对于许多公司来说，类别管理是他们采购运营模式的基本要素。I’ve defined category management作为管理主要支出类别的供应生命周期的跨功能（和交叉企业）流程 - 在哪些战略采购中显然是一个核心过程。我的同事彼得史密斯做得很好多件系列on the topic as well that’s definitely worth reading.
类别管理通常会组织供应市场的集群，为此类管理层分配对类别管理人员的责任（从较低级别的商品经理到顶级超级类别经理 - 例如，IT / Telecom是一个很好的例子）。然后，这些管理人员将根据操作模型（例如，业务行可能会托管某些直接类别的操作模型“插入”组织层次结构。组织层次结构本身由地理，产品/服务线（专注于提供的解决方案或服务行业），功能等驱动。
But wait, it gets more complicated. Some companies that outsource large swaths of connected processes on behalf of a customer might group their categories of supply along a process dimension. Other companies might start to look at natural groupings of supply based on the application of those products/services to a certain process, product, or customer – regardless of existing named supply markets – especially when this application may be very disruptive and represent a whole new category of supply. And some advanced firms may be looking to do all of these in the quest of letting supply markets tell them where to best find natural clusters of supply that can meet demand in new ways.
We call this more data-driven and supply-driven (because, sorry Gartner, Value Networks needs to be supply-driven, too) approach “Market Informed Sourcing“并且有写了很多关于它的。It’s a way that lets you “supersize” your market baskets, but not just by working up a single spend category hierarchy (which is only one of the 10 supersizing techniques that I’ve written about in this 2-part PRO serieshereandhere), but rather to dynamically find these new clusters. Why is it that customer-facing analytics are steeped in massively multivariate analytics to tease out insights on emerging clusters and micro-segments of demand, but many procurement organizations are stuck trying to crudely apply a 30-year old 2x2 matrix to massively multi-variant supply markets. This needs to change!
The problem, though, is that the underlying data models in the “modern” packaged business applications that power procurement and supply chain processes don’t help solve the problem. They are based on relational databases that are heavily hierarchical in nature, particularly surrounding master data such as spend categories, items, contracts, suppliers, cost types, organizations, etc.
所以,考虑到潜在的维度定义spend taxonomy as I discussed above (i.e., supplier types, organization, product/service types, process, customer, etc.), how do you jam the relevant ones into a singular hierarchy? Most firms try their best to accommodate these multiple dimensions at different levels in the hierarchy, but it's only a kluge. For example, how do you model all the various types of contingent labor within the dozens or hundreds of nodes on a singular spend category taxonomy? There is no solution to a massively combinatorial problem using a single taxonomy. I do try to give some practical insights on this problem in my 2-part series on how to design a spend category taxonomyhereandhere。
有多个和与间分类有关的分类，需要带来承受。It will require much richer data models and analytics (i.e., think object modeling, metadata, big data, knowledge modeling, advanced business process management, etc.) to find these opportunities and risks that are increasingly more difficult for humans to see on their own. This is a very big topic that transcends spend categories. For “market informed procurement” – or more accurately, “market informed business” via the supply intelligence capabilities built by procurement and new classes of third parties – there will need to be a whole new class of processes and technologies that drive laser-focused, right-sized workflows that are pegged via analytics to value-producing opportunities or value-destroying risks.
We will do our best to continue to stay on the vanguard of these new capabilities, and we really do welcome your insights on this set of topics.
讨论这一点：