The importance of data in the digitalisation of procurement is well documented. Half of the respondents to Deloitte’s CPO survey 2018, for example, indicated they are proactively leveraging intelligent and advanced analytics to optimise costs. A similar number of respondents said data was helping them improve efficiencies in their processes.
Elsewhere within the function, analytics is increasingly being used in terms of management reporting, which is helping to better predict demand and manage expenditure in real time.
Although the potential benefits are clear, there are still challenges and limitations when it comes to digitising data – particularly the historical kind, which could potentially provide CPOs with even greater insight.
These challenges include finding the data, structuring it in a format that can be read and analysed, and picking the technology that can then read it and provide the right output.
Marcell Vollmer, chief digital officer at SAP Ariba, believes CPOs are keen to access and glean insights from unstructured, historical data that currently sits in spreadsheets, paper documents and emails. But, he says, they first need to think about digitalising their processes first. Only when these processes are in place, Vollmer suggests, will staff be able to store and analyse that data and, ultimately, unlock its potential.
The issue here, he says, is the actual digitalisation process is tricky and requires some careful attention.
“Procurement needs to be Apple-easy and Google-fast. The user experience is absolutely key and it’s a tough job explaining to millennials why it is so much more difficult for them to order something at work than it is at home,” Vollmer says.
“The expectation of the function is a relatively simple one: increase value, reduce costs and improve the customer and supplier satisfaction.”
Organisations can achieve that by adopting technologies such as artificial intelligence and machine learning.
Although Constantine Limberakis, director of product marketing at spend analysis provider SpendHQ, agrees CPOs need to harness technologies he disagrees slightly when he says the most important job is finding out where data exists and then working to apply technology to it.
“There are data lakes within an organisation with information held on different and multiple enterprise resource planning systems, paper documents and purchasing cards,” Limberakis says. “You need to collect and aggregate it, but you first need to appreciate the scale of the task.”
He tells Procurement Leaders that CPOs need to build a map of existing sources of information.
“[You need to know] how many data sources you have and where that data is coming from. You also need to know how many suppliers you have and what your processes are.
“You need a certain amount of discipline going through back-end and master data and cataloguing manual information. Artificial intelligence and machine learning can help you improve that process and understand more quickly what you can do with that data.”
Limberakis, however, says getting 100% of an organisation’s historical data uploaded digitally is “probably impossible” given both the scale and a lack of digital talent working in the procurement function.
“You need to have the data scientists and data experts in your team to manage the collection of data and understand the intelligence,” he says.
Clive Rees, VP, international CPO, at Fujitsu, acknowledges this talent gap. “We could lose ourselves in all the data so, within our team, we decided to get the younger members to give us a view on how it will be a benefit,” he says. “They have grown up in a different era and they can help us with how, in a practical sense, we can get market data and intelligence from suppliers and vendors quicker and in a more digestible form.”
Rees’s department is still at the nascent stages of developing its digital data strategy, however. “We have looked at robotic process automation and how it can potentially help us, but we are still in the infancy of our digital journey. Looking at digitising historical data is something for the future.”
While CPOs stated in Procurement Leaders’ CPO challenger guide 2019 that data analysts/scientists were the top emerging roles they were hiring for in 2019, that talent is not currently in place to help them right now.
One of the areas in which data scientists would be able to provide the most support is in handling and transferring data where the risk of data loss is at its greatest, as well as offering valuable insight into potential cybersecurity threats and where compliance is needed for specific legislation such as the European Union’s General Data Protection Regulation. Such work is time-consuming, manual and could easily distract a buyer or category manager from their day job.
Beyond having the right expertise in the team itself, Spend HQ’s Limberakis believes it is also essential to secure buy-in from across the business to improve the function’s chances of success.
“All business units and IT need to collaborate [on something like this],” he says. “The process of digitalising data needs to involve all stakeholders because there is a high risk of not innovating. Any digitalisation project needs to be one of incremental learning. It may not be successful the first time around, but you need to keep trying. That philosophy needs to run throughout the organisation.”
For SAP Ariba’s Vollmer, this includes CPOs building a proposal of what they want to achieve. “You need to provide a vision and share the need for change with your teams to get them aligned. That includes different lines of the business, executives and suppliers.”
Procurement chiefs also need to foster a culture of change that will help everyone “embrace these disruptive technologies”, he adds.
While some think it is about getting the technology right or finding out where data is stored, Limberakis points out procurement chiefs can be the ones who complicate the picture. First, they must know what they are trying to solve.
“As a CPO, you need to ask yourself what the problem is you are trying to solve and then look at what the biggest issue is. You also need to look at the processes to solve that and take it stage by stage from there,” he states. “If you identify this, you will know what historical data needs to be embedded first.”
This need to be focused, know exactly what to look for and to start small when digitalising data is a point that Simon Whatson, a principal at procurement consultancy group Efficio, agrees with.
“There’s no point making decisions on digitalising data if you don’t know what you are going to do with it. While it is possible to upload all your unstructured, historical data, it takes a lot of manual work and time to repopulate a new tool. It is important to find the business case for it, otherwise it will be a futile task,” he says.
Peter Wetherill, a principal at Efficio, adds: “What historical data do you want, how will it help you and how will you use it? You need clear thoughts on this. What can you harvest and reuse for the future? There is value in historical data such as looking at the price of a certain product over a period of time and seeing how it has fluctuated. You can also benchmark data and look at various unit cost models.”
Jill Ivancich, senior manager of the business process services group at DXC Technology, believes failing to have a solid “grasp” over this data is holding back wider digitalisation efforts. “You need to have a data plan to understand what you have historically. To be able to take the plan and enact it is where we struggle most,” she says. “I’m not sure many procurement organisations understand what they currently have regarding data throughout the supply chain. We have to extract as much as we can and have a big data plan before moving into digitalisation.”
She reemphasises that the whole organisation must play its part in finding the solution. “The chief information officer and CPO need to come together. You don’t want silos looking at data – you need to look at it holistically,” she says. “You also need true process mapping to understand where you have gaps and where you have problems. Do you have the best processes in place? This is much better than going straight into a digital plan and saying: ‘What can we automate right now?’"
“Too many times, when you look at data, it has not been cleansed nor normalised. It hasn’t had that human touch, that intellectual experience, which is needed for data to be put into action effectively. You need to cleanse as much of your historical data as possible before getting it into a digital platform.”
Questions CPOs need to ask about the data, Ivancich says, include: When was it purchased and by whom? What were the price and the purpose? Who were the suppliers and where are the risks? Where are your service levels?
“Data is constantly evolving and moving. You’re never going to get a completely accurate view as it is a living, breathing and changing set of numbers,” she adds. “It’s still a very manual process apart from invoicing historical and current data. That is where the most accurate and efficient data is at present, as that is where most companies have invested digitally.
“The data around price points and what is being spent with suppliers is getting to a point where it is live, accurate and actionable on a digital level. It has been cleansed and normalised, then analytics is being used on top of it.”
One key challenge Ivancich identifies is ensuring CPOs do not digitise and take data from a “process that was wrong to begin with”. They should carefully review the information and processes they want to digitise, making sure it is valid, accurate and works correctly.
She adds: “You don’t want to waste your time digitising that. First, ensure you have the right process before you go to a digital plan. The mistake is taking a flawed process or one with too many steps and trying to automate it. It will still be flawed when you move it to a digital platform or tool.”
Ivancich stresses that CPOs will benefit from data and process digitalisation. “You will see operational efficiencies, significant cost reductions and more accurate data. But, like anything else you do, you need to prepare, get things in order and understand where you want to go. You need to have an outcome in mind.”
This article is a piece of independent journalism, written by an experienced journalist and commissioned exclusively by Procurement Leaders.