruntled or unhappy customers - which should be avoided at all times.
Together with the real-time aspect comes the requirement to be able to initiate workflow action from the mobile device while keeping and communicating the context that triggered the action.
Platform compatibility:
For mobile BI to be useful within an organisation, it must run on a wide range of technologies, including a wide variety of smartphones, iPads and tablets. The more technologies to be supported, the more complex it gets - that is what makes it complex for organisations who want to deploy BI information to a mobile workforce.
The organisation must either standardise on a small number of technologies and have users that do not consider their "work phone" as their primary communication device because they prefer some other brand device, or alternatively, they have to be prepared to support a large number of very different devices with different interfaces and protocols. The most practical is likely some approach in between, where a limited number of popular devices are supported. The approach taken will also depend on the facilities provided by the mobile BI tool in use.
Security:
Security is major concern for most mobile BI adopters. Contrary to mobile data capturing and maintenance applications, mobile BI does not usually go down to the level of detail where an individual`s privacy would be compromised. However, this involves displaying company-confidential information, which may be very valuable in the wrong hands (be it the public press or industry opposition.) In addition, the information is now taken off the corporate premises on a very small and "lose-able" device. It is so easy to forget a mobile phone on a counter, in the car, or even have it stolen.
As a result, security has to be controlled on the device, during transmission and for correct usage. On the device, most manufacturers provide disk and e-mail content encryption, antivirus and firewall software and some provide the capability to totally erase the contents, which is very useful if the device is lost or stolen.
Obviously with mobile devices, transmission security is very important to protect the communicated data from interception, analysis and deception. This is done through secure socket layer and virtual private network (VPN) services, where the identities of the sender and receiver(s) need to be verified along with the encryption facilities. This needs to be combined with authorisation, which specifies what data the identified user may have access to, and what he/she may and may not do with that data. Like with conventional BI applications, the big challenge is keeping the user from copying the data and incorrectly modifying it using other facilities - the best approach is not to allow the information to leave the protection of the BI application, however, this means the BI application must have all the necessary functionality.
Architecture choices:
The architecture of the mobile BI application can have significant implications on the ease of use, breadth of adoption, deployment times, maintenance effort, and total cost of ownership. This is highly influenced by the architecture utilised by the mobile BI tool vendor.
Thick-client applications that run natively on the mobile device are typically feature rich, but they are device-specific. These native mobile applications have excellent performance as the data is downloaded and stored on the device. The applications can be used even when not connected. Exploiting the functions inherent in the device, such as touch screens and scroll functions, makes the applications more responsive and interactive.
Thin-client browser-based mobile applications are much easier to standardise, deploy and maintain across mobile platforms, and even seamlessly on laptops and PCs. The data is not copied to the device - this implies that the users always access the same up-to-date data. Security controls are also centralised on the server, which makes access control more secure. However, the functionality may be a bit more limited and interaction speeds may be a bit slower.
Hybrid solutions run components on the device, which standardise the access to corporate data, regardless of the device`s capabilities.
It is obviously easier if the mobile BI solution ties in closely with the organisation`s existing reporting and BI solution, from a shared functionality, cost and licensing point of view, however, great care must be taken that the challenges described above are handled adequately by the vendor`s solution.
Flexibility:
As is true in any BI environment, there will be change. Business priorities, burning issues and market opportunities will most definitely change during the deployment timeframe of a mobile BI solution, as well as in the course of ongoing management. The architecture and deployment approach must cater for these changes, preferably in a manner that allows for quick and agile change.
Taking the above into consideration, mobile BI is not just about providing access to data irrespective of place and time. It has its own approach and agile-like methodology; a design and strategy to make sure that critical data is available in a readable format in a real-time and secured manner. To be successful, mobile BI requires meticulous planning in the choice of mobile network, the mobile BI tool and its overall deployment.
For an organisation to be fully agile it requires some form of mobile BI to enable faster executive decision making, quicker reaction to business or market changes (which are constant), better customer service and increased productivity due to streamlined business processes. If a company is already correctly utilising BI to provide business agility and the ability to make educated and dynamic business decisions, which are based on real-time data, then mobile BI is likely to already be at the forefront of their technology strategy. However, these businesses need to make sure that the challenges highlighted are addressed adequately for successful deployment.
Share