Many new features are added in every RED and 3D release; it is hard to catch up with fast-paced WhereScape software. This article aims at shedding light on those new features so our customers and prospects have a comprehensible picture as to why they would want to use them.
A modern information system serves as an integrated set of components for collecting, storing, and processing data and delivering knowledge. Building an information system – whether it is operational store, data warehouse, or accessing layer et al – has always been a difficult task. WhereScape – as productivity tool – provides a holistic answer for information systems build consisting of Automation Solution, Iterative (Agile) Solution, and Integrated Solution.
With WhereScape RED’s Tableau Integration Suite, you can automatically find the impact to your Tableau workbooks of any Data Warehouse change.You can also create Tableau extracts by drag and drop, send Tableau objects to a Tableau Server and refresh any objects on the Tableau Server as part of you data warehouse schedule.
With the economy decelerating all over, the needs for business intelligence swell; decision-making becomes crucial. It is not about making mega-bucks any longer but rather about survival of the entire business.
Nobody is questioning the vital necessity for a successful business to constantly monitor a company’s performance, its profitability, its best and worst customers, its best and worst products and its overall health especially now, when resources are so stretched. The driving force behind such business intelligence functions is a healthy data warehouse.
Nowadays the question is this: “How to achieve the same – and more for less?” How to build such a brawny data warehouse with a constant shortage of money?
Most of the organizations still start with a data warehousing project, using the ol’ traditional waterfall approach. It worked before but does not cut it any longer …
Groundwork for such a project includes taught questions with no good answers:
• Business users are asked what data they need. But the business generally doesn’t know what it wants. “I’ll tell you what I want when I see it” is their familiar mantra.
• Then business users are asked what reports they use and what is missing in those reports. But reports are “moving targets” they change swiftly, especially when decisions are about the life and death of the company.
• Where to find right people? Well, with a continual shortage of the skilled ETL developers, data- modelers and business analysts – good luck. A decent IT chap costs a fortune.
• How long will the data warehousing project take? Organizations can’t afford the common 6 to 9 months, possibly, longer with painful approvals and rework.
• How do we keep documentation up to date? What is the risk of such lengthy warehouse project? How to find an enthusiastic sponsor of the business intelligence effort?
• And many more questions…. Oh stress, the spice of life!
That is the alternative? Not to life it is – to stress!
A different approach that presents a true unique business value for a stressed out business that includes:
• Built-in methodology that supports the entire data warehouse lifecycle an Iterative approach. The Ability to stay focused on the goals.
• Spiral-like rapid “prototyp–>iterate–>deploy” venture. Quick response to changing business needs. The only questions asked: how to make the right business decisions right and what are the measures of the truth?
• Integrated, metadata-driven data warehouse development environment. All that the developer needs is on his/her fingertips, Auto-generation of procedural code and documentation.
• A low risk, proven, pragmatic approach to data warehousing. An instrument to get business users and stakeholders involved earlier.
• Effective use of time & money; quicker ROI. A process for trimming off 2-4 months a year from the development and support costs; reducing time to deliver.
• The data warehousing as a process – not a project. Well-timed accommodation to changing business needs.
• Leveraging existing, readily available SQL skills in market. Theory and best practices are in the tool, not in developer’s head.
Please find below some of the deepest pitfalls that you might want to be aware of. In addition, see how WhereScape RED can help you avoid such drawbacks.
1. Insufficient User’s Involvement
· IT departments sometimes try to run Data Warehouse/Business Intelligence projects in a vacuum. DW/BI projects are not IT projects! They are rather initiated by and addressed to Business Users.
· Lack of user involvement typically leads to either solutions that are too hard to use or constructs that do not solve any business problems.
The RED Tips
· Get Business users involved as early as possible
· Work with users to help them understand how to formulate the “right” questions. They need to be able to express the precise needs of the business to the IT development team. Create a synergy between IT developers and Business users.
· Show users frequent updates to make sure you are on the path to solving an identified problem and keep the momentum going.
· Develop guidelines and best practice principles based on successful prior iterations.
· Establish a training plan to secure consistency in DW development, tool usage and the organization’s standards.
2. Enterprise Warehouse at Start
· Some companies attempt to begin by building an enterprise warehouse. It is hard to bite, chew and swallow a very risky endeavor. Adding more team-members does not help: Data Warehouse projects do not scale well.
· Enterprise warehouses have slow ROI, cost a lot of money and, therefore, usually end in disaster.
· Painful lessons learned are more difficult to implement.
The RED Tips
· Start by building a Data Mart – the methodology is identical but scope is far more manageable. Data Warehouses need to grow gradually, analysis area by analysis area, building towards the full enterprise Data Warehouse.
· Make use of rapid prototyping and speedy development cycles; they give you swift success by
– solving a real business problem quickly;
– wining a powerful sponsor of the DW/BI project;
– helping secure financial support for future projects;
– gaining experience and outcome at the prior iteration (helps to justify the subsequent ones);
– establishing a single point of the organization’s business rules and DW and OLAP standards;
· Be aware that the initial iteration should take no longer than three months (hopefully much shorter).
3. Poor Data Quality
· The Data Warehouse, perhaps, is the best way to discover (and report) “noisy” data.
· Most “bad” data is easy to spot but some can slither into the Data Warehouse.
· Fixing data in the Data Warehouse does nothing to stop new bad data coming from the source.
The RED Tips
· Identify the resources that contribute the most into the Data Warehouse and verify its accuracy.
· Use the Data Warehouse methodology to drive repairs to source systems to prevent bad input.
· Do not fix data; reject the whole “chunk” or record if invalid data is found.
· Create “unknown” buckets in the Data Warehouse so data issues get visibility.
· Assign invalid data to a zero dimensions key; auto-add dimension entries to ensure consistency.
· Re-process the failed rows once the source issues are resolved.
4. Deficient Funding
· Even when starting with a small Data Mart, money is always an issue.
· Warehousing projects often fail to subsidize all aspects, including hardware and training.
· Data cleanup always takes longer than expected and will dissolve a significant part of the budget.
The RED Tips
· Obtain high-level sponsorship from an area of the business.
· Prepare in advance for adequate hardware and training in addition to the licensing, data and report work.
· Make sure to schedule delivery of reliable information to the decision-makers.
· Validate the funding from non-technical points of view:
– Real business advantage for the organization is pragmatic, iterative Data Warehouse building;
– Lower risk by effective use of time, money and staff;
– Implementation of red-hot latest DW/BI methodology (Kimball, Inmon, Agile development, etc).
5. Corporate Jungle
· Power struggles in the organization may occur over the resources involved in the project or the vision and focus of the warehouse, to name a few.
· Certain people may feel threatened by the outcome of warehousing projects.
· Some folks claim they own certain data and no one else is going to use it.
The RED Tips
· You only have two possibilities here: play more politics or play less politics.
· Have the organization support of a clear vision on how to solve a specific business problem.
· Very vigilantly consider opinions and “pain points” from various business groups; hang on their words.
· Carefully choose your sponsors; this should be the most influential decision-maker that you can recruit.
· Have your business top dog cut through the maze and drive the project forward.
· Provide fast offline query performance capability for ad-hoc analysis by business community.
6. Incorrect or Partial KPIs
· Even simple cubes may not contain valuable information. E.g., a cube that contains sales information is useless when trying to analyze sales reps’ performance.
· Warehouses that are built by IT often include just the existing data and not the KPIs needed.
· A lack of proper KPIs means that the warehouse falls short of solving a business need.
The RED Tips
· Work with users and business patrons to solve specific business problems; one problem at a time.
· Remember that KPIs must be focused on solving an explicit problem such as trends, distributions, and empowerments.
· Adjust the display of KPIs thoroughly to show the organization’s competitive advantage.
· Understand what metrics are used to calculate bonuses for upper management.
· Consider delivering the KPIs through a dashboard or scorecard.
7. Exceedingly Complex Cubes
· Cube complexity is a major obstacle to the success of the Data Warehouse.
· The IT department tends to build extremely complex cubes; OLAP’s with many dimensions are very difficult for end users to understand, much less use.
The RED Tips
· Keep in mind, that there are two very different types of cubes: Analytical and Reporting.
· Use as few dimensions in an Analytical cube as possible; rule of thumb – maximum of six to eight dimensions.
· Take advantage of rich OLAP functionality, provided by the DW tool such as hierarchies, calculated members and so on.
· Build multiple, simple cubes and link them when necessary using the Virtual cube concept. Alternatively, create simplified views of complex cubes.
8. Limited or Unreachable Access to Information
· Many companies can build a Data Warehouse but fail to provide users with a simple way to access the data. For example, many companies use Excel as the standard client tool but, in fact, it is one of the weakest tools available.
· Different users need different ways to interact with the data.
The RED Tips
· Always remember do not just get the data in but also get the information out; have the right tools for the right users to look at the data:
– scorecards, dashboards for decision-makers and management;
– powerful analytical tools for business analysts and SME’s;
– reporting with limited analytical functionality for operation workers;
· Design the Data Warehouse independently from end user tools.
· Expect a range of functionality from the Data Warehouse, which will mean a variety of front-end tools.
Summing up …
· Data Warehouse / Business Intelligence efforts do not succeed because of a variety of reasons, many of which are non-technical.
· What does WhereScape RED put forward for your organization:
– Offers an integrated Data Warehouse development environment.
– Promotes a built in methodology that supports the entire Data Warehouse lifecycle.
– Enables rapid prototype-iterations and deployment of the project
– Supports a low risk, proven, pragmatic approach to Data Warehousing.
– Treats Data Warehousing as a process – not a project.
– Allows for the effective use of time, money and resources.
– Shaves off two to four months per year from your development and support costs.
– Leverages existing, readily available SQL skills in market and collaborates.
– Complies with the Data Warehouse Lifecycle Management principles.