Skip to main content

Procurement Analytics Conferences Feature Emerging of Cognitive Procurement Technologies

I had the opportunity to attend a number of exciting events on procurement analytics in the last month.   First, I attended a CAPS Research event in RTP on Procurement Analytics, held right down the road at IBM headquarters.  Next, I attended a second CAPS Research Roundtable held in Tempe, AZ. Finally, I attended an event held by IBM, called Empower2016, in Orlando, which featured a number of speakers and analysts discussing the emerging technologies being developed around Watson Cognitive Analytics. This last event was rudely cut short by the arrival of Hurricane Michael, but I nevertheless was able to stick around long enough to gain some great insights into the capabilities of IBM Procurement combining the power of Emptoris with the emerging capabilities of Watson.  I also was able to sit in on numerous user group meetings sharing how companies are using Emptoris capabilities to improve their contract management outcomes.  For instance, one presenter discussed how she was using Emptoris decision tree features to guide users to the right contractual template.  All the user had to do was answer a series of questions, and the system would guide them to the right contractual template that had been developed and recommended by legal.  At this meeting, I also had the chance to speak with a number of analysts from Gartner and other software analytical companies to better understand their opinions of what is emerging.

What is emerging is certainly impressive.  I sat in on a number of demos, and saw the capabilities emerging in a number of IBM systems for procurement that are briefly described here.


Spend analysis  Real-time spend analysis combined with visualization techniques and structured queries on contracts can provide another level of value that is currently being developed by software providers such as Coupa and Emptoris.  In a second example, Emptoris is able to provide visualization of data analytics by category, by geography, etc. to permit deeper drill-down understanding of spending across a category across the business. In a final demo,   I saw how Watson is able to query the current spend analysis in and quickly return a number of parameters of interest. For example, in response to the query of “What is our spend with Dell?” a dashboard is produced that allows “drill down” capabilities to better understand the opportunities for combining contractual requirements.

Sourcing and market analysis (Watson Buying Advisor). This approach combines IBM catalog information, and uses natural language classifiers and speech to text technology, combined with mobile-enabled and smart digital technology to create a buying assistant. This assistant allows users to input their requirements (either through visual cameras, text description, or other approaches), and the system provides a list of preferred products and suppliers. The disruption in technology is that the technology interacts with the user through natural language, mobile picture devices, unstructured text and digital imagery, combined with clarifying questions, to understand the need and channel the need to the right sources. The system can provide suggested products and services based on the description or picture input by the user, and narrow the search using a decision-tree like set of questions. The user is guided to the right product from an approved supplier at a pre-negotiated price in many cases.

Contract management   IBM Watson has also invested in programs to cognitively manage contracts across the supply base (Blue Hound). Contract technology cognitive applications seek to accelerate contract analysis by identifying clauses that link to changing market conditions, specific supplier conditions, and how this relates to the contractual agreements in place that govern such changes. There is also a need to better able to compare and contract different terms and conditions across multiple contracts within a spending category, and determine alignment of agreements across both buy-side and sell-side agreements. Contracts for an entire organization often span hundreds of millions of pages, and these are rarely read and reviewed. Cognitive computing provides the capability to be able to rapidly scan contracts using specific queries and keywords to help understand exposure, limitations, best practices, and other insights. This is made even more complicate during a merger or acquisition, when another entity’s contracts are absorbed and must be rapidly integrated into the current supply base. Currently, extracting insights from statements of work given the high volume of unstructured data is a very painful process. Cognitive technology holds the promise of being able to parse key terms in a contract, and train it to become smarter, thus building a corpus of knowledge around what represents best in class contractual terms and conditions. This could down the road to an engine that could construct a contract for specific supply situations, or an expert system that drives the right activity.

Market Intelligence   An emerging technology at IBM is the increased understanding of price movements in the market (Pricing IQ). These technologies will provide market intelligence advice, and is envisioned to be deployed across technical services and other categories of spending. The opportunity is to correlate pricing with events in the market, including election results, interest rates, natural disasters, and other issues that impact supply and demand. Technologies that correlate pricing to such macro events can help drive predictive pricing for 3, 6, or 12 months into the future. Understanding price predictions can help drive contracting and hedging strategies. This technology is also envisioned to collect and refine real-time market research events and update contracts with pricing clauses in real-time. The technology may also be able to eventually work to establish how employment statistics from government websites will impact labor availability, and eventually impact labor pay rates. For instance, the trends may dictate the need to pay above market rates in some locations to avoid high labor turnover rates, and other areas where pay rates are well in excess of reasonable market rates.

These emerging opportunities for building greater insight into spending patterns using cognitive technologies are emerging and will be available in the next 2 to 4 years.