Mostly Harmless, Part XVIII

Previous Post

In the last post, the analysis process was discussed and some of the basic questions were defined. This post will address the some of the associated challenges of the analysis process, some associated best practices, and the benefits that could be expected from an appropriate e-Procurement solution.

Common Challenges

The literature will claim that there are many challenges with regards to spend analysis, and that these challenges will revolve around data cleansing, data classification, and data enrichment, but the reality is that there are generally only two real challenges: access and analysis.

• Data Access

  In order to do data analysis, an analyst needs access to the majority of the relevant organizational data that represents at least 90% of the spend in the categories that the buyer wishes to analyze. Without a centralized e-Procurement system, this can be difficult as some spend will be in the ERP/MRP, some in the accounting systems, some in the P-Card system, and some in various departmental systems.

• Insightful Analysis

  Top N reports are not analysis. Spending trends are not analysis. Automated reports are not analysis. Analysis is the ability to slice and dice the data eight ways from Sunday to allow for the identification of unusual spending patterns that represent true savings opportunity.

Best Practices

• Force Every Purchase Through the e-Procurement System

  Forcing every purchase through the system not only significantly reduces maverick spend, but it provides a centralized repository of all transactions which provides a solid foundation for spend analysis.

• Force Every Payment Through the e-Procurement System

  If the e-Procurement system does not support e-Payments, insure that it supports a record of payment against each invoice and that all payments are loaded into the system and cross-referenced with associated invoices, goods receipts, and purchase orders. This is necessary to take spend analysis to the next level. While many vendors will claim that only AP data is needed for spend analysis, the reality is that AP, invoice, and purchase order data is needed for spend analysis. Without the m-way analysis, it is impossible to tell if overpayments, which could be recovered, were made. Without the m-way analysis, it is impossible to identify all maverick spend, which is the first step in reducing future maverick spend. Etc.

• Do Ad-Hoc Analysis Whenever There is a Possibility for Savings

  Traditionally, data analysis was avoided because the cost of analysis was high relative to the savings potential. However, with the right tool, the cost of an ad-hoc analysis is no longer the five or six figures it used to cost, it’s now three or four figures — which makes even a maximum five figure savings opportunity worth analyzing, especially if a preliminary analysis can be done in an hour or two! Even if only one in ten hunches pays off, if it only costs $1,000 of an analysts time to do an analysis, and the one pay off is $100,000, that’s a 10X ROI.

Potential Benefits

• Savings

  Real analysis will identify overpayments that will lead to immediate savings. Real analysis will identify maverick spending, which will lead to more savings when it is prevented. Real analysis will uncover new savings opportunities, which will lead to more savings. Real analysis is savings.

• Better Procurement

  Finding and eliminating maverick spend through better processes leads to better procurement. Understanding spending patterns leads to better procurement. Saving money leads to better procurement.

Once the analysis is complete, it is time to review and update the catalogs and contracts, which is the subject of the next post.

Next Post: Catalogs & Contracts, Part I

Share This on Linked In

Mostly Harmless, Part XVII

Previous Post

Analysis is often defined as the process of inspecting, cleaning, transforming, and modelling data with the goal of highlighting useful information, suggesting conclusions, and supporting decision making. However, this misses the point — analysis is all about looking into data to get insight. Nothing more, nothing less. Everything else is a pre-cursor, post-decision, or just not relevant. It’s the Analysis, Stupid.

In e-Procurement, it’s all about what did the organization do, what did the organization not do, and what should the organization have done. Of course, this is a much harder question than it may seem to be on the surface. At a basic level, did the organizational buyers follow the processes? Were the purchases on contract? Did AP pay the correct amounts? Were taxes computed properly? Were owed moneys reclaimed?

However, these questions are just the starting point, not the end report. For on-contract purchases, were prices consistent? This is very important in office electronics purchases where rates are a percentage off of “best-price”. Considering electronics tend to depreciate about 2% to 3% month-over-month, over a year, prices should decrease. However, when many organizations plot pricing for a standardized computer purchase over the course of a year, they will find that prices stay relatively flat. For off-contract purchases, were prices reasonable compared to market index? How much did maverick spend cost the organization? Was any of the maverick spend justified? Did AP pay the approved (adjusted) amounts?

But the questions shouldn’t stop there. Were the spending patterns consistent? If spend on a particular category abruptly rises in a department where it was consistent year-over-year for the last three years, something is amiss. If claims for a particular employee are the same month over month when they usually follow a sinusoidal curve, something is wrong. If off-contract spend increases rapidly, something is very wrong.

Thus, directly or indirectly, a good e-Procurement system will support in-depth analysis of the process, spend, and results. This means that if the platform does not contain a true data analysis tool, it should support full export of all relevant data in a standard format that can be imported by such a tool for analysis. This is because, when all is said and done, the biggest savings will come from improving the procurement process itself. This requires insight, or lessons will never be learned.

Next Post: Analysis, Part I

Share This on Linked In

I was very disappointed after reading this recent article on “process matters too” in the CPO Agenda which discussed the recent findings of The Hackett Group with respect to P2P transactional channels as well as their recommendations for procurement organizations wanting to improve overall source-to-settle performance. If the article is to be believed, Hackett has fallen for the classification trap.

According to the article, Hackett says that a company must:

1. Possess a unified spend category taxonomy.
2. Define a rationalized set of transactional purchasing and payment processes that are then explicitly mapped to spend categories and/or associated suppliers.
3. Ensure that individual P2P transactional channels balance cash, cost and stakeholder satisfaction.
4. Integrate a channel strategy selection and implementation plan into the category management process.

In reality:

1. A single taxonomy is not enough as each business unit will need its own in order to be effective. Moreover, taxonomy is irrelevant. The only thing that is important is that the spend is captured and available for analysis. Every department and user will want to see the data rolled up differently. This is the classification trap, and those who fall into it never advance to real data analysis, which is where true savings are discovered.
2. While the company must define an accepted set of purchasing and payment processes, and while spend must be associated with the appropriate suppliers, the mapping should not be made to an explicit fixed category. Assignments must be able to change as needs change (spend by supplier, spend by commodity, spend by category).
3. Channels must balance cost and stakeholder satisfaction, but the amount of cash flowing through is not relevant. If the cost of maintaining the channel is too high (relative to the value), the channel must be abandoned.
4. This is good advice. Planning greatly increases the chance of success.

So, if you really want P2P success:

• Come up with a channel plan.
• Implement the appropriate channels and insure all spend goes through an approved channel.
• Make sure all of the spend data is accessible from each channel.
• Analyze the data in a true data analysis tool to determine which channels are performing well, which aren’t, and adjust the plan over time as necessary, and
• allow each business unit to use their own taxonomy.

Share This on Linked In