5 steps to successful analytics in large organisations:
A South African perspective
By Rajiv Daya and Zahida Khazi
We all know insights are important, however despite that, you’ve probably found yourself waiting months for reports and analytics – in fact, you may still be waiting. When finally receiving the long-awaited report, often it’s incomplete or not what you needed. Consequently, you start again, requesting another one-off report, hoping that when received it will provide the information you need to build insights.
Report building is often a manual process, taking weeks to compile, and are often only done annually, bi-annually or – at best – quarterly. Most often, this is the result of a combination of factors:
- the underlying data is unclear
- a number of business rules have to be applied
- numerous business units are involved, all using the same data, but applying different rules, resulting in different outputs
The problem is not producing analytics. The problem is making the right data available and preparing it appropriately to produce analytics and insights. So, how do you prepare your data to enable successful analytics?
The problem is not producing analytics. The problem is getting the right data and preparing appropriately
Step 1: Define the problem statement
Organisations today generate and store enormous volumes of data, all of which can be used to create analytics. But it’s a bit like finding a needle in a haystack. Clearly defining a problem statement – what you want to do with your analytics – helps create focus and structure to make sure the data answers the right questions. Without this, the analytics will be generic and superficial.
A good problem statement is specific and measurable, creating clarity about what is needed. Even with exploratory work, teams must work with business to understand the high-level business objectives being explored. This helps to identify the data to focus on to speed up the work.
Step 2: Understand the data and context
Typically, a business unit like IT or an analytics team is tasked with creating the report, but if they don’t understand the data they’re working with, the reports they produce won’t meet the business needs. A business lens is critical in providing the context necessary to guide the report development, providing a real-world application. In the end, the business will use the reports and so, they must understand the outputs.
A business lens is critical to guide the analytics development
Where processes span multiple business units and systems, understanding the various data points across the journey can provide valuable insights. But, connecting data across systems and business units is often not as straight forward as it seems. In fact, it’s a challenge many organisations have grappled with and failed to overcome. Linking data across systems and creating a cohesive data storage process is complex, and often results in delays. Unpacking the connections and understanding the business rules is essential when developing insights, especially across a journey.
Step 3: Access the data
In most organisations, data management is not centralised, and getting access to the relevant data can be problematic. Many organisations have policies against data sharing – even across business units – which limits what can be done. A good practice is to obtain sample sets of data to determine if it’s fit-for-purpose. Then, test it out with your business rules to see if the problem statement can be answered.
Where the data is stored is also key. You may have to move the data to a central repository – or a cloud platform – to run the analytics. This approach will also mitigate the potential impact on your technology infrastructure. Analytics can be intensive and can result in significant strain on the system when it needs to be available to users and clients.
Centrally accessible data is key to creating usable insight from analytics
Step 4: Trust the data
Inaccurate, incomplete or outdated data causes inconsistency, errors and unreliable outputs. This is why upfront data governance and management protocols are key. This helps ensure consistent, quality data is available, and that there is a clear process to remediate errors and issues. Understanding the source of data, and how it is derived or calculated through data lineage tracing, helps to verify correctness and pinpoint errors. Only by ensuring data quality can you enable trust in the data.
Only quality data can be trusted, so make sure you have high quality data to start with
Step 5: Enable self-service analytics
In today’s economic climate, organisations are searching for agility. To enable this, they want to be able to build reports themselves, rather than their needs being added to a backlog, waiting for priority. Mature data governance and management protocols are needed to enable the access management required for self-service analytics. Lots of easy-to-use data visualisation tools are available, although IT teams may still need to prepare the data – at least initially – so it’s ready for analytics in a central location. By empowering business users to create their own reports, the process can be significantly sped up, while providing a vital business lens.
Processes can be significantly sped up when business is empowered to create its own reports
Data is a critical asset, but it will only empower you and your business if it is appropriately prepared for analytics. To find out more about how BSG can partner with you to prepare your data for analytics, helping you effectively leverage this valuable asset, get in touch.
Get in Touch
Are you grappling with the hows, whens and whats of analytics and insight-led reporting? BSG can work with you to tailor the best solutions for your needs. If you'd like to find out more about how BSG can help enable your business to generate the right analytics for your needs, get in touch.