This paper proposes an easy-to-implement dynamic measure of market performance over time for various selling units (e.g., sales territories, sales regional offices, or the whole sales organization). It may be used as a diagnosis tool by comparing the market performances of various units, taking into account the conditions prevailing in the different markets (such as competition relative effectiveness, sales penetration, or local market fluctuations). Combining sales volume, market share, and profit variations data into an Index of Sales Unit Market Performance (ISUMP), provides managerial guidance for selling units’ evaluation or resource allocations among units. This index may account for a firm’s selected market strategy (market penetration, market skimming, etc.). Implementation in a large North American insurance company is reported.
Many managerial marketing decisions (e.g., budget and resource allocations) and sales force decisions aim at maintaining or improving the performance of various selling units over time [1-3]. A selling unit is an entity responsible for achieving some output market performance. It may be, for instance, a sales territory (either assigned to a sales team or to an individual salesperson), a regional or district sales office, or the whole sales organization. Every selling unit is characterized by its market output performance (such as sales, profits, or market share) that results from the accomplishment of the selling tasks, in a specific environment, over some longand/or short-term period of time.
In order to effectively manage and control selling units, managers must rely on valid and accurate measures of market performance [
The aim of this paper is to devise a measure of a selling unit’s market performance over time that could supplement the management’s control tool kit. When considered jointly with extra-role performance measures [
After discussing the concept of selling units’ market performance measurement, the proposed procedure is described. An application illustrates the concept. Finally, the advantages and limits of this procedure are discussed.
Assessing market performance is fraught with difficulty, and sales force researchers and sales managers have typically used different approaches to that purpose [13- 21].
Over the last decades, a major stream of sales force research has attempted to gain better understanding of salespeople’s performance and to find appropriate ways to enhance selling effectiveness [
In most cases, researchers have measured selling units’ performance either using objective sales data [
In order to assess selling unit market performance, managers have typically used quantitative and/or qualitative measures [
Alternatively, in order to explain management controls of the sales force, some authors have made a distinction between behaviorand outcome-based sales force controls [50-52]. Although controlling salespeople’s behaviors is an important and necessary aspect of sales management [53,54], it would be worthless if the proper behaviors did not eventually translate into sales, profits, market shares, strengthened customer relationships, or clients’ satisfaction. In practical situations, managers do evaluate selling unit market performance from at least some objective quantitative measures of sales performance [
Territory sales volume, market shares, or profits, which are typically observed by managers, are inadequate market performance measures when considered individually. As stated above, although selling units are instrumental in building sales volume over time, sales are also influenced by a host of factors which are beyond the unit’s control. In most cases, managers cannot easily disentangle what parts of such outcomes must be attributed to a selling unit’s marketing program (attractiveness of the offers, advertising, company’s reputation, etc.), or to environmental factors (competitive strength, economic trends, etc.). As a result, sales volume alone may give a distorted picture of this unit’s market performance [35,59]. Although territory market shares bring the additional dimensions of industry sales, competition’s performance, and the firm’s market penetration, they suffer from the same problems as sales volume. In addition, all outcome measures of performance tend to over-emphasize short-over long-term performances. Finally, selling unit profit measures are useful to watch, in as much as the units are given responsibility to negotiate prices and/or are left to allocate their efforts and resources among several product lines with different profit margins [
Sales variations over some benchmark period (after removing seasonal effects) are frequently used and provide better measures because they indicate a unit’s success (or failure) in sustaining a given level of sales over time. Taken alone, however, sales variations can be misleading because they may also be caused by many factors beyond a selling unit’s control (for instance, a shift in territory environmental conditions, windfall sales in any of the two periods, or simply random factors affecting sales in one or both periods of time).
Customer satisfaction could be a more appealing measure, because it involves an important long-term firm’s objective. Unfortunately, customer satisfaction variations are not always easy to measure. In addition, research suggests that salespersons’ confusion can result from control systems that combine both customer satisfaction and short-term performance measures such as sales [
Market share variations in a selling unit’s territory provide a clear indication of a firm’s standing in its market. It measures a company’s market penetration, and, to some extent, the selling unit is responsible for it. Market share brings the additional dimension of industry sales, and consequently, of competitive performance. Competitive sales performance constitutes a natural benchmark for evaluating selling units’ market performances. In other words, it is essential to account for the evolution of the competition’s sales levels for evaluating selling unit market performance.
When territory sales, industry sales, market shares, and profit variations are considered jointly, they provide a more complete assessment of a selling unit market performance. Various market conditions require different efforts and abilities [
Another important dimension of selling unit market performance is the time frame over which it is measured. While building long-term relationships with customers has become a prevalent strategy, short-term sales have often become inadequate measures of market performance. The current emphasis on sales activity based at the expense of outcome-based controls is a logical consequence of this evolution. Consequently, there is a need for devising selling unit’s market performance measures that 1) are meaningful to managers, 2) are based on easily observable and quantifiable data, 3) are valid and reflect the ability of selling units to progress, taking into account its market evolution, and 4) can accommodate shortas well as long-run market performance assessments.
Like DEA, the selling unit market performance measure used in this study is also a multi-criterion procedure. It presents, however, several advantages over DEA: 1) it measures “true” market performance, discarding environmental factors that influence the sales results of every selling unit; 2) unlike DEA which is an extreme point method, it is not sensitive to some unusually high or low unit performances, and consequently to outlying observations; 3) it requires much less computations (against one linear program per selling unit in DEA); and 4) it is more easily understood and applicable by managers.
Managers assess a selling unit’s market performance at time T1 by comparing the performance this unit has achieved over some period of time t (typically a few weeks or months, ending at time T1) to some selected benchmark. Depending on the situation and management’s goals, the selected benchmark can be 1) the outcomes achieved during a comparable reference period of length t (such as the preceding period T0 or some other selected period, or an average of several selected benchmark periods), or 2) some target outcome levels to be achieved during the considered period of time (for instance, some sales objectives or quotas). The analysis can be carried with one or several types of appropriate benchmark.
The choice of a benchmark is crucial and requires careful managerial consideration. Selecting a preceding similar period is appropriate whenever, during this period of time, external factors have not differentially and significantly affected the various selling units.
At some time T1, four main factors reflect different aspects of a selling unit’s current market performance variation over the benchmark: 1) the total industry sales (IS) variation in this unit:; although a selling unit cannot be held responsible for industry sales variations, one should recognize that it is more difficult to improve a firm’s market position in a declining than in an expanding market; 2) the unit’s sales volume (S) variation over the benchmark situation; and 3) the corresponding selling unit’s market share (MS) variation
because performance should account for a selling unit’s ability to improve or maintain the firm’s market position in its territory; and 4) the corresponding gross profit variations. Being measures of variation (in decimal form), is1, s1, ms1, and π1 can either be positive or negative (or null) and will typically vary between −1 and +∞1.
In order to provide a simpler explanation of the underlying concepts and rationale of the proposed method, a simpler case that does not consider profit variations is discussed first. Then, in a following section, the method is generalized to the case including profit variations.
These three dimensions are somewhat interrelated. Increased sales in a decreasing industry sales territory imply an increased market share (Case 1). Decreased sales in an increasing industry sales market imply a decreased market share (Case 6). No inference about market share variations can be made, however, on the basis of sales and industry sales variations alone when both sales and market industry sales increase (decrease). In such cases, market share increases or decreases, depending on the relative sizes of the sales and industry sales increase (decrease) rates. Therefore, one should explicitly take market share variations into consideration. The six possible cases defined in
The six situations described above are represented graphically in
Although theoretically, s, is, and ms can possibly vary between −1 and infinity, in most usual cases, their values are likely to be relatively close to zero (NO = status quo, i.e., no change over the benchmark: s = is = ms = 0). The six zones shown in
In order to compare the market performance of various selling units, one can compare the percentages by which each one has moved on its vector compared with the status quo N0. A formal measure called ISUMP (Index of Selling Unit Market Performance) can be used to that effect.
Let successively be sales variations (s), industry sales variations (is) and market share variations (ms) in Selling Unit i’s territory (all in decimal form) (subscript i omitted unless necessary):
with, andwhere:
sales level achieved by Selling unit i in period 1, in dollars;
sales level for this unit in the benchmark situation, in dollars;
industry sales in period 1, in dollars;
industry sales in the benchmark situation, in dollars;
market share achieved in this unit’s territory in period 1, in decimal form;
market share for this selling unit’s territory in the benchmark situation, in decimal form.
Replacing S1 and IS1 by their values in (1) and (2) in Equation (3) leads to:
Given specific values of is, one can determine the linear relationship linking market share ms to sales volume s variations, as shown in Equation (4). Thus, for; for for is = 0, ms = s, and for. In the same way, because the three concepts are interrelated, each one can be expressed as a function of the other two:
As shown in
In order to compare the market performance achieved by various selling units, one must assess by which percentage each unit has moved on its vector from the status quo. For a given unit i which has achieved sales variations of si0 and market share variation of msi0 in period 1, the market performance increase (or decrease) in comparison with the benchmark situation 0 (or Index of Selling Unit Market Performance, ISUMP) is given by:
By replacing si0 and msi0 by their values in (1 - 3) and rearranging the terms leads to another expression of the ISUMP:
ISUMP = 100 means no market performance improvement, ISUMP > 100, some improvement, and ISUMP < 100 some market performance decrease. In order to assess how a selling unit’s market performance compares with the overall higher order entity (e.g., a given sales territory within its corresponding branch office) market performance, this index can be supplemented by an adjusted ISUMP defined as2:
The analysis can be supplemented by locating the position of every selling unit on the six-zone map (tips of market performance vectors in
The same principles apply when profit margin variations are added to the analysis. In this case, there are twelve situations (and twelve corresponding zones); each case identified in
by the s, ms, and pm variations.
There are two ways through which a selling unit can increase profits: increasing sales and/or the profit margins on the goods or service sold (either through negotiation of higher prices and/or selling more profitable products). The profit situation is therefore characterized by:
where PM0 and PM1 are the profit margin rates respectively in the benchmark situation and achieved in Period 1. Defining the profit margin variation as leads to:
A simple extension of the previous analysis leads to an estimate of the Index of Selling Unit Market Performance (ISUMP):
or using Equation (10):
In many cases, a firm may equally value sales, market share, and profit achievement. Alternatively, when introducing a new product line, a firm may select a strategy of fast market penetration. In this case, sales and market shares may be assigned a higher weight than immediate profits. In other instances, the firm may select a market skimming strategy. In this case, profits may be given more importance relative to market share. Other strategies may be pursued. In such cases, selling units that are given the responsibility to implement market strategies, and it seems natural to assess the units’ market performance at implementing them. This issue can be easily handled by management assigning different weights reflecting the relative importance of the three outcomes. Let:
weight assigned to market share increases
weight assigned to sales volume increases
weight assigned to profit margin increases with
In this case, a straight application of these weights to the corresponding dimensions lead to a new expression of the Index of Selling Unit Market Performance
(ISUMP):
In addition, when management wishes to assess selling units’ performance at implementing different market strategies for different product lines, the analysis may be carried out for the various product lines separately. Then, management may assign different weights to the product lines reflecting their relative strategic importance. A weighted market performance index is computed for every unit, and compared to the corresponding higher level selling entity’s ISUMP.
This procedure has been implemented in a large NorthAmerican insurance company which sells directly to customers with no intermediary involved. Like many North American insurance companies, this firm lacked adequate means to properly assess the market performance of the various selling units (sales agents and sales managers) at developing their territories over time. Thus, the general sales manager was highly concerned about the current market performance evaluation procedure where sales managers were essentially relying on one single criterion, i.e., the sales volume achieved in each territory. As a result, several salespersons grew dissatisfied with this criterion: they argued that even though their sales grew slowly (or even decreased) their territory market share was increasing. Market share data were collected from internal services and communicated to the sales personnel each month.
In this case, the application involved all the branch offices covering the North American market. The company used 713 insurance agents under the supervision of sales managers in 28 sales territories. Because of space constraints, only the results of the 28 sales offices are reported here. In other words, the branch office has been selected as the selling unit. For that purpose, the required data are the summated results across territories belonging to a selling unit. Those were calculated by cumulating the results of the insurance agents they supervise.
Selecting an adequate time period length to assess performance variations is an important decision. Too short a period of time (e.g., monthly data) could lead to wide market performance variations and may hide the true longer term performance of some selling units. Alternatively, too long a period of time (e.g., more than one year) may not be flexible enough, especially if management bases some financial rewards on short-term performance. This is why it may be worthwhile to carry the analysis with various time period lengths, whenever possible. In this application, quarterly performance results were not judged by management to be stable enough. Consequently, a one-year time period length (t = one year) was deemed most appropriate.
In addition, management selected market performances achieved during the previous year as benchmarks for comparing current market performances. During that year, no new product had been launched, marketing investments had been normal and the market had remained pretty stable. For illustration purpose, for a sample of sales territories at both ends of the performance spectrum (ST1-ST3 and ST26-ST28), the first eight rows of
The percentage variations (in decimal form) in sales (s), industry sales (is), market shares (ms), gross profits (π) and gross profit margins (pm) in Year 2 over Year 1 have been used for computing the Index of Selling Unit Market Performance (ISUMP), using Equation (11) or (12). In this application, gross profit margins were estimated as:
The interpretation is straightforward: the ISUMP index relates market performance in every sales territory, relative to the benchmark year, taking into account the evolution of industry sales and consequently, the impacts of competition and other environmental factors in the territory.
Considering first the Year 2 results (excluding profits), as can be seen in
exploited the growth opportunities of their markets: their sales volumes increased in an expanding market, but not sufficiently to maintain the firm’s market position at its original level. Finally, two sales territories (7 percent), performed pretty poorly: their management failed to exploit the market growth opportunities and weakened the firm’s position in an expanding market (Zone 6). Only in one sales territory (4 percent) management could strengthen the firm’s market position in a decreasing market demand, but not enough to keep the same sales rate (Zone 3). No sales territory has been observed in Zone 5.
The graphic illustration of
Adding the profit dimension to this analysis (bottom of
Of the sales territories considered in this analysis, 22 (79 percent) fell into an a zone (profit margin increase). Only six (21 percent), fell into a b zone and experienced a profit margin decrease. Consequently, because of these overall good results, the whole company experienced a high ISUMP index of 119.51. The highest performing sales territories show successively ISUMPs of 205.95, 164 78, 147.84, 147.80, and 142.18. The three lowest performing territories show ISUMPs of 87.38, 98.26, and 99.02. This shows that including the profit dimension into the analysis sheds some new light on the various territories’ market performances. For example, ST5 with a below average sales performance (adjusted ISUMP = 98.81), had a superior performance when profits were also considered (adjusted ISUMP = 123.71). From a managerial point of view, sales territories that depart sharply from an ISUMP value of 100 should be considered for further diagnosis and possible corrective actions.
One of the most frequently used bases for evaluating a unit’s sales performance is sales increase/decrease over the last (similar) period. In the reported case study, variations of selling unit’s sales as a measure of market performance does not fully account for the different territory situations (industry sales variations or market share evolution). As can be seen in
This analysis has been extended to the cases where the firm would pursue either market penetration (MPS) or market skimming strategies (MSS) (versus an undifferentiated strategy giving equal importance to all three objectives). In such cases, top management could assign different weights to the three objectives (sales volume, market share, and profits). In these occurrences, the firm’s management should clearly communicate those weights to all the concerned managers, and inform them of the strategic market priorities. As an example, management could assign the following weights:
Market share increases α = 0.1 α = 0.6
Sales volume increases β = 0.3 β = 0.3
Profit increases γ = 0.6 γ = 0.1
Total 1.0 1
The results obtained after application of Equation (14) are shown in
*Spearman’s Rank Correlation = 0.100.
in the ISUMP index so as to reflect the selected market strategy makes it possible to assess how effective every selling unit has been at implementing this strategy.
Following implementation, the company’s top managers in charge of business development in North America were impressed by the outcomes of this project. For the first time, they could make a sound analysis of the market performance in the different sales territories in terms of business development, which was not done before. For Year 2 annual sales territories’ evaluation, top management used the results of this analysis for making a better allocation of its resources. They had at their disposal a sound basis for denying additional resources requested by some sales agents and assigning them to sales territories that could profitably take advantage of market opportunities. This method provided top management with a powerful tool that could help assess their own strategies and decisions and point to possible resource allocation improvements. In addition, managers could identify the zones with large untapped company potential for growth. Although the ISUMP indices are only indicators of market performance and do not provide a diagnosis, they could allow management to identify and question the sales managers in charge of every territory in order to find proper explanations for their market performance and plan actions to take advantage of every market opportunity.
The proposed procedure has several advantages: First, it provides short-run and/or long-run sales market performance assessments that, as can be observed frequently in practice, are dimensions that many sales managers like to watch very closely. It reveals also which selling units could and/or should take advantage of market opportunities in order to reinforce the firm’s market position, a typically long-term objective, generally implying customer relationship and loyalty building. Second, the proposed market performance measure is simple to compute and understand within a firm and its various selling units. It is based upon the three major ingredients of market performance, i.e., increases/decreases of 1) sales, 2) market share, and 3) profit, compared with some selected benchmark(s). These are elements over which selling units are generally recognized to exert direct influence, and that are easy to measure. As a result, this method can be implemented easily and at low cost as part of a CRM application or business intelligence tools. Third, the proposed market performance measurement process is fair. It provides selling units with evaluations that are commensurate with their actual market performance, by making the generally reasonable assumption that environmental conditions equally affect a firm and its competitors. Fourth, the procedure is dynamic, and can be applied over several consecutive periods of time, and/or with various time lengths. As a result, it can be a useful device for tracking selling unit market performance over time, for shorter or longer periods of time, depending on the objectives. Fifth, the ISUMP index can be applied to assess the market performance of various sales entities (from territories assigned to a sales teams or an individual salesperson, regional or district sales offices, to the whole sales organization) and for various product lines, providing a common basis for making useful comparisons among and across those entities. Finally, a firm could use the ISUMP indices for allocating financial rewards (such a quarterly bonus) for short-run market performance [
The proposed procedure has also a few limitations: First, this method is applicable in cases where several firms compete in the same market and when the considered firm does not hold too large a market position compared to its competitors. Second, random sales variations due to environmental uncertainties or unusual circumstances are accounted for only implicitly. For instance, an unexpected windfall sale would increase S1, s, and consequently, provide too high a market performance evaluation for that period. Although the ISUMP indices are effective market performance diagnosis tools and point to more and/or less efficient selling units taking the territory conditions into account, they do not diagnose why such market performance levels have been reached. In other words, the ISUMP index can identify possible problems, but does not provide a diagnosis. Third, as a corollary, such market performance analyses should not prevent management from watching other more qualitative aspects of selling units’ market performances. For instance many human aspects (such as, for instance, the characteristics of the salespersons or the sales teams in charge of the selling units, e.g., their experience or career stage) should be kept in mind when carrying such an analysis. This index may be considered along other indicators in order to obtain a complete picture of every selling unit market performance. Finally, the proposed method requires access to sufficiently reliable market share data for every selling unit. Note, however, that in many industries (like the pharmaceutical industry) firms have access to such syndicated data on a regular basis.
This paper has described a relatively simple procedure for assessing various selling units’ market performance, not only in the short-run, but also accounting for a firm’s market position improvement (or decay) a typical longterm firm’s marketing concern and a more relevant selling unit’s market performance measure. The assessment formula is simple to explain and easy to administer. It requires only three sets of data: selling units’ sales volumes, market shares, and profits in the corresponding territories. In addition, the outcomes of a selected market strategy can be assessed easily. A firm can integrate this procedure into its CRM or other sales intelligence system easily and at low cost. This procedure has been illustrated with an actual case study. It has been shown to provide more adequate results (from a marketing point of view), and more equitable market performance assessments than more complex (or even simpler) comparable procedures.
There are several ways in which the proposed procedure could be extended or refined. For instance, it could be modified to account for various variables, such as random factors and uncertainties. It should be kept in mind, however, that these refinements would come at the cost of making the procedure more complex. Consequently, they should be included only if the new benefits are worth the additional complexity.