Why Purchase Price Variance (PPV) Should Be Banished From Procurement Measurements and KPIs

One of the biggest challenges to overall program impact and improvement in all but the most advanced procurement organizations are the raw elements that many procurement organizations measure themselves against: key performance indicators (KPIs). Of these, purchase price variance (PPV) is particularly obnoxious in all but certain cases. PPV measures the difference in price paid for multiple purchases for the same SKU, part or service. It is typically employed in standard costing environment in an ERP system for SKU-based items where actual PO prices are tracked compared to the existing standard cost.

This methodology is great for the financial accounting function. The PPV can be calculated easily by the system by accumulating the PPV until the new standard is calculated (and those variances posted to the appropriate general ledger account). A favorable PPV (i.e., price is less than the standard) is also known as a purchase price reduction (PPR). This all seems straightforward for the accounting department, but it’s not a great way to judge procurement performance, at least not on its own. Why?

There are numerous reasons why PPV can be such a misleading figure. In this two-part Spend Matters Plus series, we explain why PPV is a KPI that procurement organizations should stop measuring internal and individual performance against.

The Arguments For Putting PPV in the Rubbish Bin For Good

On its own, PPV is a somewhat narrow KPI for measuring individual procurement performance. There are many reasons to put PPI in the performance measurement rubbish bin for good:

1) As the last decade of global roller coaster commodity and currency prices has shown, declining and rising cost variables outside of a buyer’s control can have a significant effect on individual part costs not reflected in the actual performance of a procurement manager. At a minimum, some type of price benchmarking to the market is needed to augment the book value-centric approach of PPV.

2) Procurement often has no control over many of the most important factors that can drive cost for a given order, especially in a manufacturing setting. For example, expedited order requirements (e.g., smaller and faster protection runs, air freight, prototype parts) can dramatically affect SKU pricing and related PPV, if such cost elements are pegged to a SKU and its related PPV calculation, which may or may not be the case. If these costs in a total landed cost calculation are not included in PPV, then TCO must by definition be measured in order to get the whole cost picture.

3) Programs based on volume commitments (order quantities and tiered/volume discounts when certain volume requirements are hit) can trigger reduced pricing that PPV alone cannot measure. Without contract-based discount schedules modeled in the system, such a scenario requires back-end reconciliation to understand whether the submitted/lowered invoice price (assuming the supplier actually submits it correctly) requires a reconciliation of the resulting invoice price variance (difference between PO cost and invoice cost), or whether the PO system needs updating. By the same token, rebate programs and rebate income (e.g., in which a supplier provides a rebate in the form of an actual cash payment or credits to the buying organization) are also not likely to be accurately reflected within reported PPV.

4) Showing negative or rising PPV may be in conflict with important programs with strategic suppliers and customers. For example, when balance of trade programs get tricky (e.g., when an organization is selling more to but purchasing less from a strategic customer), it may be in the best interest to have savings show up elsewhere on a P&L (savings reflected as rebate income) so it does not have to be reported directly in quarterly or annual reviews.

5) In the case of direct spend, direct SKU comparisons are difficult across different facilities and geographies based on individually (and locally) negotiated pricing with suppliers (even from the same parent), variations in shipping and packing requirements, and the service of suppliers (e.g., certain facilities not being able to receive in full truckload quantity, etc.). Such pricing variation is often encountered during spend analysis, and while a significant amount of savings is indeed found here, the “apples and oranges” problem that we just discussed for the same SKU really shows up. In other words, it’s not that the PPV is necessarily bad here, but it only shows variation rather than performance and causal factors relating to the performance. Only TCO analysis to analyze other cost elements (transportation, taxes/duties/tariffs, storage, payment terms, currency, cost of lead time, etc.) can really identify all of the variations happening and how best to find opportunities for each cost element. So, even if you think PPV is necessary (to post to G/L accounts), it is by no means sufficient to judge performance and improvement opportunities.

The bottom line is that PPV fails to measure true lifecycle costs. For example, as a supplier, I might offer a price concession that sounds outstanding on paper but then I bury additional costs in other clever ways. These could include higher costs for non-standard orders/delivery, related services, support, and so forth. Or they could result in fees charged by others. For example, suppliers have told Spend Matters that supplier network fees that certain providers of discounted e-procurement toolset pricing charge for the buying organization often then bring additional and hidden charges or higher costs by the suppliers transacting through the e-procurement system on the sell side.

6) New supplier, inventory, supply chain finance and working capital management programs (e.g., vendor managed inventory) can affect unit cost in ways that do not clearly reflect the total cost (or benefits) of purchasing a SKU, part, or service at a given unit price. In this case, even if you can track the effects of different payment terms, you may choose to offer dynamic discounts to the supplier directly, or you may use a third party in a supply chain finance program to do so on your behalf, and then you receive a rebate payment from the provider eventually. Good luck posting that back to your SKUs to include in a “holistic PPV” (an oxymoron) calculation. If you are treating the rebate as revenue, then it’s pretty much all out the window anyway.

7) You might ask “why treat the rebate as revenue?” rather than just drop the benefits to the bottom line. Is it just overzealous conservative accountants? Not necessarily. In certain industries, it can be in procurement’s interest to show savings (or revenue) elsewhere than in direct PPV given the need to share “nearly” full transparent cost information with customers. An organization might, for instance, show a 1% favorable (downward) PPV to customers based on annual cost reduction commitments, but have achieved a 3% total cost reduction through optimizing for other cost areas in the trading relationship that do not have to be disclosed with the ultimate customer. For example, a newly introduced scrap reclamation program or multitier raw material demand aggregation, buy/sell or trading company models can hide potential savings and revenue from actual PPV that must be shared with customers.

8) Smart but lazy procurement managers are expert at gaming PPV. They are able to bury unavoidably rising costs through the clever gaming of the system (or attempting to dismiss PPV as a metric when it does not work in their favor for some of the above reasons we’ve covered). In the best of cases, this can come from changing SKU numbers and part ordering requirements. In the worst, it can come from back-door dealings with suppliers that lead to changes in material specifications/usage, production processes, or the like in a hidden manner. A buyer can ask a supplier to submit yearly 2% price reductions for the SKUs, but then bury other costs in other service charges not captured in the cost accounting system at the SKU level, which are pegged to the buyer.

9) Another example that is less nefarious than the previous one is the implicit incentive not to lower component costs (and total lifecycle costs) at the time of design. If best practices are followed, and all the costs are taken out during conceptual design (or even detailed design), then there is no excess cost left for the poor buyer seeking favorable PPV performance. So, the buyer actually has an incentive to specify higher cost components and then drive downstream engineering changes to reduce material cost savings (e.g., consider the GM example of safety ignition switches here).

10) PPV does not accommodate the variation inherent in your commodity strategy itself. For example, if you are exposed to highly volatile commodity prices (e.g., metals, food, energy/hydrocarbons), then you may choose to “beat the market” or “smooth the market.” You may also do this based on your similar strategies with your customers. Depending on your strategy, your PPV performance will obviously vary. Not only that, but trying to impose a standard costing methodology on top of volatile commodity purchases can create a huge amount of back office reconciliation effort that costs real money that will never get pegged back to the item PPV.

The above discussion has been centered squarely on direct materials purchases. But many practitioners realize that these problems are also relevant for indirect and services spend. Much of the rich data gained from invoice-level information does not make its way into a PO (the information that’s most commonly used in spend analysis efforts that track PPV). The result: aggregate or incomplete information used to determine PPV. And in certain cases, the additional manual homework required to hunt for the data to enrich PO information can lead to the introduction of data that leads to discrepancies from manual entry or measurement variations and approaches. We discussed the issue of non-price cost factors in No. 5 above, but in indirect, particularly for manufacturing plant-based MRO, this is a special type of hell reserved for activity-based costing staff that hopelessly try to allocate the hidden costs buried in big cost pools back — just to the product-line level.

This list is clearly just a start. We could keep going (and we’re sure you could add to it, as well). There’s also the most critical factor of turning to better metrics for gauging procurement performance based on measuring total cost associated with a given purchase including logistics, inventory, taxes/duties/customs and other factors. This topic goes beyond this discussion, but there are many good information sources, consultants and tools that can help in building out a competency in total cost management. It’s not easy, but it’s essential as one competency area required to collaborate with suppliers to reduce TCO. Major automotive OEMs like Honda and Toyota (and some U.S. and European firms) try their best to use highly granular cost modeling/tracking in conjunction with their suppliers to manage “exquisitely” detailed causal analysis of cost drivers to collaboratively reduce them. Simplistic PPV scorecards that seem to be part and parcel of “old school” supplier relationships obviously won’t cut it here.

So, it’s time to bury PPV, right? Not necessarily. As we mentioned before, PPV merely shows variability. It is not good or bad in its own right. Still, understanding variation is a great way to identify savings opportunities. In Part 2, we will write about such a case.