A better procurement incentive model

Discouraging hedging on purchasing performance targets may take convincing, but in the end, everyone wins.

Motivating procurement teams ought to be straightforward: Set the right targets and incentives, and then conduct a fair evaluation of buyer performance. Nevertheless, many procurement managers feel they have yet to find the best way to implement these processes. In practice, most conventional incentive schemes inadvertently encourage managers to deliberately underestimate the savings potential in their categories in order to ensure positive performance ratings. This “sandbagging” has a number of unhelpful effects: It reduces the overall effectiveness of the procurement process, makes budgeting difficult and can lead to antagonistic relationships between top management and category teams.

As we will argue in this article, CPOs don’t have to put up with these suboptimal conditions. A target-setting methodology originally developed in the command economies of the Soviet Union can motivate purchasing managers to set the most realistic targets they can, while still encouraging them to beat those targets if they are able.

Target setting in practice and the challenge of sandbagging

In an ideal world, companies would have a clear picture of true cost versus price. Procurement would have a deep understanding of supplier costs, gained through close relationships and the use of clean-sheet analysis. Knowing this, they would use predictable demand and high bargaining power to secure the best possible price. In the automotive industry, for example, purchasing teams use complex models to create cost benchmarks, then work with suppliers to establish the “gap to benchmark.” Closing that gap, or at least a portion of it in a given timeframe, becomes the target for both the buyer and the supplier.

In most situations, however, this ideal approach is not possible. The spend in a given category may not be high enough to justify large-scale data collection and analysis. Or the company doing the buying may not have sufficient bargaining power to get suppliers to make their cost calculations transparent, or to agree to savings and other improvements based on the identified gaps.

Where full market transparency is not available, we see a spectrum of approaches to target setting, often depending on company culture. One common approach is for management to ask the responsible category leads for their “bottom-up” estimates of how much they believe they can save (and the savings figure may also be negative). Yet when an individual’s personal performance rating or financial bonus depends on the figure he or she states, the temptation to hedge is high. In a fully bottom-up process, many people would probably try to minimize their risk of falling short of the target. They would thus offer a relatively “safe” or low savings estimate, wrapped in various reasons why a higher target is just not possible.

At the other end of the spectrum, management sets savings targets purely from the top down, based on top managers’ perception of what savings are necessary and possible. By excluding employees with the highest category expertise, however, this approach runs the risk of demotivating them with targets perceived as unfair—especially if a flat target is applied across categories. If category leads think the targets are unfairly high, they may not be motivated to try and meet them, while if they believe the actual savings potential is higher than the target, they have little incentive to make this known to management.

In practice, many companies try to combine the top-down approach with bottom-up verification. While this is certainly good for participation or to stretch aspirations a little, the basic problem remains the same. By making top management believe in lower savings levels, category leads increase their own chances of meeting or even outperforming their targets and thus getting their year-end bonus. We have even observed the perception among top management that overshooting a less aspirational target is superior to setting truly aspirational targets in the first place.

This sandbagging of savings potential leads to various negative effects on the procurement department. Initially, it creates high uncertainty among the company's senior management about the real savings potential in the upcoming period. This uncertainty leads to distorted incentives, and prevents companies from successfully forecasting and planning upcoming budgets and savings.

To correct this effect, some companies resort to the “additional target,” suddenly set by management in the middle of the fiscal year—either because the procurement improvements so far are not enough to compete in the market, or to unlock additional reserves that were ”hidden” during the initial target setting process. This strategy has a number of serious shortcomings, however. It is likely to make the organization even more cautious about target setting in future periods, for example, and can force purchasing teams to focus on short-term commercial topics at the expense of measures that may lead to larger or more sustainable long-term savings.

A different way of incentivizing your organization

What could a better incentive model look like? We would like to propose a modified version of the Weitzman scheme.1 This is an incentive system originally developed in the USSR and reported on by MIT professor Martin Weitzman as a method to reduce uncertainty in the central planning forecasts of state-run companies and improve target estimates. Its inauspicious origin might be one reason why the approach has had little or no uptake elsewhere. However, we believe that adapting the Weitzman scheme to the procurement world offers a valuable alternative to current practices. The scheme stands a good chance of producing results superior to today's target-setting because it creates incentives for buyers both to communicate an accurate but aspirational savings target from the start, and to go the “extra mile” and surpass these self-set targets.

This “truth in target-setting” incentive model can be executed in two simple steps, which most of today's approaches also have in rudimentary form:

Step 1, at period start, the category lead defines the savings he or she believes are achievable in their area of responsibility. This is set as their performance target, and a share of these savings (percentage a) is used to define their incentive. For simplicity here, we will define this as the base bonus that will be the starting point of the lead's bonus calculation at the end of the period.2

Step 2, at period end, the base bonus is adjusted to reflect the lead's over- or under-performance relative to the target set in step 1.

  • If the lead exceeds the savings estimate, the additional amount is also counted towards the lead's bonus, but using a smaller share (percentage b) as the multiplier than the one used for the basic bonus in step 1 (b<a).
  • If the lead falls short of the initial target, the missing amount is also counted, but as a penalty subtracted from the base bonus. To encourage them to prioritize hitting their target, the penalty is calculated using a higher share (percentage c) than for setting the base bonus in step 1 (c>a).

Compared with today's target-setting processes, this incentive logic has three significant advantages:

  • The category leads have a clear motive to communicate exactly the savings they believe in. Any amount over the target they bring in at period end will be weighted lower in their final bonuses than if the leads had communicated the potential from the start.
  • At the same time, overestimating savings will have a direct negative effect, reducing their bonuses. Assuming risk neutrality, the leads therefore have a clear incentive to communicate the real bonus estimates they believe in.
  • Finally, even if the estimated savings target is reached, the category leads still have an incentive to continue adding to the savings, since they count towards the bonus, only with a smaller multiplier. The much smaller multiplier helps encourage more accurate estimates—and thus targets—from the start.

(See sidebar for a fictitious example of how this would work in practice.)

Getting implementation right

To make this novel incentive scheme work in their own purchasing departments, companies should focus on three critical elements:

  • Picking fair and acceptable percentages for the base bonus and over- and underperformance adjustments
  • Managing the effects of price volatility
  • Integrating the approach with the company’s wider business goals, incentive schemes and IT system

Making it fair

Spend volumes and the ease of achieving savings will differ from category to category. Therefore, companies may not wish to link the bonus directly to the cash savings number achieved by each category lead. A pragmatic alternative is to define an absolute bonus number for good performance in line with the category lead's tenure and additional salary components (e.g., in line with the previous year's bonus payment) and divide this by an estimate of the savings available in the category to give the base bonus share (percentage a). This initial estimate will inevitably be a top management approximation and it must be made with some care: for categories that recently have been tackled, savings estimates will be need to lower than those for untouched categories, for example. The exact percentage can be adapted in direct negotiations with the responsible category leads. Critically, whatever the agreed percentage in their category, the formula still motivates the lead to provide as accurate a savings forecast as they can in order to maximize their final bonus.

Managing volatility

Another topic that often interferes with target setting mechanisms is market price fluctuation, like that found in pure commodity markets. Such fluctuations distort results and create high uncertainty for category leads when they have to assess their potential savings to set the basis for the bonus. In these cases, we recommend measuring not the absolute savings within a category, but rather the category's relative performance against a given index (e.g., the MEPS index for steel3 ). In markets with high currency fluctuations or inflation, we also recommend adjusting the savings by weighting them with these fluctuations.

Integrating into company systems

The relative simplicity of the proposed model should make it quite straightforward to integrate it into a company’s IT and performance management systems. Some care is required, however, to ensure the overall incentive scheme does not only focus on savings but also incorporates other total cost elements relevant for the purchasing company. It is therefore advisable to include aspects such as product/service availability, quality, and other dimensions of performance into the incentive calculation. Similarly, care must be taken to ensure procurement staff is not incentivized to favor short-term savings over long-term supplier relationships. Sustainability can be ensured by paying out only part of the year-end bonus immediately and connecting later payments to the sustainability of savings.

* * *

Given the smart and simple incentive model described here, now might be an excellent time to rethink the way procurement sets targets, bonuses, and incentives.

About the authors: Björn-Uwe Mercker is an expert principal in McKinsey’s Munich office, Wolfgang Schnellbächer is a consultant in the Stuttgart office, and Jan Wüllenweber is a director in the Cologne office

Related Articles