BI – Will you grow up already?
BI – Thoughts (7) – Cadran publishes a series of articles about the ideas surrounding Business Intelligence in combination with Oracle JD Edwards ERP. In these articles various considerations and reflections are discussed, which can be helpful in making the right decisions in the implementation and application of both systems. The previous articles dealt with the star scheme, facts, dimensions and the interplay between these two. This article discusses the development cycle.
Maturity of the organization
Companies, regardless of their size, go through a growth path that is comparable to the human growing up. From birth to seniority, a number of stages are completed. Of course, this happens with the operation of the business processes and with the management of the company. Automation is not always an aim, especially for smaller companies, but it is a very important tool. Processes are captured in procedures and automation can greatly contribute to their efficiency and effectiveness. Looking at Oracle Business Intelligence we can see that added value can be achieved when it comes to proactive management and signaling in the form of management by exception. But Rome was not built in one day. It is an illusion to implement Oracle Business Intelligence in a short time and to use it as a powerful all-encompassing resource across the full breadth of the company. This simply requires time. An organization will have to allow itself this time. The implementation and acceptance of this also takes time again. Time, which is scarce in daily operations. In addition, such a solution will initially require work and time instead of reducing it. This is because the deployment of a good solution will irrevocably reveal problems that relate to the quality of data and the optimization of processes. This will have to be anticipated and that is in itself also an effort that takes time.
Small circles / Baby steps
Oracle Business Intelligence is capable of providing the organization with guidance information across the entire breadth and depth on an operational, tactical and strategic level. If this is offered as a big-bang in the organization, there is a risk of short-circuiting. There is a surplus of management information, for which organizational and administrative support must exist in order to be able to respond. The key word in this is small development circles.
This is independent of the chosen solutions. It may be that Oracle Business Intelligence Applications for JD Edwards Finance has been purchased and is being implemented. This immediately contradicts the above, because an oil tanker with a shipload of information becomes available. However, it is also possible to start small here, by focusing on a small sub-area and within that a small sub-topic.
As an illustration of this, within Finance you could zoom in on Accounts Receivables and analyze Days Outstanding. By only using this piece of information in the organization, the application is introduced on a very small scale. The acceptance is therefore low and can be the beginning of a slowly expanding whole. We can go live.
A Proof of Concept is also a good start if it is not set up too broadly. A example of this is an insight that speaks to the direct imagination, namely margin reporting. Within this context, a small analysis can start with an insight into the turnover and profit per period per business unit. We can go live.
The secret is in small development circles. A simple sub-area, as proposed above, must be accepted as small as possible and taken into production. It will then provide insight into actual production data and offer added value. As soon as this plays a functional role in the operation, this can be repeated again and again with a small extension.
From here, the reports, analyzes and dashboards can broaden and deepen and the amount of subject areas can slowly increase. Each time it applies that this should be done in small circles. Adding a subject area, such as Purchasing, can start with a simple report of eg Open Purchase Orders per Status. As soon as this ‘report’ is available, it can be taken into production. From there, the broadening and broadening of this reporting area can take place.
In this way, results are quickly visible and acceptance in the organization can be achieved quickly. In addition, the product will go live quickly and prove the added value in the organization. An additional advantage is that when users see their usefulness and convenience, they will be more inclined to want to do more and more with the application of Oracle BI, as a result of which the product will increase more quickly in adulthood.
The image above shows in Oracle Business Intelligence the steps in the majurity path. Where the application is primarily used to replace existing historical operational reports, it gradually grows into an increasingly tactical and strategic tool with predictive and common qualities.
Maturity of ICT
The implementation of Oracle JD Edwards as an ERP system is a challenge in itself. Studies have been carried out, which show that an ERP system in an organization takes about three years (after go-live) to become an adult tool in the organization. The use of a Business Intelligence solution usually has a similar development time.
When an organization has been working with Oracle JD Edwards for some time, many problems, security issues and data problems have already come to light and have been dealt with. This is a very big advantage when Oracle Business Intelligence is implemented. It does not make much difference whether it involves Oracle Business Intelligence Applications or Oracle Business Intelligence Enterprise Edition (being the foundation and toolset for a green field approach or an accelerate solution).
Increasingly, implementations of Oracle Business Intelligence and Oracle JD Edwards take place in parallel. This is a challenge, but not an impossibility, as long as there is close coordination between both projects. The definitions in Oracle BI are very dependent on the implementation issues that are answered within Oracle JD Edwards. The challenging elements are further:
- The complexity of building a bridge between two shifting banks
- The lack of representative data
- The lack of historical data
- The lack of all conceivable situations, facts and dimensions
- Failure to test well and thus accept the Business Intelligence solution
It is therefore advisable to implement an Oracle BI solution only when there is data. This data only arises when Oracle JD Edwards (longer time) is in production.
With a go-live from Oracle JD Edwards, data is usually converted from the older systems and data sources. This applies to master data and current transactions and items, but it is often decided not to convert historical data. In combination with an implementation of Oracle BI, this could possibly be presented in a different light, as one immediately wishes for historical comparison possibilities. If this does not happen, it will have to be accepted that such insight into time course and trends will only offer added value with time. Here too it is possible to start the circle small and to grow slowly. If one is live for a number of days, comparison with previous days could soon be possible. After a quarter, comparison at a monthly level will play a role. And so this is also slowly expanding.
Testing and accepting – An eternal cycle
As Oracle BI operates on a representative set of company data, the application will become alive faster. The side effect is that it will provide information about the quality of the data. The quality of the master data, as well as that of the transactional data, as well as the integrity and consistency of the data, quickly and unmistakably comes to light. Oracle BI will therefore not only improve the business processes, but also the administrative processes. This too is an endless cycle, which needs to be kept sharp. In the automation it is suggested that an Oracle BI solution should be subjected to a complete user-acceptance test at least once a year, in order to reveal data problems.
When testing and accepting Oracle BI, the following situations should always be considered:
- Is every fact good when it is requested unlimited and unfiltered and without dimensions?
- Is every dimension / attribute good when it is requested unlimited and unfiltered and without any facts?
- Is every fact good when it is requested against all possible usable dimensions and attributes?
The term ‘good’ can be understood as:
It is always advisable to keep a critical eye on the figures. After all, an accepted and properly functioning analysis on a dashboard can at any time display incorrect figures when something goes wrong with master- or transaction data somewhere in the operation or registration. When these figures are used for decision making, it is evident that they are ‘good’.
In the following article the translation of the above is done and applied to all reporting options within Oracle JD Edwards in line with the Oracle BI Continuum. From small to large, from flat to multidimensional, from historical to predictive and from operational to strategic.
Want to learn more?
Author: Rick Brobbel
BI Consultant at Cadran Consultancy