Daily Archives: September 19, 2017

How Do You Identify The Day After Tomorrow’s Supply Chain Paupers?

Well, assuming the day after tomorrow comes and they are still around the day after tomorrow, they will be easy to spot. Not only will they still be trying to use Excel, but they will still be using Excel and will only recently have started exchanging documents using XML, using last decade’s e-Procurement technology.

They will not have advanced to modern e-Procurement applications, yet alone modern sourcing or supply chain visibility solutions. They will be in the process of simply moving from paper to e-Paper, trying to still conduct RFIs through e-mail with Excel (and just uploading the results to the first generation decade(s)-old e-Procurement solution), and generally trying to keep their outdated procurement processes in tact.

However, as we now know, first generation procurement and sourcing, focused primarily on e-document exchange, simple RFXs, the odd auction, and basic reporting is not enough. You need modern e-catalog management for procurement spot buys, analytics for opportunity identification, optimization for at least TCO management (if not TVM), and SRM for supplier information, relationship, and performance management.

But this is not enough. These day’s, there’s never enough time to sift through all the data to identify the opportunities, never enough time to collect enough market data to qualify even the ones you have identified, and certainly never enough time to construct category specific models on even a fraction of those to determine if they opportunities will be realized with an appropriate sourcing event — which can take years of experience to properly identify.

You need a next generation solution that can automatically collect, maintain over time, and trend market pricing data; run all your data through multiple types of automatic analysis and compare your spend against historical spend and market data (and look for variances); pull out the categories with opportunities; run trending algorithms to project your demand against expected contract prices based upon projected market demand, supply / demand (im)balance, and economic factors; calculate the potential savings if nothing was done; use historical data and automated reasoning (enriched with context) to (probabilistically) identify the best sourcing or procurement strategy; and then use appropriate workflow automation to automate as much of the event as possible (and if it is a spot-buy under a threshold, automatically procure from a catalog, an approved supplier under contract, or a three-bids-and-a-buy RFQ against approved suppliers).

In modern terms, the next generation solutions will be Cognitive Sourcing or Cognitive Procurement solutions. While they are not true artificial intelligence, with enough data and great models, you don’t need true AI to automate acquisitions where there is no strategic value and no significant value to investing human time. Good examples are office suppliers, janitorial services, and sometimes even laptops. Yes, replacing laptops across a large office can be in the millions, but laptops against a standard config are commodity. Just do an automated auction [with ceilings and floors] against a set of approved suppliers and let the most aggressive supplier win.