The Future of Spend Analysis

In this post, we welcome back Eric Strovink of BIQ.

At the incessant prodding of Michael Lamoureux, here are some prognostications for Spend Analysis in 2008.

  1. The distinction between data-warehouse-based spend analysis tools and Business Intelligence tools will erode, as it is already doing. The only requirements on the BI side are a mapping engine that can handle spend transactions reasonably, and an improved ability to import data from sources other than the ERP system. A third party may well step up to the plate and offer these (and perhaps other) add-ons for a particular BI platform, creating instant (and possibly fatal) competition for existing spend data warehouse suppliers.

    The consequences of this erosion will be more decisions in 2008 to use existing BI tools within the enterprise for spend visibility, rather than to acquire procurement-centric tools from an e-sourcing vendor. These decisions may well be made by IT or by Finance, perhaps overruling or ignoring Procurement completely. This erosion will mark the beginning of the end for big-ticket data warehouse spend analysis implementations.

  2. The distinction between spend analysis tools and spend data warehouse tools will widen. It is impossible to perform ad hoc analysis with a fixed-schema data warehouse, especially in a space like Procurement where many different views of the data are required before useful insight can be obtained. BI solutions to spend analysis will suffer from this problem as well.

    This will mean additional Procurement business in 2008 for data analysis vendors such as SAS. However, the real keys to capturing the spend analysis business will be the flexibility of the analysis tool, its approachability by business users, and its ability to quickly create and manipulate large datasets. Data analysis providers (as well as decision optimization providers) will remain niche players until business users can operate their products without assistance.

  3. OLAP technology is awkward, and presents many limitations for spend analysis. This will become increasingly apparent as more general purpose datasets (datasets which may seem simple, but whose organization is far more complex than mere A/P spend cubes) are considered by spend analysts. Spend analysis providers will need to provide simpler and easier ways to build datasets that overcome (or shield) business users from the limitations of the underlying OLAP data organization.

    Thus, we will see in 2008 the beginnings of an ability to cross-link loosely-related datasets, to build and control those linkages quickly and easily, and to create cross-dataset analyses and reports from a unified perspective. Although these advanced capabilities are already present from enterprise database vendors (and are featured prominently in their marketing literature), from a spend analysis perspective they are just laboratory curiosities given the huge effort, expertise, and expense required to set them up and maintain them. The breakthrough for spend analysis will come when ordinary business users with no IT skills can build and explore disparate and loosely-coupled datasets just as easily as they can link spreadsheets together.

  4. Invoice analysis and commmodity-specific analysis, such as that being performed by The Buying Triangle and Opera Solutions will dominate the “what’s new” frontier for spend analysis practitioners during 2008. Invoice analysis carries with it the promise of immediate refunds or other consideration from incumbent suppliers, and the consequent ability to fund entire spend management efforts through careful analysis of past contract compliance.

    Invoice analysis means negotiating with incumbent suppliers from a position of knowledge and strength, while the relationship is good, rather than dismissing the supplier outright, or blind-siding the supplier with an RFP. Even if the ultimate decision is to go to the open market, or to dismiss the supplier at the end of the current contract, there is money on the table to be recovered now. There is no reason not to go after it.

Leave a Reply

Your email address will not be published. Required fields are marked *

*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>