Slow is smooth and smooth is fast say the Navy Seals, or so the legend goes. Procurement professionals tend to say the same thing: Quick decisions can’t possibly be good decisions.
But the current climate of trade wars, natural disasters and climate change have helped to shift the role of procurement professionals from one of repetitive stability to a state of constant flux.
Agility is the buzzword of the day. The Hackett Group even called it a “critical development zone for procurement teams in 2019.”
Technology: The building blocks for agility
Agile procurement is playing offense and defense. Maintaining good supplier relationships while always having a plan if sourcing needs to shift. Anticipating disruptions before they arrive. Shifting quickly and decisively when need arises. Failing fast and learning fast. For some organizations, agility represents a completely new way of thinking, and this cultural shift can’t be foisted upon a procurement team overnight.
Technology can help build the scaffolding for faster moves that are still smart and strategic. According to A.T. Kearney, most “routine procurement processes” will be automated within three to five years. Automation will still be implemented and managed by humans, which means those with the power to allocate funds need to be as discerning as ever. Much of the industry has taken the first step, David Natoff, procurement and finance operations management consultant at Blue Sphere Consulting, told Supply Chain Dive.
"Some tasks that used to take 20 minutes now you can do in under a minute."
David Natoff
Consultant, Blue Sphere Consulting
“Most organizations today, at least the large ones — all of those have put in procure to pay systems. Often those are part of their ERP solutions. What it means is that a lot of data entry just is not required today,” he said. Procure to Pay software adds digital infrastructure to the procurement process and has reduced, in theory, process time.
E-procurement systems have reached broad-based adoption in 37% of procurement organizations and 29% have adopted e-invoicing platforms, according to Hackett research.
“Some tasks that used to take 20 minutes now you can do in under a minute,” Natoff said.
Simply digitizing the process will only increase the velocity. Four emerging technologies can convert this speed into the agility the procurement function needs.
-
Machine learning
-
Natural language processing
-
Robotic process automation
-
Big data analysis
Perhaps the most integrated and robust form of tech in procurement departments today, according to experts consulted for this report, is machine learning, a form of mathematical modeling often used to build artificial intelligence (AI) applications. Users might not even know what they are interacting with it. Machine learning is a big tent of capabilities and touches each of the technologies that follow on this list.
A fundamental tool for speeding up procurement processes is machine-learning powered virtual assistants. Whether in the form of chatbots or prompts within tasks, machine learning can help procurement officers push tasks forward, catching errors and inconsistencies as it goes.
"It’s like Turbo Tax for procurement where it asks you a series of simple questions and helps predict how you’re going to respond and gets you to the point of an order faster,” Patrick Connaughton, research director at Gartner told Supply Chain Dive.
Machine learning’s capabilities extend far beyond this, but the nudges and corrections some procurement software offers have a fairly low barrier to entry in terms of behavior change and can drive big benefit by removing some of the human errors from processes. Quick adoption and fewer errors lead to faster operations — a stepping stone on the path to agility.
Natural language processing is another form of AI that can read text and decipher meaning. The procurement process is full of text-heavy documents, agreements and conditions. Reviewing supplier contracts quickly is a key task needed to get goods moving, and language processing software can reduce this time by flagging suspect clauses.
“Let’s say you’re in merger environment, and procurement suddenly has thousands of new contracts that they’ve got to figure out and manage. So they use these AI tools to quickly do the analysis to find terms that are worded differently but mean the same thing,” said Connaughton, adding that some software can even ascribe a contract a risk score.
Beyond contracts, natural language processing can spot invoice anomalies and even sniff out fraud.
"It helps when the system is able to recognize that things are out of balance and send an alert. So it might not be that you take the person out of the equation, but you give that person additional insights — additional data," said Connaughton. Those insights effectively expand the cognitive power of a procurement team without increasing headcount, while speeding up the fundamental processes to boot.
Robotic process automation (RPA) takes the insight machine learning can provide and turns it into action. While machine learning observes procurement professionals’ behavior and gleans the right way for processes or documents to play out, RPA follows pre-programmed “if/then” instructions. It’s a more formulaic and less fluid technology, but that rigidity doesn’t make it less valuable.
RPA can speed up processes of funds authorization, Natoff explained. "Unfortunately there is a lot of procrastination that goes on, and I think a lot of that goes back to the fact that many of these organizations don’t have the right tools in place," he said.
RPA can make one event or finished task automatically trigger another. If the machine learning algorithm spots a problem with a contract, RPA can send it to legal for analysis, without the need for humans to send emails.
RPA can’t fix bad process design, but it can remove variables from a troubled process, enabling managers to see where the flaws and slow-downs lie.
Currently, 30% of procurement organizations have adopted RPA in some form, according to Hackett. The research group expects adoption to more than double in the next two years.
Machine learning, natural language processing and RPA — these can all increase the velocity of procurement execution.
"At some point the goal is that we have a touchless process and we don’t have humans engaging on that process anymore."
Chris Sawchuk
Principal, The Hackett Group
In fact, Chris Sawchuk, principal and global procurement advisory group practice leader at the Hackett Group said non-strategic procurement functions may eventually be completely automated.
“At some point the goal is that we have a touchless process and we don’t have humans engaging on that process anymore,” said Sawchuk.
In that event, the work procurement professionals will be doing is decision-making. Technology can help here too, but leveraging agility to the front-end of the procurement process is a more difficult task.
Big data analysis, another application of AI and machine learning in many cases, can help procurement officers see into the future — at least small slices of it.
How will a hurricane in Texas change energy availability, and what are the alternatives? How will climate change shift soy production? How will a supplier’s plant relocation affect the buyer’s supply chain risk? How can a facility network protect against supplier risk?
"What’s hindered adoption of network optimization tools in the past is that it does require a ton of data that most companies don’t really have."
Patrick Connaughton
Research Director, Gartner
Big data analysis can take massive amounts of data from disparate sources and turn it into actionable insights based on historical outcomes in the context of hypothetical scenarios.
"Scenario modeling is not new, but it’s been very difficult to do in the past, and now what we’re doing is applying some of these newer technologies to it that make it easier," said Sawchuk. In the past, it was difficult for procurement officers to incorporate several different types of risk into one decision — what Sawchuk called "holistic risk."
Big data analysis can weigh multiple sources of risk and detect risk cascades. Then the agility advantage comes when professionals can run these scenarios at speed and make decisions that mitigate problems before they hit the balance sheet.
"It’s about getting insight and getting it faster than anyone else," Sawchuk said.
The challenge here is having the right data on hand to run scenarios. Some tech providers bring data with them, like weather or commodity data, but a supply chain would need to be fully digitized to take advantage of this capability.
"What’s hindered adoption of network optimization tools in the past is that it does require a ton of data that most companies don’t really have. Even internal data about all the locations and how inventory looks — it’s really a big task for a lot of companies to pull that together," Connaughton said.
Right-sizing implementation
Adoption of these tools is still fairly low, but procurement executives are increasingly seeing their value.
The good news is adopting most of these technologies doesn’t necessarily take a large-scale effort to reap the equivalent benefit.
"You still have your large ERP and you may still have a procure to pay tool and then supplement it with a whole bunch of new technologies that are specifically focused on smaller processes," Natoff said.
Most smaller tech vendors with more niche focus build their tools to plug into or run in the background of larger all-encompassing systems, adding capability sometimes without even being visible to the users. Adopting emerging tech in small pieces is also an easier sell to the powers that hold the purse strings since the barrier to entry in terms of cost and implementation time is so much lower.
"In some cases," Natoff said, "these technologies are costing as little as $5,000 per year to buy and can be implemented in hours, not day or weeks."
This story was first published in our weekly newsletter, Supply Chain Dive: Procurement. Sign up here.
