Operational Data Warehousing in 25 Tweets

Topic: Operational Data Warehousing

To help educate people about what operational data warehousing (OpDW) is and why they should care, I recently issued a series of 25 tweets via Twitter over a two-week period. Every tweet is a short sound bite drawn from TDWI's recent report on OpDW, which I researched and wrote. Most of the tweets focused on a statistic cited in the report; I like to call these "stat bites." The other tweets are definitions stated in the report.

As a change of pace for TDWI Experts in BI, I'd like to share these tweets with you. I think you'll find them interesting because they are a compact -- yet amazingly comprehensive -- study of OpDW, albeit in Twitter style.

I was tempted to edit the tweets, to spell out arcane acronyms and abbreviations, but I left these alone because I think that almost everyone in TDWI's audience can figure them out. Even so, I cut out hash tags, tiny URLs, and repetitive phrases. I issued the tweets in groups, on related topics. I've added some headings here to show that organization. Otherwise, these are raw tweets.

 
What is Operational Data Warehousing?


1. Operational Data Warehousing integrates apps for biz operations & biz intelligence, often in real time.

2. Operational Data Warehousing enables time-sensitive biz practices like OpBI, JIT inventory, real time.

3. 51% of survey respondents know what Operational Data Warehousing (OpDW) is but don't have name for it.

4. 43% of survey respondents know what Operational Data Warehousing (OpDW) is but call it by another name.

5. Some respondents call OpDW an ODS (19%), OpBI (15%), Real-Time DW (12%), or Active DW (6%).

 
Who's using OpDW today?


6. 66% of organizations surveyed practice a form of OpDW today, but only 17% run in real time.

7. Only 3% of surveyed folks said "don't know" if their org does OpDW. This shows high awareness of OpDW.

8. 56% of organizations surveyed are committed to Operational Data Warehousing (OpDW), deployed or in dev.

9. 29% of orgs surveyed are considering OpDW. Only 15% have no plans for it. Hence OpDW usage will grow.


What are the common barriers and benefits for OpDW?


10. Top 3 Barriers to OpDW: cost (44%), lack of sponsorship (31%), and a DW that can't do real time (28%).

11. 46% of survey respondents selected "business decisions and strategies" as top benefit of OpDW.

12. 39% chose "business performance and execution" as the second place benefit of OpDW.

13. In third place among benefits of OpDW, 35% surveyed chose "data freshness or timeliness."

14. 44% of survey respondents chose "cost" as top barrier to Operational Data Warehousing (OpDW).

15. 31% chose "lack of business sponsorship" as the second place barrier to OpDW.

16. In 3rd place among barriers to OpDW, 28% surveyed chose "current DW can't handle data in real time."

 
What are the common best practices and use cases for OpDW?


17. Real Time Operational Data Warehousing (OpDW) cycle has 4 steps: Recognize event, Capture info about it, Analyze, React.

18. Op apps integrated via OpDW: financials (38%), CRM (36%), call center (31%), ERP (26%), HR (21%), etc.

19. Analytic apps integrated OpDW: OpRptg (55%), OpBI (43%), active DW (37%), customer analysis (33%), etc.

20. Today 86% of data values in a data warehouse are updated daily or less frequently (week, month, year).

21. Only 14% of data values are updated every few hours or sooner. Real-time data is rare, but critical.

22. EDW queries execute in seconds (34%), minutes (38%), hours (12%), etc. Potential for RT query is there.

23. Most EDWs are designed & optimized for latent reports & OLAP, not real-time operational data warehousing (RT OpDW).

24. 58% of EDWs support only rpts & OLAP. 35% support those plus analytics, real time, & other workloads.

25. Where to process RT OpDW workload? 24% in EDW. 22% outside EDW. 23% both. 31% don't know.

 

Practical training, executive buy-in key to BI user Adoption

During any business intelligence (BI) deployment, organizations focus on the technical details – for example, determining the number of servers needed to support the system, choosing the data sources to be integrated, and deciding how often to refresh the data warehouse.

But in a BI deployment aimed at non-power users, addressing cultural barriers to ensure that the system is actually adopted throughout the enterprise is often just as important as mastering the technical specifications. Giving short shrift to these non-technical concerns could result in a BI manager's worst nightmare – a successfully installed, complex and expensive BI system that no one uses.

While power users, like business analysts, need little incentive to adopt BI technology, more casual BI users, such as marketing and sales associates and line-of-business managers, are a different story.

For adoption to truly take off among these users, it is important for IT and BI staff to educate them on how the technology will help them achieve personal and departmental business goals. Less important to these users are the BI system's various eye-catching features that might impress a more sophisticated user.

If a salesperson has a goal to find a certain number of new customers in a quarter, for example, the BI staff should show him or her how the new report or dashboard will help achieve that goal, rather than giving a lecture on how many different views of the data are available.

The best way to sell corporate executives, and ultimately casual users, on BI is "by realworld war-gaming and simulations, which can provoke seminal 'a-ha' moments".

"[That is] the best and maybe the only way to demonstrate the merits of real business intelligence."

At the Canadian newspaper, executives were having trouble understanding circulation reports. The data just "didn't mean anything to them,". Recognizing the importance of executive buy-in to maintain BI funding and to encourage wider adoption of the technology.

But whether it's focusing on practical training, facilitating communication, or getting executives to lead by example, the IT and BI staffs to remember that promoting end-user adoption of BI systems is an ongoing process, and not an easy one at that.

Remember: "It's not going to be done in one day or two days”, "It's hard."