Case study

The company ABC is in a consumer products industry and uses a direct store delivery (DSD) system. Under DSD, manufacturing companies take responsibility for shipping products from their plants and warehouses directly to retailers, and for stocking the product on the shelves (versus using the retailers warehouse and distribution systems). The focus of this study is the Route Representative (which we often refer to as a DSD employee), who is responsible for four tasks: ordering products, delivering them to the retailer, stocking retailer shelves, and negotiating additional shelf space. With the exception of delivering the product, all of the other tasks have a direct impact on sales, and require the driver to intelligently analyze sales trends, manage relationships with store managers, and creatively market the product within the store. Ordering affects product mix and avoids out-of-stock situations, stocking (location, neatness of the display, rotating older products to the front, etc.) affects how customers choose which product to buy, and gaining additional shelf space for displaying the product (at the expense of competing products) affects total volume sold.

Retail establishments sell a wide variety of products, so each manufacturer has minimal impact on the flow of traffic into a store. Instead, competition for consumer dollars at the retail level takes place among products that are substitutes in the consumers shopping cart: tobacco competes with alcohol, various drinks compete with each other, breakfast cereal competes with ready-to-eat breakfast items, over- the-counter weight loss aids compete with comfort foods, etc. At the margin, manufacturers increase sales through branding and marketing campaigns, price promotions, gaining additional shelf space at the retail establishment, and varying the product mix available on the shelves. Of these, DSD employees can directly influence the last two.

The firm groups routes in a local geographic area into sites with common management over- sight. A site consists of a contiguous geographic area covering a distinct consumer and labour market. For example, a small- or medium- sized city and surrounding communities might constitute a site. Each site has a leadership team, supervisors, and DSD drivers. To keep the span of control and complexity at a reasonable level, each site ranges from approximately 80 to 120 routes. Very large urban areas contain multiple sites, but the area covered by each site within those large urban areas is drawn to correspond to natural community divisions.

A. Initial and New Pay Schemes

Both the schemes, i.e., initial and new, are linear commissions on the dollar value of net sales revenue (gross sales net of returned / unsold product). There are a number of advantages to this general approach in both the original and pilot schemes. A linear scheme based on net revenue is easy to understand by employees and supervisors, and to administer. There is transparency because there is little dispute regarding the dollar volume of total sales: employees observe which products sell at what prices, and which do not and are returned to the warehouse. The performance measure should provide relatively well-balanced multitask incentives and thus not distort incentives much away from optimal behaviour. As long as profit margins do not vary much across individual company products, revenue provides incentives to manage trade-offs between sales of different products in the same store. It also motivates the employee to effectively allocate effort to different tasks. The measure also motivates the employee to avoid excess inventory, since unsold product is not included in sales. (The employee is not charged for the opportunity cost of excess inventory in the store or truck, but that is likely to be a small cost for this firms products.)

The company had two concerns with the original pay scheme, which they tried to address with the pilot scheme. First, the original scheme did not explicitly motivate DSD employees to care about sales growth, on which managements incentives are based. Second, the firm wanted to increase the strength of the incentive. In addition, although not an explicit concern of the firm, the incentive plan may motivate Ratchet Effects from how targets are set. We examine all three of these issues in the empirical work. Next, we describe the initial pay scheme in detail many of its features were maintained in the pilot scheme and then how the firm adjusted it in the pilot scheme.

i) Initial Pay Scheme

The commission in the initial pay scheme is fixed and depends on route size. The company divides DSD employees into three route-size categories. Within each category, it attempts to design specific routes a set of stores in the same geographical location so that expected compensation will fall in a fairly narrow range, assuming equal ability and effort of employees. The commission rate for each route type is set to equalize expected compensation across the three types, so the highest volume routes have the lowest commission rate, and vice versa.

Though employee compensation under the initial pay scheme does not depend on a sales target, bonuses for frontline supervisors and leadership depend nonlinearly on the site meeting its growth targets. This was a main source of disconnect between the companys objectives and DSD employees incentives under the initial compensation scheme. Sales targets are set for the DSD employees too, in an attempt to encourage them to meet growth targets, and supervisors regularly met with the employees to identify ways to increase sales. However, under the initial pay scheme DSD employees are rewarded directly for increased sales, but only indirectly for sales growth, and their commission does not vary with the growth rate.

Setting of sales targets for employees at various levels is done using a complex combined top- down and bottom-up approach, which is described in some detail. This approach applies to setting target sales in both the initial and pilot compensation plans.

First, dollar volume sales targets are set for senior management, based on the total sales growth target set by the company. Those cascade down to site targets for the middle and frontline managers through the annual budget and planning process. Sales growth rate targets for supervisors are determined by targets for their DSD employees. Though differences across routes might in principle lead to different sales growth targets among supervisors, in practice individual route differences equal out across supervisors. Thus, supervisors have essentially the same sales growth targets as the overall site target (i.e., if the sites target is 2 percent, the target for each supervisor is also 2 percent).

Sales targets for DSD employees are first and foremost determined by the sites overall target volume sales: if the sites target growth rate is 3 percent, individual route targets are set to yield 3 percent growth. Thus, the highest volume routes have the largest dollar volume targets, and vice versa. Deviations from that distribution are allowed in unusual circumstances. For example, sales growth at new retail outlets are historically higher than at older establishments, so a route with a high concentration of older outlets might receive a slightly lower sales target. Such adjustments to a routes target are made by site leadership (not the supervisor), but only by making offsetting changes in targets for other routes at the site. This process ensures that gaming of targets by supervisors is minimized, and that adjustments to individual route targets are the exception and not the rule.

Relative performance of the employee may be accounted for indirectly because further adjustments are possible in cases where historical sales growth is significantly lower than expected given route characteristics. Employees deemed to have provided relatively low effort might be given higher targets in the subsequent year. In practice, however, the single most important determinant of an employees target sales is the overall growth goal for that region.

Note that this approach seems to imply a Ratchet Effect: an employees sales this year raise the target sales level next year. However, under the initial compensation scheme, the commission rate is constant, and compensation is therefore independent of whether or not the employee meets the growth target. Ratchet Effects may instead be created by the firms practice of route rebalancing. If expected sales on a route grow too large, the job may become difficult for one employee to handle because of the physical requirements that the employee deliver and stock the products sold. In such cases the firm adjusts the set of stores in the employees route to bring volume back in line with the companys view of physical limitations of DSD employees in performing the job. Lower volume also makes it easier for sales growth targets to be achieved, though at the cost of creating more routes in aggregate. Such rebalancing typically takes place at most once per year. Route rebalancing is one of the most disliked and distrusted management processes. The company uses lump-sum pay adjustments to try to ease the impact on total compensation, but the adjustments are imprecise. As a result, employees might exhibit a Ratchet Effect whereby they lower the growth rate of sales to try to avoid larger adjustments through route rebalancing.

Now consider the intensity of incentives in the initial plan. DSD jobs typically have high sales volume of small, inexpensive products. It is not uncommon for sales on a route to average $10,000 per week or more, for total annual sales around $500,000. DSD employees typically are high school graduates or have some college education. Assuming target compensation of approximately $40,000, the commission rate would need to be 8 percent. Higher expected sales would mean an even lower commission rate. The nature of the job and high volume of products sold mean that relatively little time is spent on pure sales activities such as negotiating for additional display space on a regular basis. Thus, a key question is whether the commission rate in the initial plan provides sufficient incentive to elicit effort. For example, additional sales of $500/week, added up over all employees routes, could significantly help a company meets its profit objectives. Yet at 8 percent commission, and assuming a marginal tax rate of 25 percent, an extra $500/week in sales translates into only about $30/week extra in take-home pay. This is one reason the company changed the compensation system.

ii) Pilot Pay Scheme

The pilot pay scheme was designed to provide stronger incentives for employees to meet or exceed their targets, so that local management would be better able to meet overall growth goals set for them by higher levels of management. This was accomplished by establishing base compensation BC, or minimum salary paid regardless of volume, and a steeper commission rate paid out only for sales growth above a threshold T. Interestingly, this is precisely the kind of change modelled by Oyer (2000), with the same intention of increasing the incentive intensity in a desired range of output. For sales growth above TS, compensation under the pilot scheme was higher than under the original scheme, and vice versa for sales below TS.

The company set the new compensation formula so that compensation is equal in both plans when sales equals the growth target. The base pay is intended to provide a safety net to workers

in the case of natural disasters or other calamities that might cause consumers to stop purchasing products other than bare necessities for extended periods of time. In normal circumstances, workers are expected to achieve at least minimum expected sales M > T, so that base pay BC will be non-binding. While a worker will not face termination for performing below M in a single month, consistently performing below M will trigger termination (this was also true under the old system). In the pilot plan, the worker knows BC, T, and the commission rate, but M is implicit.

One potential concern in sales incentive plans is that employees might time sales or returns of items around the end of a bonus period (e.g., end of fiscal year) in order to maximize bonuses across the two years (Oyer 1998). In principle that concern might apply to the pilot scheme, for employees with performance very close to Threshold Sales T, since sales below T receive no commission. However, employees with sales so low are highly likely to be terminated, eliminating that concern.

By changing the relationship between pay and sales growth, the new incentive scheme had the potential to change Ratchet Effect incentives. Under the original scheme, an employee would want a route with the highest volume possible to maximize total sales, even though growth might be minimized. The threat of route rebalancing created a potential Ratchet whereby employees might restrict sales growth to avoid having an account taken away. Under the new scheme, with compensation aligned with sales growth instead of sales volume, employees on lower volume routes should have an easier time meeting sales growth targets than employees on higher volume routes (assuming the same growth rate target). Thus, under the new pay plan employees might no longer resist route rebalancing, and might even welcome it if they have confidence they can more easily grow sales with the stores remaining on a route after rebalancing. This does not ensure a lower Ratchet Effect under the new scheme, but suggests that it might be lower under the right circumstances.

Questions based on case study:

1. Discuss the main elements of initial pay scheme followed by company ABC. What do you think were the reasons for changing this scheme to pilot pay scheme?

2. Highlight important characteristics of company new pay structure.

3. Indicate the advantage and limitations of the two adopted pay schemes of the company. How such changes in the initial pay structure will affect the staff performance, justify your statement with proper examples?