You need to consider the following seven key mobile BI steps to determine how best to structure and manage your data for multi-tiered, hybrid computing. They are based on data consumption and data governance (i.e., the rules around how data is used in a company) best practices that I describe in my article, “A Primer on Mobile BI Data Governance.”

Tip 1. Create Ownership of Your Data

Determine if data management is centrally controlled or distributed. This debate is more philosophical than practical, but it sets the underlying tone as a guiding principle throughout the organization. Data rights are important and can often vary from one department or business unit to another. For example, make sure your supply chain team has visibility into native manufacturing orders and purchasing requests.

Another approach might require that information be managed by manufacturing inside SAP ERP, for example, and summarized into BI cubes and Mobile Business Objects (MBO) data sets for the supply chain group to view and analyze. Do your internal IT teams manage those cubes and perform routine data performance tests on the operational data, or is that the responsibility of a regional business unit or operating function? These situations and questions set the tone for how business is done and who conducts the work to govern the data, as well as how the data is available to be consumed in cloud and mobile apps.

Tip 2. Structure Your Data

If you have multiple platforms of data that may be difficult to summarize and parse, there needs to be a strategy to bring the data together, if not physically, then logically. While data nirvana may be an expensive pipe dream to organizations, a clear look at how data summarizes and parses is critical to mobile BI performance.

For example, if I need to consume purchasing history and customer location information in a mobile sales and delivery app, those data sets need to be available in MBO or HANA so that users can quickly bring up the information, review it, and make real-time decisions as to where to spend their day in the field.  If purchasing and customer data exist in separate systems, those information sets need to be brought together and summarized. In an MBO environment, some data histories may not be needed by the mobile sales app, and these data histories need to be parsed to provide the best app response. Make sure you can deliver the data when and how it’s needed for users. Taking the time to understand the inner workings of this data pays off in the long run.

Tip 3. Eliminate Zombie Apps

Often there is unused or obsolete data that can be committed to offline storage and repositories. This step can reduce your operating costs to maintain your data storage environments. In addition, be wary of uncontrolled pockets of information — what I call rogue or zombie apps. This problem is typical of organizations trying new apps without proper data and application governance throughout the enterprise. Users or departments try an app, decide it isn’t for them because of function or support, and stop using the app. For example, as a remnant of an old bring-your-own-device policy, employees might have downloaded a digital music app, let the app consume data in a proof of concept, and then forgotten about the app.

But those programs don’t just go away. Instead, they can continue to use information, driving costs up and creating data management headaches. Just get rid of them and transition to a governed data and application model. Several third-party system management tools can detect application and data patterns in network landscapes and isolate zombie apps, which makes it easier for companies to remove them.

Tip 4. Remove Fiefdoms and Kingdoms of Data

Even if companies have good data governance practices by business unit, a distributed approach can lead to data fiefdoms and kingdoms. In this situation, directors and vice presidents typically have such a distrust of operating IT teams and structures that they insist on handling their own data governance issues themselves. While this may work for a time in a classic on-premises environment, the model breaks down in a mobile and cloud computing environment because data needs to be aggregated, summarized, and parsed for mobile consumption.

If the executive king cannot be convinced to behave like a good corporate citizen, at least summarize the data into BI cubes that are well-structured and well-managed so other groups can leverage the information in their enterprise mobile and cloud apps. Data can be replicated as a work-around, but the best way to address this head on is to not include the data until the data owner abides by a “data bill of rights” or some other data governance policy that applies to the department or function.  By doing so, decisions can be made without an operational vacuum.

Tip 5. Align Corporate and Individual Performance Management Systems

This advice has less to do with data management but everything to do about how data governance works in practice. The old adage “people behave how they are paid” applies in most companies. If data governance is important to your organization, make sure that corporate performance objectives call that out and that individual performance management systems support team behavior and good citizenship accordingly.

For example, if a mobile initiative is important company-wide, then management should have as a company goal the successful launch and operation of the mobile initiative.  Once management team members have a stake in an individual performance plan, their behaviors support corporate goals and objectives.

Tip 6. Ensure Your HR Policies Allow You to Use Data in Cloud and Mobile Apps

Although this advice is obvious, there has been more than one head slap in corporate HR circles because of IT groups getting ahead of HR policies, particularly in bring-your-own-device practices. Even though policy-wise HR trumps IT here, often HR is so disenfranchised that the department doesn’t know what’s going on in the IT area.

Several years ago, it was common practice to include in corporate policies a ban on social media and mobile computing for data use and storage. Now mobile and cloud apps often demand access to these very same environments. Not only does this dilemma make it look like your HR and IT departments don’t communicate well with each other, it also, in extreme cases, may be cause for employee termination even though the mobile app you provided is approved by management. In legal situations, the employee handbook, no matter how outdated, is the core source of policies and procedures determining employee discipline. This situation can be messy, so it’s best to avoid it with a well-thought-out and well-documented approach for how employees can use data on mobile apps.

Tip 7. Cleanse Your Data Regularly

This step is perhaps the least glamorous on the list, but nevertheless it’s important. According to a frequently used metric in data governance circles, poor data quality accounts for 40 percent of the reasons business initiatives fail or achieve poor results. Your team can take care of the first six tips I address and still not reach its goals with business users if they don’t follow this last tip. So take the time to maintain your data and tune it up, just like apps.

William Newman

William Newman, MBA, CMC is managing principal of Newport Consulting Group, LLC, an SAP partner focused on EPM and GRC solutions. He has over 25 years of experience in the development and management of strategy, process, and technology solutions spanning Fortune 1000, public-sector, midsized and not-for-profit organizations. He is a Certified Management Consultant (CMC) since 1995, qualified trainer by the American Society of Quality (ASQ) since 2000, and a trained Social Fingerprint consultant in social accountability since 2012. William is a recognized ASUG BusinessObjects influencer and a member of SAP’s Influencer Relations program. He holds a BS degree in aerospace engineering from the Henry Samueli School of Engineering and Applied Science at UCLA and an MBA in management and international business from the Conrad L. Hilton School of Management at Loyola Marymount University. He is a member of the adjunct faculty at both Northwood University and the University of Oregon with a focus on management studies and sustainability, respectively.

If you have comments about this article or BI Expert, or would like to submit an article idea, please contact the editor.

You may contact the author at wnewman@newportconsgroup.com.

If you have comments about this article or publication, or would like to submit an article idea, please contact the editor.