E-Sourcing Forum » If you are going to do a “Proof of Concept” (POC) Spend Data Classification “Test” with a Potential Supplier

If you are going to do a “Proof of Concept” (POC) Spend Data Classification “Test” with a Potential Supplier– Do it Right (Part 1)

November 30th, 2009 at 09:25am Rod True - Spend Radar

In today’s world of vendor selection for company Spend Analysis needs, the process “du jour” is to have the vendor conduct a free “Proof of Concept” (POC), Pilot, or classification test. Increasingly, this selection process is based on older standards for Spend data classification, and has become an incomplete test of a long term, successful Spend Analysis and larger “Spend Visibility solution” for your company’s Sourcing programs. Classifying data in some manner, which is an important step in ultimately finding cost savings, may be testing the size of a vendor’s pre-sales group, and may not be at all reflective of many other very important and critical capabilities that will make your company find and manage larger cost savings over time.

Some “Cons” and Missing Links

Usually a POC is done in a short time-frame; classifications are to UNSPSC, with minimal company sourcing categories and programs information.
How easy is it for a vendor to provide a good end-result, but hide difficulties and secrets, by having numerous resources put on the project (pre-sales people)? How was the classification really accomplished? The “how” will matter much more over time for your company’s on-going sourcing need.
Conducting a POC classification to UNSPSC does not test the ability to how a vendor can “roll-up” to the company’s sourcing categories, and how the classification can support the company’s strategic sourcing programs. Data classification must ultimately end up in rolled-up sourcing categories, and then be managed from there (category management – another post).
AI isn’t BI. Artificial intelligence works to a standard, like UNSPSC. But UNSPSC is being used less and less, as it is difficult to source from UNSPSC codes. If UNSPSC is not used to source, why “auto class” to it? A company needs real business intelligence specific to their sourcing programs, which of course is different from company to company, and is not conducive to AI.
Auto classification isn’t cutting it – How can “auto classification” be intelligent to a company’s sourcing categories, if every company’s taxonomy is different at a detail level?
Where in the process are you able to view the feedback and control mechanisms to refine data accuracy? You don’t want to discover problems in the months ahead, when you do more detailed sourcing projects.
You don’t really see the speed – to accomplish data classification specific to your business needs quickly and easily.
You don’t see transparency of your data – where it is, where it will be, and what happens to it.
How does this position your company for a real implementation (pre-sales people won’t be implementing)?
How does this show structure and repeatability for future refreshes of data?
Usually a POC does not incorporate new company data collection capabilities for advanced Spend Analytics, which will happen over time as the company matures.
Usually a POC does not focus on reporting, opportunity assessment, Spend measurements, and compliance capabilities.

In Part 2 I will discuss the benefits of doing a “Proof of Concept” and how to do it in the most optimal way.

Email Forward & Bookmarks

Close this Window Bookmark and Share This Page

Save to Browser Favorites / Bookmarks

Email This to a Friend

Copy HTML:

If you like this then please subscribe to the RSS Feed.

Entry Filed under: General, Optimization, Spend Analysis

2 Comments Add your own

1. Anonymous | November 30th, 2009 at 5:33 pm

Dear Sir,

I find components within your post extremely inaccurate and actually poor advice. It appears that you are purely trying to deflate existing processes within the marketplace that are working with incredible success and are supported by 3rd party analyst groups.

I understand your need to differentiate your company offering by writing reports highlighting “your view” but be careful on stating some of your claims because they are off target based on what existing spend classification leaders can deliver today.

I appreciate your views and look forward to future blogs from you on Spend Classification. I am interested in reading your thoughts on an optimal PoC. It certainly must be part and parcel to how Spend Radar works
2. Rod True | December 3rd, 2009 at 1:28 pm

My blog above points out drawbacks of which to be aware regarding the approach when doing a POC. My advice is reflecting what we have seen in the marketplace over the past year and more. We see a significant shift occurring. Companies are maturing and focusing on smarter cost savings and Spend management capabilities. This involves broader Spend classification beyond UNSPSC (categorizing directly to sourcing groups), more opportunity assessment (smarter Analytics beyond cubes), and more ongoing category management (specific BI around categories and structured refresh of Spend regularly for measurement and management). As such, the vendor selection process should embody a wider range of review than choosing a vendor primarily based on “how fast can you auto-class my Spend to UNSPSC”. We are seeing that this perspective is no longer the best means to determine if a vendor is right for your company for ongoing Spend Management. Many companies we work with try to analyze their data in the UNSPSC structure. In most cases, they quickly realize that it does not make sense for their business. Additionally, the typical POC does not test (or validate) transparency into the process…the feedback and control mechanisms for accuracy…the ability to easily change classifications to match sourcing programs/supplier groupings/etc.

My Part 2 blog (forthcoming) does not say to not do a POC, they do serve a purpose. It focuses on how best to incorporate seeing your data within a vendors application, alternative sourcing classifications, and where a POC fits within the overall selection process.

Some HTML allowed:
<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Trackback this post | Subscribe to the comments via RSS Feed