Procurement’s P2P Pet Peeve: Bad Spend Data

Spend Matters welcomes this post from Zycus.

During a recent Spend Matters webinar entitled, “Forget the Amazon Metaphor - P2P Needs to Be More Like a GPS,” attendees were polled regarding their greatest challenge with Procure-to-Pay (P2P). While there were a fair number of votes for oft-cited issues such as spend compliance or user adoption, what stood out is the fact that the biggest pain point actually focused on the by-product of P2P – accurate spend visibility (or lack thereof) – the biggest challenge for more than one third of webinar attendees.


The polling feedback underscores the need to focus P2P initiatives on more than the clichéd goal of providing end-users with an “Amazon” shopping experience. Of at least equal importance – what’s in it for procurement? Procurement, it seems, wants P2P to provide real insights into spending – in real time – and is frustrated with consumer-like P2P shopping UIs that engage end-users, but which don’t make them any more proficient at correctly coding and classifying their transactions as they “point and click” to fill their shopping carts, thereby thwarting procurement’s goal of leveraging accurate spend visibility from the P2P process.

Spend Matters Research Director Pierre Mitchell framed the dilemma for the procurement organization, which he says, “wants real-time spend analytics from P2P, instead of the forensic analytics they currently have to perform after the fact, to correct misclassified or miscoded orders.”  Co-presenter Scott Fitzgerald, procurement director for The Mentor Network, illustrated the challenge for his organization, where harried requisitioners, eager to complete the transaction and get back to their real jobs, often simply “use whatever GL or Commodity Code they used for their last transaction, regardless of what is being requisitioned.” The result: upwards of 20 percent of requisitions have historically been incorrectly coded or classified in P2P. Further exacerbating the problem is the volume of P-Card transactions placed outside P2P altogether – with millions in spending emanating from thousands of individual card holders – that offers “no line item visibility into the products and services purchased on the cards,” according to Fitzgerald.

In order to rectify the situation, The Mentor Network is deploying Zycus’ P2P solution that leverages Artificial Intelligence for a Guided Procurement System – a GPS of sorts – to help ensure that users not only make compliant purchases from approved, contracted vendors, but also that the transactions themselves are “compliant” – bucketed in the correct category for spend analysis purposes and applied to the correct general ledger account for financial reporting. An early proponent of guided procurement or Guided Buying, Mitchell stresses the need for the P2P system to guide users at the very outset of the process, or as he sees it, “the moment of truth” when the user is searching for what they need. “The system needs to have the intelligence to know when the user enters a search term like ‘mouse,’ are they really looking for a computer peripheral or a lab animal?” The AI-based, Guided Buying approach automatically classifies the user’s search term and suggests the most appropriate category for the search, yielding only relevant results. Where traditional keyword searching would produce results for computer products and lab animals, or any other item containing the term “mouse” in the product description, with Guided Buying the user gets only relevant “hits,” or as Fitzgerald puts it, “three results instead of 300 – just the products we want them to see.”

Guided Buying also ensures that users can efficiently complete off-catalog requisitions – for services or non-standard products – just as easily as a catalog-based requisition. Off-catalog requisitions often comprise the majority of indirect spending, and off-contract or “maverick” purchases. The Guided Buying model helps the user navigate the off-catalog requisition process, by first auto-classifying the request to the correct category and presenting the user with approved, contracted suppliers and a category-specific Guided Form. Guided Buying then prompts the user to specify their requirements and ensures the requisition is properly coded and classified and will follow a compliant process, e.g. the right category-specific workflow. In this way, Mitchell notes, “Organizations can shrink cycle time for all types of requisitions, especially by eliminating the back and forth that so often occurs between buyer and requistioner, to ferret out the user’s exact requirement.”

For his part, Fitzgerald sees big benefits accruing to his procurement organizations as a result of the Guided Buying initiative for P2P. “First, we work with the business units for purchases go through the P2P process – this will allow us to reduce the risk and inaccuracy associated with the P-Card transactions. We will still get the benefits from the card program – capturing increased rebates and eliminating invoice processing – but with greater control since these orders will get prior approval through the requisition approval workflow and orders will be charged to a centrally administered ‘ghost’ card account, instead of an individual card holder’s account.” Furthermore, Fitzgerald elaborates that the initiative is designed to “clean the data” at the source, or at the point of requisition, for all P2P transactions – P-Card or otherwise – by setting up a defined mapping from categories to the appropriate General Ledger accounts, allowing the system to automatically determine the appropriate account based on the category assigned through the AI tool. So while Fitzgerald sees the enhanced Guided Buying experience – fewer clicks and no confusing commodity or account codes to enter before check-out – as a clear benefit to his end users, the dividend for his procurement team will be reliable and real-time spend visibility at a granular, line-time level to let procurement do what it does best - find and execute on savings opportunities.

Share on Procurious

Discuss this:

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

This site uses Akismet to reduce spam. Learn how your comment data is processed.