Market Insights Manager at Derivco Durban, Mike Mallett shares with us what he has learnt building his own BI team and gives us insight into how you can build your own.
Start with your people!
It sounds cliched but it really isn’t, because you require people to engage with, implement and realise your BI strategy. Who you choose is critical to your success. Not everybody is cut out for BI. It requires specific technical and soft skills because you are moulding raw, meaningless data into information and presenting it to people who must understand and act upon it. The insights you deliver will drive critical decision making and as such your stakeholders will place significant trust in the abilities and output of your team. You must identify people that are willing and able to deliver from this perspective.
Build a diverse team
Providing a BI solution requires a broad range of skills. You need a rock-solid data foundation, so you need an architect capable of designing, developing, and implementing a data architecture. You likely already have lots of raw data, but that’s all it is. Your architect must transform your raw data into accurate, trusted and consumable data. You need BI developers to convert this consumable data into information (reports, tools, dashboards etc.) and they will work closely with your analysts to turn this information into insights. Your analysts will typically manage key business requirements and stakeholders, and drive the responsibility of turning these insights into actions and business decisions. Of course, the only way you can guarantee accuracy of your data, and hence the validity of your analysis, is if your solution is thoroughly tested by a tester who knows how to test BI data. Finally, you need a person who understands virtual and physical infrastructure, with the ability to create and support robust, secure and scalable systems to host your BI solution. Don’t forget yourself too. Assuming you own responsibility for the solution, you need to spend a great deal of time understanding and documenting the business drivers and technical considerations. Your ability to analyse and strategize for the options available are key.
Know your customer and their requirements
Knowing why you are embarking upon a BI journey is crucial. Business requirements are a big driver in influencing the way you design and implement your solution, but they can be diverse. Are the factors influencing your solution internal or external, are they financial or behavioural? How quickly do your customers want the solution, how accurate must it be and what are they willing to spend? The project management principle of You can have it Fast, Accurate and Cheap – pick two is a great question to apply to your BI strategy. Discuss the options with your customer, understand what they want and propose your solution based on that, but they can’t have all three.
Cry now, laugh later
You start with your people, but your data is everything, it really is. You absolutely must get it right. Spending time up front building a solid data platform is time consuming and painful, but it lays the foundation for your longer-term success. You cannot skimp on data integrity and the value of doing additional work in the short term (the crying part) leads to far more advanced BI capability, efficiencies and benefits in the longer term (the laughing bit). With more and more customers (in the digital space at least) moving away for prescriptive reporting requests in favour of access to data to create their own reporting solutions you have another reason to get your data right. Customers can only consume your data effectively if your schema, data fields and definitions are understandable, and your business rules consistently and clearly applied throughout your BI suite. So, buy the time to do your data right and manage your stakeholders throughout the process. The benefits will follow and they will be impressed.
Try to Provide iterative value
Great BI takes time, especially if you are building your solution from scratch. If your initial business requirements allow you to focus on a concentrated area of your data, build that area first and deliver insights and business value before moving onto the longer-term solution. But don’t do so if you are going to incur technical debt or throw away development in the process.
If you’ve built a team of varied skills you will also be more enabled to provide iterative value. While your data architect is focusing on a longer-term strategy your data analyst or scientist can work with other sources or sub-sets of data to provide valuable insights. Look for these opportunities; delivering iterative value provides business value to stakeholders, trust in your longer terms strategies and faith that they will see return from their investment in your team.