Being part of a leading software company offers many advantages—and sometimes disadvantages. A key element I particularly appreciate in my work is that I can be part of large – or even huge – scaled analytics journeys, comprising customers and BI competence centers that serve thousands of users. And in today’s digital economy, they all struggle with similar challenges.
Let’s focus on the business users and their biggest requirements for analytics, and how this often brings us into the self-service dilemma.
Enterprise end users require, at minimum, the following features:
- Self-service capabilities: Business users require great autonomy in their analytics work. They want to easily create, deploy, and share their business analytics content themselves, without being too reliant on their ICT or BI competence centers. The data analysts among them even require to access non-corporate data to blend with corporate data in the search for new insights.
- Agility and flexibility: Agility has become an almost magical word, and I hear every user talking about it. Users nowadays require full flexibility when using analytics. This means easy access on any device and the ability to change graph types on the fly. It also means being able to swap measures and attributes at any place in the analytics dashboard, storyboard, or report. Users also require drill-anywhere capabilities, and the ability to drill to the transactional level if applicable. Agility requirements are determined by business decision-makers, who must fully understand the tools needed to manage processes, market fluctuations, and customer needs.
- Online, real-time information: As you might expect, all the users I have met want data to be accessible in real time, and ideally, online. I understand that need; driven by agility, users absolutely must have the latest data to respond to any fluctuation in process or market.
- Consistent metrics and metadata: This one’s a no-brainer: Virtually all users report having had negative experiences with consistency in metrics and metadata. In any type of business analytics applications (reports, storyboard, workspaces,or dashboards), users expect consistent metrics, the use of definitions, hierarchies, prompts variables, and other metadata-related content. End of story!
- Good governance: Oh, yes—end users are concerned about governance. Though users may seem to want unfettered access to everything, deep down they understand that authorization processes and security are critical and must be treated with ultimate care. Key here is SSO (single sign-on): Would you like to log on and enter your credentials 75 times per day? Nah, didn’t think so. So SSO is a must-have.
- Visual appeal: Basically, this refers to the user experience. Since analytics are widely shared – often extending to your customers’ customers – they must be visually appealing to attract attention. The concept of visually appealing analytics is more complex than you might think. Visualizations need to have the creativity, effect, and structure to communicate the precise message that is intended.
There are more players involved as well: ICT or a BI competence center need to maintain the analytics platform, ensure standards and guidelines, and assist in ad hoc analytics. Also, large enterprises have huge amounts of automated analytics, such as reports or dashboards that are essentially “fixed” and need to be distributed at regular intervals, which require managed reporting and dashboarding. Finally, all this consistency and governance that end users require must be deployed somewhere.
The self-service dilemma
So here we are: The large enterprise is using business intelligence toolsets and users are looking for self-service and agility. Typically, this is when the self-service dilemma starts. Users, architects, and IT leaders are all very well-informed these days, and many consider self-service BI the ultimate tool for every end user. And they have a point: End users get full flexibility and self-service capabilities with a very low learning curve. It offers powerful visualization capabilities and users can easily blend their data with external data.
However, they are forgetting about managed dashboarding applications for enterprise-wide analyses. Managed dashboards are centrally created dashboards on corporate data that are managed and distributed – with good governance – to end users.
In many cases, managed dashboarding can cover all end user needs in a remarkably powerful way. Self-service BI is really best suited for data analysts. The question then becomes a matter of choosing the best BI component to address any self-service dilemma an enterprise might have. Here’s how to do that.
Self-service corporate data
A key question: Do your users need to access only corporate data, or do they need to blend corporate data with outside data sources? If users need to analyze only corporate data, it is important for them to have real-time, online access to transactional-level data. In such situations, managed dashboarding is a perfect solution.
Users who need access to external data, or data that is not part of the enterprise corporate system, require self-service BI on top of managed dashboarding. Often these users include data analysts.
For example, imagine a dashboard application that provides analytical insights on purchasing information. The dashboard covers every purchase order detail (a common misperception is that dashboards cannot cover detailed information—this is wrong; they can do this perfectly) and provides insights in spend analysis, vendor optimization, and vendor fraud detection.
A user can slice and dice any data type, drill into details, and gather KPI information, including trends and thresholds. Now imagine that user wants to focus on a specific vendor and needs to combine the dashboard data with data from, say, Dun & Bradstreet, to verify the vendor’s financial credibility. This data exists in an external system, which the user has access to, but it is not part of the managed dashboard. This where our user turns to self-service BI.
Managed dashboards for corporate data
For users who work only with corporate data—in other words, only with data objects provided through the application—managed dashboarding tools are very well suited.
For more on the power of data analytics in the enterprise, see Big Data In Practice—What Can We Learn From Successful Organizations?