There are only two moments that matter in a piece of data’s life: the moment it is created and the moment it is used. The problem is that these two moments are rarely connected.
The fact that this connection is often missing is one of the major obstacles for companies hoping to capitalise on artiﬁcial intelligence (AI) and machine learning in their supply chains.
Deloitte’s 2018 survey of chief procurement ofﬁcers found that more than 45% believed the lack of integration and poor-quality data were preventing the effective application of digital technology.
The majority (78%) of CPOs in Deloitte’s report also said cutting costs was a priority.
The value of digital transformation
Of course, costs can only be cut so far. There’s no question that price is an easy measure of value, but there’s a limit to this strategy. Suppliers are looking to make a profit, so when prices inevitably hit the ﬂoor, the ‘easy win’ of cost reductions hit the floor with them.
That’s why the focus for companies should be on adding value instead. This can come from improved supplier management and increased supplier insight.
As often seems to be the case, technology is seen as the solution. According to a report from consultancy Oliver Wyman, 84% of procurement organisations believe digital transformation is going to change the way they deliver their services in the next few years.
The areas where technology is expected to have a major impact when it comes to creating value include reducing companies’ wage bills and building stronger supplier relationships.
Bringing data to the fore
One of the issues for procurement organisations is that transactional activities already consume a lot of time. This is why it’s particularly important for them to make the most of the opportunities offered by technology – which in turn means making the most of good data.
Machines can only learn if the data used by the technology is clean, meaning accuracy and completeness are prerequisites if you want to benefit from AI. If you have poor-quality data, there’s little advantage in spending money on the latest software.
The truth is, however, that data has been the poor relation of the corporate world for many years now. Recording and inputting data has been treated as an administrative function – necessary, but in no way part of the strategic agenda. Those who create the data don’t use it, while those who use it can’t alter it.
Different departments need different things from the data, meaning nobody takes ‘ownership’ of it. It’s precisely this attitude that is hampering the ability of many companies to benefit from AI.
Embracing data governance
Companies that do utilise data efficiently stand to gain a lot. A 2014 report from IDG Research Services revealed that companies with effective data grow 35% faster.
What’s clear is that companies need a culture of data excellence, including creating a process of data governance that tasks one individual with understanding data use among different departments and co-ordinating it between systems.
It needs to be viewed as a strategic role, able to take an overview of corporate data needs. Consequently, objectives must be deﬁned around business outcomes to prevent the issue being siloed by a single department, such as IT.
A ﬁrst step towards establishing data excellence may be cleansing and consolidating existing information to provide a ﬁrm foundation for future use.
No substitute for human intervention
Technology can only provide limited help for this critical step, which accounts for around 80% of the time spent on machine learning projects. Truth is, there’s no substitute for human eyes ensuring that there are no double entries or oversights lurking. Too many companies over-think this process.
Of course, you need to spend some time establishing your data standards by considering what data you need and why. But once those standards are set, it’s simply a case of working through existing data and comparing it, piece against piece. Key to this process is managing data centrally – only then can the job of data cleansing become hardwired into an organisation’s mindset.
Also, bear in mind that data governance is an ongoing task since business databases can change and decay by as much as 30% a year. New information must be handled in a systematic way if it is to be of use. New suppliers need onboarding, existing supplier details need updating and redundant suppliers need deleting.
If there is a clear path in place then data be managed in a way that is useful for the organisation’s future. Technological innovation can of course help, but data is the underlying foundation that empowers and enables this. To understand your data is to understand your business.