By Irene von Buddenbrock
Data doesn’t work for everyone, so stop reading. It’s true, business intelligence (BI) tools and hype slipped sadly into the Gartner Trough of Disillusionment in 2012. It is also true that BI overpromised and underdelivered in many expensive use cases. However, today what stands true is that BI has withstood the tarnish it earned back in 2012 as it has steadily evolved thanks to industry expertise and emergent technology. Artificial intelligence (AI) and machine learning (ML) have provided BI with that extra edge that is needed to cut through the data and deliver both relevance and value to the organisation. According to Karien Bornheim, CEO of Footprint Africa Business Solutions (FABS), it has become critical to invest in solutions that ensure the delivery of relevant data to the business.
“The challenge for most organisations is that they have too much data,” she says. “Everyone does. This is one of the reasons why BI ended up losing traction and value at one point – it was sifting through every last inch of data, even that which had little to no value. This has led to a fundamental change in how people are using and approaching data today.”
The truth is that not all data is interesting, valuable or relevant. The lakes and pools of Big Data are not packed full of insights that can transform the organisation or breathe fresh life into ancient products. Some data is just not going to make any difference to the business. However, there are ways of improving how you leverage, store, approach and manage your data to ensure that you do get the relevant insights.
“One of the first steps in ensuring that you invest in relevant data is to improve how you collate and collect the data,” says Bornheim. “The methodology you use to collect and store your data is critical to how the data will serve your organisation. This is defined by the type of data you need – identifying the information that your business can use to further its growth and ensure its sustainability.”
The type of data your organisation prioritises will depend on the industry. Retail, for example, would rely on data that reveals shopping behaviours, customer preferences, and customer movements; while in the manufacturing industry, the data would track fleet movement or productivity. At this point in the process, the business requires a strategic view of its goals and a realistic understanding of what the data is expected to deliver. Without a solid foundation that comprises strategy, output, goals, and deliverables, the data will do little more than spew information that has minimal value.
“Once you know exactly what you expect from the data, then you need to focus on which data you will be collating and where this will be stored,” says Bornheim. “The data has to be stored and managed meticulously, otherwise it just starts to rot. Not in the physical sense, but in terms of the value that it offers. Data has to be timeous as well as relevant. Thorough data storage will also make it easier to analyse and access which will improve how insights are delivered.”
Data is also going to arrive in storage looking less than spotless. There will be typos, errors, low-quality information and dodgy data analysis. At this point, data needs to undergo a thorough and regular cleaning that normalises the information, ensures that the information is of high quality, and that removes glaring errors that could affect the analysis. This is also the point at which the business needs to relook its siloed approach to information sharing and gathering. Silos will minimise the reach and relevance of the data considerably.
“If you consider how data from finance could potentially impact performance in sales, it makes sense to pull data through all silos to create a holistic view of the business,” says Bornheim. “You want to know how performance in one area of the business could be impacting on performance in another area. How sales can benefit from customer insights in finance or how production can benefit from insights generated by marketing. These pockets of data can offer the organisation remarkable visibility if they are brought together.”
These are just some of the steps that the business has to take before data can offer anything more than hit and miss insights. There are, of course, other factors to consider along the road to highly optimised and relevant data delivery, but these are often specific to the business and the industry and would benefit from collaboration with partners that have an in-depth understanding of data nuances and applications.
“Consider collaborating with a company that understands your business and how your data can benefit you over the long and the short term,” concludes Bornheim. “Data can become one of your most valuable resources and the processes that surround extracting and analysing it does not have to be prohibitive or excessively complicated. Just start with a reliable partner that will set you on the right path using a methodology that fits your business and your budget and grow from there.”