Caesars Entertainment uses data to predict consumer behaviour © Getty
Caesars Entertainment uses data to predict consumer behaviour © Getty

How to get value from procurement data

Rebecca Ellinor Tyler is former editor of Supply Management
4 May 2018

When it comes to procurement analytics, there’s a lot of noise. SM digs beneath the hype and reveals how to ensure you get far more than just garbage from your data.

No one wants to be making multi-million pound buying decisions based on flimsy data. But, according to David J Ward of Avidus Partners, that’s exactly what many procurement functions do. “Most procurement organisations, even relatively sophisticated ones, cope with a snapshot version of yesterday’s truth,” says Ward, who held procurement leadership roles in organisations including Merck and GSK before becoming a consultant. Ideally, he adds, procurement should own the whole space of supplier information, and have access to a wealth of procurement analytics. Rather than asking stakeholders for facts and figures, procurement should be able to inform others, using up-to-date graphical data that can be quickly translated into decisions.

Of course, some supplier expenditure analysis has been done for many years. Procurement has long compiled figures from finance or other departments to form a picture of their organisation’s spend profile. But while for some this means using enterprise resource planning systems to automatically extract information, many more are still populating spreadsheets with immediately out-of-date data. Neither approach can hope to be as comprehensive as those gleaned from crunching much larger, richer sources of information.

Done well, procurement analytics is not only cleverer – able to spot trends, avoid duplication, flag up poor processes and identify potential risks – it can also draw on multiple sources of information from both in and outside the business. This can include everything from the scores of credit ratings agencies, to social media chatter, to unorganised, unencrypted content in the deep web. And the enriched picture it forms can be used to detect problems, negotiate better prices, select a more favourable time to buy, move quickly to seize opportunities and boost efficiency. It can equip procurement with hard facts to be used in the boardroom and at the negotiation table.

CPOs know this. The results of the Deloitte Global Chief Procurement Officer Survey 2018 show for the second year running that leaders think analytics technology will have the most impact on procurement in the next two years. But it also found that, while there’s an appreciation of what’s possible, progress has been slow. Almost a fifth (17%) have no digital procurement strategy, and of those that have, less than a third believe it will enable procurement to deliver on its objectives and improve value for their business.

So what’s holding them back? One issue is the usual resistance-to-change challenge. There’s a certain ‘so what?’ faction who are achieving savings and have little inclination to evolve their approach. “It scares people,” says Chris Sawchuk, a principal and global procurement advisory practice leader at consultancy The Hackett Group. “They’ve been in the job a long time and done things a certain way. This is reshaping what a procurement professional does in the future. It requires different competencies and some may not be able to make that transition.” But over the next three to five years, he says, they’ll have little choice.

“Leadership is a key success factor,” adds Lance Younger, EMEA head of sourcing and procurement at Deloitte. “Leaders must be more disruptive, innovative and digital.” For digital transformation to have real impact, procurement leaders need to review and refine their digital strategy to make it more action-oriented, agile and scalable. Any change is clearly tricky without senior support and some investment is required but it’s actually relatively inexpensive cloud-based tech. “It requires zero infrastructure and takes weeks not years to get it running,” assures Ward.

When questioned about the key applications for procurement analytics, leaders responding to the Deloitte study listed cost optimisation, process efficiency and management reporting among the top advantages. Hearing aid manufacturer Sivantos Group has been using analytics to assist with cost savings for the past seven years. Group CPO Shahriar Tabrizi says the level of information, together with good structures, processes and personnel, has helped his department achieve double-digit savings throughout.

At electricity company SA Power Networks, the procurement team has been using used vendor spend classification in SAP and Microsoft’s Business Intelligence tool since 2014. Head of procurement Charlie Hollis says the biggest benefits have been a reduction in manual work to produce the information and access to live – and therefore current – vendor spend data. This has enabled the team to make informed category decisions around its approach to market and work on supplier relationship management, and realise cost savings.

Real-time data enables quick and more accurate decision making, which can mean seizing opportunities or avoiding pitfalls. The way Sawchuk sees it, competitive differentiation in procurement organisations will come from the ability to get and  respond to these insights faster than others. “Procurement needs to stop being so bureaucratic and taking too long to make supplier sourcing decisions,” says Ward. “Rigid structures and the category management of the past is yesterday’s news. Organisations should take a more flexible view of spend. Once you can see more in real time, you can be more flexible.”

It can also help mitigate risk. Scanning masses of data from multiple sources means systems can start to be predictive and identify potential problems like fraud or impending bankruptcy. Is a supplier always asking to be paid early and paying subcontractors late? The collapse of Carillion could have been flagged months before as a credit risk. But if you’re “set in stone on a category management plan, you might not have a big supplier going bust in the mix,” adds Ward.

The ability to flag up inefficiencies is not to be sniffed at either. Ward says the application of procurement analytics at one company enabled it to improve its cash management to the tune of an extra $1bn in the bank. “Most contracts have payment terms and most clients want to manage cash cleverly. Analysing payment terms meant they could see where they could wait to pay until they were contractually obliged, which massively improved cashflow.” 

But one major block is capability. Deloitte found 51% think their teams lack sufficient skills to deliver their procurement strategy. Only 3% think staff possess the skills required to maximise use of digital capabilities. Yet despite this, investment in training is falling and only 16% plan to enhance such capabilities. “People need to double-down on investment in this space,” says Younger. “Spend training budgets on analytics capabilities not negotiation training.”

Hollis at SA Power Networks agrees you need skills within your team to analyse data to make the right decisions. As Ward puts it: “You still have to bring your brain to the party.” Without individuals with curiosity all the data in the world won’t be much use.

This is where Antti Suorsa, head of strategic transformation and analytics, for sourcing, at telecoms company Telia, considers himself very lucky: “My business intelligence manager has been working in sourcing for a long time but comes from a maths/analytical background. He’s a unicorn in the field.”

There has been an increase in the number of procurement professionals with terms like ‘analytics’ in their titles in recent years. Vodafone has a chief of cognitive procurement, IBM has a chief analytics officer and Citigroup is hiring data science PhDs, says Sawchuk. These businesses, however, tend to be ahead of the curve, developing tools to not only improve internal capability but to sell services to external customers. A September 2017 report by Research and Markets named SAP, SAS, IBM, Oracle and Microsoft as leading suppliers of predictive analytics.

Data quality and integration can also present challenges. Connectivity is crucial and the information spewed out by systems is only as good as the data that goes in.

For Suorsa at Telia, data classification presents a challenge. “The information is categorised by accounts, so if they say ‘laptop’ or ‘computer’ it affects how it is shown in the analytics. We now have an algorithm to do a lot of the classification, which is a big advance.”

Charles Jobson, acting executive director at NSW Procurement in Australia, says he’s encountered challenges with poor quality, timely delivery of information and obtaining a complete set of accurate data across government. Ensuring data is clean and accurate requires ongoing maintenance. To overcome some of these challenges, he recommends professionals obtain a single source of truth, provide timely and relevant feedback to stakeholders and build good relationships to understand what they need and explain what procurement requires. “If you are meeting their needs, they are motivated to provide good quality data,” he says.

It’s even earlier days for predictive analytics. While the topic generates a huge amount of talk, there’s little capability just yet. Deloitte found just 3% have ‘fully deployed’ such technology, while 13% are scaling it up, 18% are in a pilot phase and 32% are just now considering it.

“We’re scratching the surface of what we can potentially do,” says Sawchuk. “I tell my kids, ‘if you want a secure future, focus on data’.” Whether it’s to predict spend, bankruptcy, hurricanes or disease, “everybody will soon have to be comfortable working with data.”



Phone company and mobile network operator Telia Company comprises around 20,000 staff across Sweden, Finland and the Baltic states. It manages an annual operating spend of SEK 46bn and capex of SEK 19bn. Antti Suorsa is head of transformation and analytics for sourcing. He manages a team of two that’s been in place for a little under a year, but has had full-time people dedicated to analytics for more than two. Static reporting tools have been replaced and now anyone in the company can produce reports on all third-party spend. Suorsa says it’s been an eye-opener for stakeholders and they’ve received “tonnes of positive feedback”.

This leaves his small team to concentrate on more advanced elements, including benchmarking suppliers, managing relationships and performance, governance and forecasting.

“We’re leading the way on what’s possible and then we’ll take it to the rest of the organisation.” He says the chief advantage is that current data earns procurement a seat at the table and the ability to make a proven impact once there. “It’s a stepping stone for getting involved early in discussions, to save money.” While it doesn’t yet have predictive analytics, pilots are underway to better forecast spend. One is examining investment plans to identify common needs across its collection of independent divisions. “While everyone might agree it makes sense to share information and cooperate, it’s not really in anyone’s scope to do so. Having a cross-business unit that sees all these things just makes sense.”



For at least the past five years, Australia’s New South Wales government has collected monthly spend data from more than 100 entities across all 10 clusters of its principal departments, covering around AUS$30bn spend a year. It consolidates it into a central data warehouse and uses the data to provide reports that help the government make better procurement decisions.

It also provides insights to assist with compliance, reduce consumption and manage tail spend. It can help track and manage spend with SMEs, Aboriginal and disability suppliers, and has the capability to project government spend based on past trends.

Acting executive director Charles Jobson says benefits include:

- Identifying spend categories to prioritise activities, such as contract compliance;

- Segmenting and identifying key suppliers to partner for development;

- Identifying opportunities such as enabling spend diagnostics;

- Understanding current supplier profiles and spend baseline prior to sourcing; and

- Benchmarking spend across government agencies.

It also recently introduced artificial intelligence for categorising spend data and is looking into using it for predictive analytics and other applications.


10 steps to introducing pure gold analytics

City of Westminster, London, SW1P 4DR
£62,519 - £70,859
Department for Transport
Milton Keynes, Buckinghamshire
up to £36,000
Niftylift Ltd
CIPS Knowledge
Find out more with CIPS Knowledge:
  • best practice insights
  • guidance
  • tools and templates