The new “cognitive” buzzword is getting a lot of people interested in modern Sourcing and Procurement technology, and that’s a good thing, except when it isn’t. (How can it now be? Not all providers truly offer cognitive capabilities, not all are equal among those that do, and not all are right for your organization.)

And unless you truly understand what cognitive sourcing can do, when it should be used, what technologies you need to power it, and how to properly apply it, the answer is no cognitive sourcing is right for you.

When it comes to sourcing, a sourcing solution must meet a number of requirements in order for it to be considered cognitive. It must be capable of:

• supporting advanced cost models
  to allow for an accurate determination of should cost
• supporting sophisticated automated data collection to populate those models from market indices, statistics bureaus, public (government) data repositories, etc.
• supporting a large repository of trend analysis algorithms
  to help an organization understand market dynamics
• support sophisticated analytics
  to help organizations slice, dice, and compare all the insights extracted by the cognitive platform
• support advanced optimization
  to analyze the cost models and all the supply and logistics options available subject to business constraints

If you look at each of these requirements in comparison to an average Procurement organization with some semi-modern Supply Management technology

• they have some cost modelling capability in their ERP
• they have some automated data collection around risk and commodity costs through providers like D&B and Ecovadis and Market Index data providers
• they have some familiarity with trend analysis in their inventory management systems
• they have adopted a spend analytics platform, which may be a generation behind, but still gives them some cost insights
• but they have no decision optimization at all

So if you really want to get cognitive, get optimized. Without a good understanding of what optimization can do, and how to use it, how do you expect to figure out when to apply, and not to apply, cognitive sourcing technology properly.