When procurement goes digital: Day 3 - Data
Today we tackle part three of a five-part series of articles in which we look at the often neglected key enablers of aprocurement digitalisation journey in the South African context.
So far, we’ve discussed ‘purpose’ and ‘people’ as the first two of these enablers. Today, we’re discussing forgotten enabler number three: Data.
Neglected enabler of the day: Data
We can probably write a book on data rather than an article! For the moment, however, we’ll only touch on some high-level but key thoughts around the subject of data when it comes to digitising your procurement world.
Firstly, let’s start with why data is so important in your digitalisation journey. One of the main drivers of digitalisation is to automate repetitive tasks and provide your procurement team with insights to allow them to focus on the activities that require fancy footwork and experience to ensure the best outcome. We may get to a point in the future where artificial intelligence may equip systems to do the thinking and sourcing of information for themselves, but, for the time being,systems are still dependent on a structured source of data to perform most tasks, formulate suggested responses and actions, and provide us with information in various forms (analytics, alerts, views etc.) to allow us to successfully do our work. Without the required data, we constrain a digital platform, making it very hard for the solution to add the value it’s supposed to.
‘Required data’ is defined mainly by using three high-level properties:
Data completeness is relatively self-explanatory and refers to the fact that the system is provided with all the sets of data, each containing all the content required, for the system to provide the functionality for which it is used. Completeness is mostly ‘ensured’ by mechanisms like mandatory fields, forcing the user to populate these pieces of information.
Standards refer to the format and rules applied when entering data into a system, for example, the format in which a telephone number for a supplier is populated, the structure of a supplier name (full names, abbreviations accepted, whether it starts with upper case letters, and so on). Standards make it easier for a system to interpret data, as it can use specified standards as a template to help interpret information, for example, determination of possible duplication of master records. In some systems and for some data inputs, the use of prescribed standards can be controlled using input masks (templates) and rules as control checks on incoming data, but for a large proportion of data, defects need to be highlighted with analytics and addressed via human intervention.
Data integrity refers to the correctness, and as a result trustworthiness, of data available to the system. Ensuring the integrity of data is the most difficult of the three properties and is largely dependent on manual intervention. Due to the nature of this activity, the integrity of data can not only depend on evaluating data on face value as it often requires validation from external sources, for example, the email address of a supplier may comply to the format of the supplier email address domain, but further investigation could reveal that the actual email address does not belong to a person working for the relevant supplier. This validation would require access to the supplier’s HR data base and, as a result, would most likely require assistance from the relevant supplier contact person.
In the case of data integrity concerns, it really is a case of prevention being better than a cure. For this reason, most organisations that are serious about valid data do not only clean data as part of a project activity every couple of years, but establish a data management capability.
This capability can vary in size and scope per organisation and will involve a combination of some of the following elements:
technology (for example, master data management platforms)
processes, procedures and controls related to how data is provided to various systems
analytics and performance indicators.
Discussion around the implementation of a data management solution will require a chapter in the book we should write on data. But, in short, the main purpose of this function is to ensure that all incoming data created by any party is created according to the agreed standards and templates and to ensure that the quality of data is controlled by an entity for which this is a first priority. The result is that all systems have the data required to function optimally and that, in most cases, all stakeholders have quick access to information that allows them to make decisions related to the well-being of the organisation.
Examples of the structure of supplier email addresses mentioned so far may not get you too excited, but think of the following: My ‘Top supplier spend report’ can show me that my Top 20 suppliers make up 70% of my spend. This is helpful, but only to an extent for prioritising suppliers for savings opportunities and effort allocated to manage the relationship. But let’s say the quality of my data allows the system to do a simple thing like identify multiple supplier numbers belonging to the same supplier and likewise for items. Now my top supplier spend report can group spend accordingly and I may realise that my top 20 suppliers actually make up 85% of my spend and two of these suppliers have four contracts each for the same items, but across three business units at different prices and with different payment terms and that 90% of spend against this supplier entity actually sits against the least favourable of the contracts … that seems slightly more useful doesn’t it?
At face value, data is probably the least interesting element of digitalisation, but if handled correctly, it will allow your digital platforms to show you things about your business that may just blow your mind!
Emile Olckers is a principal consultant at Supply Chain Partner. He applies his time enabling customer procurement and sourcing optimisation and visibility using market leading digital platforms. Able to analyse customer challenges and provide suitable solutions using a combination of technology and process enablers, he is passionate about partnering with customers to deliver on their procurement digitalisation vision .Emile has spent his 18-years in business providing customers with insight and visibility into their inbound supply chain operations and leading the development of pre-populated business intelligence tools for SAP to enable customer visibility in very condensed periods of time. He has been involved with customers spanning multiple sectors: manufacturing, mining, petroleum and chemicals, transportation and airlines.