Self-service BI part 5
This is part 5 of a 6-part series on self-service BI. The series will address primary dimensions of self-service BI:
- Part 1: The Night is Dark and Full of Terrors: Gear Up with Self-Service BI
- Part 2: The Hand of the King to Support Self-Service BI: BI Competency Centers
- Part 3: The Many Faces (and Use Cases) of Self-Service BI
- Part 4: Breaker of Chains: How IT Needs to Support Self-Service BI
- Part 5: Unburnt: Data Strategies to Support Self-Service BI
- Part 6: Winter is Coming: What to Do Next
Dude. You need a data strategy for self-service BI. The unburnt is one who can stand in the fire and some out unscathed—this *will not* happen without a new, flexible, scalable data strategy. This self-service BI data strategy should help identify key principles / tasks:
- Assist in finding the right sources.
- Manage metadata for ease of consumption.
- Perform routine cleansing and de-duplication.
- Promote to enterprise sources, with views.
- Move / Copy data as little as possible.
- Make sure archived and deleted information is excluded.
Of course, your data strategy should be shared with the Information Governance team. That team needs to understand how data not only supports enterprise transactional data, but also analytic data.
What do we mean by Information Governance? For a full answer on the model below, check out this blog series.
In short, Information Governance is a discipline. Technology definitely supports all of the key activities (that’s why the outer ring encompasses the whole circle), but the majority of the work is not a technology problem. Much like developing an analytics strategy and self-service BI strategy, and information governance strategy includes defining processes and metrics for your data, understanding who can create / read / update / delete the data, deciding which data is fit for which level of governance…the list goes on and on.
Self-service BI will add new requirements to your data strategy. And with increased data visibility, you need a solid data strategy to remain unburnt.
- Which data sets will be promoted to enterprise sources?
- Where does cleaning and transforming take place?
- Where do calculations take place?
- Is there a feedback process in place for dealing with found errors and inconsistencies in the data?
- How will you move the data to avoid creating loads of data silos?
- Are your user permissions for sensitive or out-of-area data set appropriately?
- How will data get promoted to enterprise sources?
- How will you share common reference data and enrichment sources?
- How will you manage third party data sets, like Dun and Bradstreet numbers or demographic data?
A key dimension is the data preparation that is required for analytics activities for some companies, as much as 80% of their analytics effort is spent in preparing the data (see the TDWI report linked above)! Make sure there is a thought-out strategy for data preparation for self-service BI.
Are you best friends with the Information Governance team? Have you created a holistic data preparation strategy?
In the next section, we’ll talk about how to combat winter by starting on your self-service BI journey.