3 Companies Integrating Artificial Intelligence into Procurement

Artificial Intelligence (AI) and machine learning (ML) are fast evolving into every sector of our economy. So much so that referencing them in the tech space is fast becoming akin to a buzzword drinking game. “Okay everyone takes a swig when we hear AI used in a sentence! Take two when you hear machine learning and AI together!”

Many of us know about the range of uses for this technology from how we pick hotels online to listening to our kids talk to little boxes in our living room. “Alexa play the Narwhal song!”. AI and all of its variants are having a progressively bigger impact on our lives but has AI reached the world of buying and procurement?

According to Pierre Mitchell at SpendMatters: “Much of the problem with AI is that it’s misunderstood as a technology area that’s somehow different and separate from day-to-day processes and technologies.” Ultimately many of the changes that are happening today are not a deep combination of machine learning and neural networks, they are a combination of relatively simple algorithms that use a large data set. These algorithms, however, can have a huge impact on procurement processes.

Some very big players are doubling down on AI bets that they made in the procurement space over the past few years and these bets are really starting to pay off. Let’s take a look at how three companies are starting to make an impact on using AI in the procurement space.

The SAP Ariba vs complex decisions and cognitive bias using a Intelligent Digital Assistant

Back in 1999 (remember Y2K?) Ariba was founded as a way to move procurement from the age of paper and manual processes into the digital age. Ariba was one of the first B2B internet companies to go public and it made several big moves over a tough decade to rise to prominence as a procurement powerhouse. In 2012 Ariba was purchased by SAP and now provides a full range of cloud-based procurement, spend management and supply chain services. In 2017 SAP Ariba announced plans to create an intelligent digital assistant designed to help with everything from risk assessment to product sourcing.

“Intelligent bots are facilitating a whole new paradigm of interactions that allow people to be more efficient and smarter in what they do,” said Dinesh Shahane, (former) Chief Technology Officer, SAP Ariba. “We are investing in these types of applications and technologies such as artificial intelligence and machine learning to deliver a next-generation user experience for our customers that promises to transform procurement as we know it.”

The core driver for SAP is to build a more efficient set of processes based off of tasks that human beings are not as good at as a digital assistant would be. For example, people are not good at calculating all the variables that are altered by a single decision. Another area where humans struggle is with bias.

So ultimately SAP Ariba is aiming to take procurement into a world where procurement officers and people who work in purchasing and supplier management have a new set of tools that can augment their decisions based on a treasure trove of underlying data.

Suplari vs risks and cost inefficiencies using aggregated data

Suplari is a much newer enterprise in the procurement space founded in 2016 by Nikesh Parekh. Suplari was founded to help enterprises manage their suppliers and costs using machine learning and AI. The idea is that by aggregating data across different systems and then applying machine learning an organization can drive cost benefits and efficiencies, extend the reach of procurements influence and increase visibility and introduce new capabilities or tools into the procurement teams fold.

“All this data sitting within a company largely goes unanalyzed,” said Suplari CEO Nikesh Parekh.

AI can help provide insights into pricing irregularities, usage anomalies, contract variance, and suspicious spending or even fraud. Other menial tasks like data clean-up, data duplication and alerting and reporting tasks can greatly speed up procurement while lowering costs.

Another area where AI can help is with bandwidth and prioritization. AI can help to rank the different procurement projects through a lens of prioritization based on a lot more data. This helps to streamline which projects a CPO tackles and in what order. This is all done by having a system that can aggregate data and assess impacts, prioritize and then learn from outcomes and adapt.

Lastly, Suplari is aiming to tackle new capabilities that were not even on the radar a few years ago. How about gauging supply chain risk? Or understanding historical purchasing and growth to make predictive insights on future need? How about using that data for negotiation of better pricing? Even though these are areas that are not fully baked you can see the potential for CPOs to create a much greater impact for procurement.

Coupa vs inaccuracy and compliance using AI Classification

Coupa among other things is focused on what companies are doing on the spending front and how they can help you find efficiencies using various solutions including AI. This is obviously a huge field that could provide incredible opportunities on the purchasing front. I think most of us can draw from direct personal experience to know what a huge challenge this is.

I worked for five years at a major university as an IT Director and we had no way (at least a few years ago) to connect what anyone was doing below a certain level of spending. That meant that likely there were tens, hundreds and sometimes thousands of licenses for things that could have been far cheaper if purchased in bulk. Think of how many professors have their own grants and budgets. New Adobe Photoshop license for a new paper? New Macbook pro for online teaching? New Canon camera for research? How much hardware, software and other things are bought with almost total independence from a central purchasing strategy?

The challenge, of course, is that you need to capture all of the data and then have a system that knows what to do with that data. Fundamentally, you need to have a system that tracks what is happening and can make predictive calculations about how to meet the needs of employees and business goals while also ensuring efficiencies at the same time. Ideally, that system also tracks what hundreds or thousands of other company spend patterns look like and it learns from them then cycles that back into recommendations for you. That is what Coupa is putting AI to work doing. They have been so successful that in 2018 they were named to the Deloitte Fast 500 list which is a list of the 500 fastest-growing tech companies. “These companies are innovators who have converted their disruptive ideas into products, services, and experiences that can captivate new customers and drive remarkable growth.” said Sandy Shirai, vice chairman, Deloitte LLP

So where does that leave purchasing and procurement today?

The Global CPO from Deloitte put it bluntly when talking about how many procurement teams in big companies are using AI today. He said “The application of predictive and cognitive analytics is almost non-existent.” So we are definitely not out of the starting gate yet in any huge numbers. This is likely driven by a couple of reasons:

A major barrier to implementing AI is simply the costs. To run a platform like SAP Ariba, Supplari or Coupa you need some serious pocket change. You also need to have a large team and a lot of suppliers and a lot of data. You then need time to integrate all of it. That’s great if you are Airbus but not if you are a 100 person tech company buying SaaS.

Most of these AI solutions today are really aimed at large enterprises because they are complex and they tackle a lot of the more complex aspects of procurement like risk management, supply chain management, and big data analysis.

Another major barrier for most of us is education. We simply don’t have time to sit down and learn about what AI solutions might be relevant to our business. We are all busy running our daily lives.

Ultimately we can compare AI today to computers in the 1970s. It is still very much in its infancy and it is being used most effectively by companies that have big problems and even bigger bank accounts. This creates a huge realm of opportunity however as the technology matures and solutions become available at both ends of the scale. It is close to becoming accessible to companies of all sizes. So for SMBs there will shortly be a raft of SaaS companies that will be using AI and Machine Learning to tackle a lot of business challenges.

So are we really that far from asking our AI purchasing system what risks it sees with specific supplier … while we make french toast? We are much closer than you think.

Thanks for reading.

Phil Chatterton is the Co-founder and President at Warmcall, Inc. a screening bot company for purchasing (warmcall.com). If you are interested in learning more about warmcall you can watch a quick video here.