May 19, 2020

Procuring fast answers on everything from steel tariffs to the Coronavirus

AI
Procurement
covid-19
Johannes Ritter, Director, Bus...
4 min
Procuring fast answers on everything from steel tariffs to the Coronavirus

Johannes Ritter, Director, Business Value Consulting EMEA, ThoughtSpot, looks at the current status of procurement in these uncertain times. 

Whether it’s about something as routine as supplier costs for a specific component or assessing Coronavirus risk exposure in an extended supply chain, procurement needs answers - and fast. In everything from negotiations to crises, time is of the essence. The sooner you are armed with reliable facts, the more attractive (and profitable) options are available to you.

However getting answers to even the simplest questions often takes several days, sometimes weeks. By the time a data analyst produces a static report, the answers in it are often out of date or need more clarification. Since the reports aren’t showing live, native data, there’s no way procurement can use them to scenario plan and make optimal decisions. 

The slow pace of data sourcing is baffling to business people who rely on procurement. Surely ‘the system’ should be able to easily cough up basic information? But for most companies, especially large enterprises, this isn’t the case. Supplier contracts and their data are often inherited through mergers and acquisitions and never get integrated properly. Information is stored in multiple, often incompatible systems, in different structures with inconsistent labels. People come and go. Systems come and go. All of which makes getting simple data answers very difficult and slow. 

Ironically it’s the stealth fragmentation of systems and data that lead to the waste, inefficiency and excess costs that procurement people are paid to investigate. And it’s these problems that make it so hard for those same people to track down answers and negotiate better deals. Most companies employ only a small team of data analysts with the specialist skills to extract, aggregate and build reports on this information for a huge number of people across the whole business, not just procurement.

This problem gets exponentially worse when data questions are complex, or lead to more questions: “Are we utilising our logistics assets appropriately? What is the current and historical time and capacity utilisation for every vessel, aircraft, truck, and railcar? What are the trends? Are all of those assets available when we need them per the contract?” This example came from a real list of 55 questions that a large oil company’s procurement operation struggles to answer.

Changing the game with AI and search 

Now let’s go to a happier place: what if you could type any procurement question - or series of questions - into a search engine like Google, using natural language, and get answers instantly - even in the midst of a critical negotiation? 

Sounds futuristic, but this is exactly where technology is heading to make procurement more effective. Another large, conservative oil industry company that operates in more than 120 countries, is benefiting from these technologies today. After several mergers and acquisitions, the company was running 20 ERP systems and had massive data volumes with more than 220 million records. This scale and complexity made it impossible for procurement to gain insights fast enough to make optimal, profitable decisions.

Today 250 procurement people in 38 countries are using search and AI-driven analytics to get answers to a staggering 20,000 ad hoc questions every month. That’s an average of 80 questions per person, per month - a volume that previously would have seemed completely unthinkable. 

The system is providing insights in countless ways, for example, it can now identify bottlenecks that led to capacity problems during the ramp-up in oil production at the beginning of 2019. It can detect problems at plants or site locations that weren’t previously visible so that manufacturing processes can be optimised and the right amount of stock is manufactured. The system also exposes the impact of external market factors, like Trump’s trade tariffs. These affected oil production costs because steel is used to develop pipe, valves and fittings.

The time the system saves is equivalent to eleven full-time employees. This frees up procurement capacity to focus on more difficult and high-stakes projects like mitigating supply chain risk associated with the Coronavirus - a crisis that affects supply and demand. The International Energy Agency’s (IEA) recently cut its growth forecast for oil to the lowest levels since 2011. This means procurement and supply chain operations must plan collaboratively, using reliable, timely information to make the best decisions for the company.

SEE ALSO:

While search provides easy user experience and fast answers, AI flags those all-important trends and outliers. AI’s ability to trawl through vast datasets and spot patterns in vast stores of historical sales data, for example, makes it possible to improve forecast accuracy and reduce wasteful spend on products that could become obsolete. AI’s ability to spot outliers can pinpoint the regions where, for example, demand has reduced significantly.

Whether it’s routine procurement questions, trade tariffs or the Coronavirus, companies today have to respond quickly and optimally to whatever comes their way. Fortunately, AI and search technologies are finally providing the fast answers procurement needs.

For more information on business topics in Africa, please take a look at the latest edition of Business Chief Africa.

Follow Business Chief on LinkedIn and Twitter.

Share article