Guest post by
Himanshu Bhardwaj, Consultant - Business Intelligence, Oracle Practice, Enterprise Solutions, Infosys Technologies Ltd
'Should I do it on my own' or 'should I have someone else do it for me' is a question that mankind has asked itself since time immemorial. This question, commonly put as 'Build vs. Buy' in the IT parlance, has been commonly debated in the Business Intelligence (BI) space as well. From BI's perspective, the Build vs. Buy debate can be discussed in two contexts. One point of view could be to consider home grown made-from-scratch reporting applications as 'Build' and standardized platforms and packages like OBIEE as 'Buy'. The second point of view could be to consider applications developed on tools like OBIEE as 'Build' and prebuilt packages like BI apps as 'Buy'. Since standardized tools/packages have clearly demonstrated their superiority over home grown reporting applications in terms of performance as well as scalability, the first debate is pretty much settled. So, for the purpose of this blog, we will focus on the second point of view that this debate can take.
Since we believe that the choice between OBIEE and BI Apps is purely circumstantial, we will not attempt to choose an unconditional best between the two. And, since this debate is more meaningful if discussed from the perspective of 'when to put what' rather than 'which one is superior', we will attempt to arrive at a set of parameters that should be evaluated before coming at a decision.
Faster deployment, lower TCO and better ROI are perceived to be some of the primary advantages of BI Apps. The prewritten ETL and the prebuilt metadata and reporting content provides a handy jump-start thus reducing the deployment time to less than half. Less time implies less effort leading to lower costs and superior ROI. And once we add superior functionality which is otherwise hard to replicate, adoption of industry best practices in terms of KPIs, potentially better security and performance and higher adoption rate by the business community to it, the balance looks to be tilting in favor of BI Apps. But there is more to the story.
The Source Systems Aspect
Rosy as it may seem, the truth is that all these benefits of a packaged solution come with a huge rider. Before they materialize, they will ask for a certain set of pre-requisites to be present. As we move from a non-customized standard source system (say an EBS with little or no customization) to a customized standard source system (say a highly customized EBS) and finally to a home grown non-standard source system, the benefits will start getting flimsier and flimsier. So, an honest and thorough assessment of the existing source systems should be the first step in the 'Build vs. Buy' decision making exercise.
The Customization Aspect
Proponents of BI Apps say that more than 80% of the metadata would be made up of non-unique content irrespective of the industry. However, this may not always be true and sometimes not even for organizations operating in the same industry and servicing similar types of clients. One of our clients in the financial services domain got merged with a same sized competitor and we realized that the common denominator is much smaller than it is perceive to be. It was seen that though the two companies looked at the same metrics but calculated them in very different ways. This resulted in re-building of the Oracle BI repository as per the 'new' definition and a lot of customization had to be done. So, in such scenarios a Buy option would imply a much higher license cost without any offsetting of effort or cost in customization.
The Cost Aspect
A back-of-the-envelope calculation tells us that the licensing cost for a couple of modules of BI Apps, for a user base of 500, can set an organization back by more than a million dollars! So, the savings resulting from lower deployment and maintenance cost of BI Apps should be enough to offset the higher licensing costs. But this may not always be the case. Thus, a Cost benefit analysis of this nature is a must before making a final decision.
BI Apps is undoubtedly a fantastic product which can offer a lot of value to its clients. But there are certain circumstances where a Build option will prove to be a much safer bet. An honest and thorough assessment of Source Systems, Reporting Requirements and Level of Customization required along with a Cost Benefit Analysis is quintessential before we go ahead and recommend it to the client.