By Michiel van Staden, Data Analytics Lead, ABSA

Leadership spend most of their day in meetings, making decisions. Operations are mostly busy doing operations, taking decisions. In the spaces between, there is hopefully some time for all to engage the relevant colleague communication platforms, decision-making.

When running, we run. Every now and then we might pause to catch our breath, engage those around us. Whilst running, there are apps that feeds us relevant info on our progress as we go and on pausing, gives us just what we need to decide on the way forward.

Coming from 13 years in data analytics experience across fraud prevention, credit risk and operations to digital and marketing, I’d like to talk about practical ways to feed decision-making in workplace meetings, operations, and the spaces between, with relevant information.

Reports Coming From Systems

Within our organisation, we have many different systems collecting information. Some of these have been developed to engage potential new customers, others to process applications. There are systems to manage accounts whilst the relationship is in good standing, and still others for when that relationship goes through challenges.

Based on the relevant function, each one of these processes would generate datasets that would then be stored accordingly. Much of this does now reside on the same centralised data warehouse, but there are still nuances in terms of dataset-specific access and formatting.

Often this would dictate the way reports are developed. Datasets relating to applications would for example form the basis for a sales report, often taking shape around the pieces of information that happen to have been stored. In developing this report, the temptation is often great to include as much information as available, some of which might even not be accurately captured or well understood.

From this position, getting the business to actively access and use these reports is an uphill battle. Even for those very close to the relevant system process, field names as captured in the data can be totally unfamiliar, whilst in parallel they are still struggling with navigating the reporting tool and trying to make sense of the overwhelming amount of information.

Already being pressed for time, this does not bode well for adoption, but with time it is possible to get there. Demand is there to track business performance, and thus leadership is forced in a sense to make the reports work. As stakeholders continue to engage with the reports, questions do emerge around the fringes.  For example, enquiring what role prospecting to potential new customers played in sales and what happened with the relationship post sale.

At this stage there might very well already be prospecting and existing customer relationship reports. In some cases, the data specialist responsible for sales reporting might also be close to those, but more probably not. Whether to merge these reports or keep them running in parallel, with the increasing likelihood of overlaps as they grow, can be a very complicated problem to solve. Can the different data sources be compiled practically into one report? Are the other reports being used? Will the overlapping numbers be consistent or conflicting?

When not handled well, you can easily end up with a myriad of reports, containing duplication and inconsistencies, whilst becoming increasingly too large, complex, and unwieldy to be of any practical use. Add to this ad hoc requests for specific pieces of information increasingly landing on data specialists laps because stakeholders are not able to effectively find the information themselves.

The Business Does Not Know What it Needs

When you do ask the business what intelligence they need, they are not able to tell you.  Not having a view of what is possible and available in terms of reporting makes it very difficult to devise specific requirements for what the business will need and be able to practically use.

As data specialists, our first task is to get to know the data on offer very well. We have to figure out how to access the data coming from different systems, need to ask the questions towards understanding exactly what each piece of information means and ultimately need to become very comfortable in weaving data from these different sources together into an end-to-end picture.

Additionally, it is also key to understand the business. What is the business strategy? How does it make its money? What are the key processes? What does the business offer its customers? And lastly, take time to listen to and understand your audience. What does their typical day look like? What challenges are they facing? How do they make decisions?  What are their thoughts around data? How comfortable are they in working with data?

Only once you’ve got a very clear understanding of all of these components can you engage your stakeholders very practically, giving them the business intelligence they need to do their jobs better and make informed decisions.

BI in the Moment

Knowing exactly what data is available and which processes are core to the specific business area, sit with your key audience to understand what the key meetings are in their schedules and unpack what information would better enable them to make the critical decisions in those sessions. Give them what they need.

Also spend time in the various operational functions. Unpack what systems they are using and how they make decisions within those processes on a day to day basis. Give them what they need.

• Understand your data
• Understand your business
• Understand your audience
• Give them what they need

In the spaces between there might still be additional, maybe slightly more generic information that could give your audience the latest on business progress or performance, informing their decision-making more holistically. Actively work with your stakeholders towards narrowing down a handful of key metrics, giving them what they need at a glance, in an email or messaging body. Clicking through to reports on a daily basis quickly becomes very tedious and discouraging.

Before adding any new metrics or views to a dashboard, review all the existing ones. Are they still relevant and value adding? Or is the report starting to slide into becoming less valuable?  Then something has got to give. Nobody wants to keep generating reports nobody uses, and nobody wants to be inundated with reports they cannot use. Give them what they need. BI in the moment.