When planning a Business-Intelligence strategy, most companies focus on technology: what database types to use, which ETL software, SAS suites, and cloud storage to set up. In the last 10 years, while working as a data driven SEO, I’ve seen many BI strategies fail and only some succeed. In my experience, the key secret to a successful BI strategy lies in its human dimension. The employees involved with the implementation of your BI strategy face several challenges. They have to:
- implement the ETL processes
- understand the data
- find and fix errors
- use the dashboards
- live the KPIs
If the employees of a company refuse to work in a data driven setting, your strategy will fail. If you choose technology which is not easy to use and could not be customized for the needs of the end users, your strategy will fail.
Likewise, if you only choose a BI technology because “the big companies use it”, or because of some fancy buzzwords (like “Marketing Cloud”), you could easily get caught in a spiral of:
- exaggerated expectations from the management
- runaway costs (for software licenses, employee training as well as expensive specialists)
- little usage of the tools (because of complexity and flat learning curve)
The first key component of a successful BI strategy is expectation management: communicate, listen and plan carefully. And above all: involve the employees directly in the implementation process.
The second key component is the choice of a technology mix, that is easy to learn, implement and use. This results in a steep learning curve for the users, which in turn contributes to high acceptance of the new system. I would recommended a Linux based Open Source infrastructure, with databases like MonetDB, PostgreSQL or even MySQL (those who really need BigData, may combine Hadoop with Hive). Such a system should be used in combination with a powerful BI Software like the Pentaho-Data-Integration Suite (which is also Open Source but offers a premium service). If your BI problems are closely related to management, than you could combine Pentaho-Spoon for ETL with tools like Jedox to get Excel-based dashboards including writeback capability.
These few simple steps will put a powerful and affordable Buisiness Intelligence Strategy within your reach:
- Hire one experienced Head for your new BI Team. If the stakes are high or if you need a big team, then you also need one experienced senior for every base component (like a server sysadmin, a data base management specialist or an expert on cloud infrastructure).
- Choose a tool setting that is simple, easy to learn, scalable and as open as possible to reduce your costs (you get savings on license fees, you will keep training expenses low and limit your dependency on high-priced specialists).
- Select some slightly technology savvy employees from the departments scheduled for switching to a data centric working mode. Then train them by the experienced team members and let them implement the necessary tasks themselves. (You may also use them as in-house trainers for the rest of your staff later on.) If this is not enough workforce you also have to hire some new employees to fill the team. They need not to be overly qualified (and thus expensive); ideally, you should fill the required positions with people who are interested but not yet so experienced.
- Start! 😉
What you get is a dynamic, self-learning team that truly lives its ETL processes. People who intuitively eliminate flaws in the data sources, breathe the data flow and implement what the departments really need. Or in other words: you get what you really need in a cost efficient way! 😉