In the practice of indirect procurement, purchasing software to support various departments and operations is a concern those in charge should take quite seriously.
However, the consideration grows much more complex in regard to big data analytics. According to a 2013 report conducted by SNS Research, global investment in data analysis solutions is expected to expand at a compound annual growth rate of 17 percent between 2014 and 2020.
This technology comes in many forms, and iterations are often constructed to befit the needs of certain verticals, such as manufacturing or health care. For instance, a business specializing in qualitative data analysis tools may deliver solutions through cloud computing, a practice known as software-as-a-service.
Listed below are three steps procurement officers should follow when surveying big data analytics tools.
1. Determine how operations will expand, contract and change
The average professional would maintain that looking for a solution designed to last a company 10 years is somewhat of a lost cause, but this isn't necessarily the case. Many enterprises are still using database engines as old as Windows Server 2003, so it's fair to reason that purchasing software that will last for a decade or longer isn't necessarily a tall order.
What may scare spend management specialists away from making a large investment in an analytics solution is the uncertainty of how much a business may evolve over a decade. For instance, if an organization is due to become more data-intensive over the course of a certain period of time (i.e. involve "low-level" workers in data analysis projects as opposed to just data scientists), then procuring a solution that is easy for the average Joe or plain Jane to use is essential.
2. Assess the complexity of your needs
If a manufacturer with 40 factories equips each of its facilities with Internet-connected devices, its data analysis needs are going to be incredibly high-end. An analytics solution of this caliber should be capable of performing the following functions:
- Aggregating the information from sensors distributed across a single facility or more in real time.
- Allowing machines to make "smart decisions" based on the data produced by Web-connected electronics.
- Providing professionals with accurate visualizations of intricate situations.
- Choosing data representations based on the nature of the intelligence that was formulated, a feature that is particularly advanced.
- Sending reports to machine specialists detailing a problem with a mechanism and what can be done to fix it.
The list goes on. In contrast, a finance team may not require an analytics program equipped with the aforementioned applications. As it happens, those who are well-versed in Excel and other basic data analysis solutions should have no problem creating the kind of in-depth reports they require to give their clients or employers educated advice.
3. Scrutinize potential providers
After you've determined the evolution of your business and scrutinized your current and future needs, you need to figure out which company will be able to support and accommodate you. Generally, you have a choice between one of two corporate types:
- Umbrella developers: These are the kind of enterprises that create software of every kind imaginable, such as Microsoft, Oracle or Google. While these companies offer sophisticated engines, their support resources are also dispersed across other verticals to help customers using their email, productivity and operating systems.
- Specialists: This contingency is made up of enterprises that specifically focus on creating analytics software and nothing else. Tableau and Datawatch are two companies that operate around this model. The advantage of procuring software from one of these organizations is that they use all their talent and resources to set new standards for what makes an optimal analytics program.
Choosing a data analysis software provider should only occur after procurement officers have thoroughly assessed the needs of the business.