🙃The paradox of an agent

This article addresses the motivation of active sales agents and the establishment of sales benchmarks under circumstances where:

the number of Customers and their "value" are both critical; the initial transaction size is not a reliable indicator (a significant Customer might initially place a small order);
the Customer count per month is low, yet any Customer could unexpectedly place a substantially large order (which would be incorrect to use as a benchmark);
it is important that the agent tries to conclude a contract on terms favourable for the company when the deferment of payment is used.

The article provides a comprehensive solution for all four challenges within the context of active sales technology.

Designing 'smart salaries' Salary is the “gene of the company" which, upon examination, reveals much about the overall "health" of the organisation. The extensive experience (since 1993) of TRIZ-RI specialists in consulting indicates that there is scarcely any consultancy work that does not involve the creation of salary models for employees and/or the establishment of key performance indicators. More details ...

Problem 1: Why are the words "good" and "wealthy" not synonyms?

A. Conditions

How can we provide a sales plan to a commercial agent who actively seeks new Customers, when not only the number of Customers is important but also their "value", which is difficult to formalise?

It is evident that "bigger" Customers are always preferable, but how do we measure their size? It would seem that a partnership with a large retail chain as a supplier is preferable to engaging with the small shop next door, yet the small shops are also significant (due to their abundance). Furthermore, it is crucial not to differentiate in our conduct towards different Customers: there are no "big" or "small" clients – we should treat all of them with the same care.

B. Traditional incorrect answer

Defining the significance of a Client solely through income: the one who purchases in large amounts is a "good Customer", while the one who buys in smaller quantities is also good, but "has not purchased much".

Fair and square, such an answer indicates a lack of strategic planning in sales. A competent company should establish its operational strategy properly and should aspire not just to make a lot of money at once. Therefore, if we define the significance of Customers by the sum of their average transactions, we need to discover a specific approach. A robust strategy still includes a line of products (services), varied in price, but serving different roles in the entire range. For example, products with a high turnover rate compared to their low price (or the same standard services offered "in crowded places") are utilised to increase the frequency of visits, subsequently leading to the sale of more expensive items.

Typical examples of different strategies are described in the article "The Trend of the Development of Commercial Services".

An entrepreneur addressing the "motivation" issue should have "a game plan". The incorrect answer lacks such a plan, i.e., it does not include any market segmentation.

Typical Example:

Clearly, a hotel Customer who pays for a suite is valuable. However, Customers who opt for a standard room are also valuable:

They often travel for business and usually stay in economy class;
Some might then return with their families and rent a large, expensive room if they enjoy the service and, of course, if the location is not too remote;
The influx of economy business trips leads to an increase in the rental of conference halls;
Conferences lead to the hotel being fully occupied and a rise in restaurant income;
Customers recommend the hotel to their colleagues and business partners. Some of these referrals will book a suite then;
Clients post reviews on popular websites (for example, www.tripadvisor.com and others). This flow of information gradually shifts the booking process towards such sites;
And finally, as more Customers book standard rooms, the likelihood of the aforementioned events occurring increases.

If the hotel only offered suites, it would face bankruptcy. However, by efficiently managing different Customer segments, the suites are usually fully booked as well.

C. A step towards the correct answers

Let's identify the cause of the problem. The issue arises because a single-factor system (based on the volume of transactions) is applied to a multifactorial situation (various contexts of a Customer value).

Eliminate this cause. Standardise the Customer types.

Example:

A wholesale company "Creative Enthusiasts" supplies goods to a diverse range of companies:

It supplies to large retail chains,
It supplies to various small but numerous shops,
It supplies to corporate Customers, who purchase these goods for their personal use,
It supplies to small regional supermarkets,
etc.

At the same time, Customers do business differently: some order in small quantities, others in large volumes. Thus, the "average transaction" values of different Customers vary significantly. Additionally, some Customers request payment delays, some cannot operate without them, some take the products to market, and some pay immediately but demand discounts, etc.

Step 1: Let's formalise all this variety. For example,

Step 2: We will assign a score to each Сustomer type.

To do this, we need to:

a) Imagine a scenario in which each Customer of a specific type pays an equal, typical sum. Let us assume it amounts to USD 5,500. (Of course, you must envisage your typical deal).

b) We ask ourselves a control question: "In this scenario, where Customers are paying equal sums for their first deal, are they on the same level in terms of potential?" For instance, an international chain of supermarkets (with which we will continue to work in the future) that places an order of USD 5,500 today, is most certainly not on the same level as a one-time client who pays the same amount.

c) If Customers (for the same order) are not on the same level, then what is the exchange ratio between them?

For example,

Of course, for your particular business, you will need to create your Customer typology and define your "exchange ratios".

Step 3: The benchmark (planned) number of new Customers that the agent is required to attract each month will be expressed in scores.

That is, we will multiply it by a "scoring" factor. By summing the obtained results, we will calculate the overall plan for the number of new customers (in scores).

Here it is:

The ratio of the actual output to the benchmark will show us the performance for acquiring new Customers.

Problem 2. Why someone small happens to be big

A. Conditions

How should we encourage agents to find new Customers if the first transaction is not indicative? Large Customers may often place small trial orders. In a conventional performance system (e.g., one based on percentage), an agent will find that his initial transactions are disproportionately small compared to his efforts.

B. Traditional incorrect answers

Incorrect answer 1: "So what?"

It is sometimes suggested, "So what? Let them improve their dealings and search more actively for other Customers. If they work productively, they will earn a good salary for the month. After all, their duty as agents is to persuade Customers and to increase order volume. As the saying goes, 'A dog that trots about finds a bone' – that's precisely their job."

Despite its apparent logic, this answer is flawed because:

a) The underlying assumption is: "Let number of deals increase, and 'statistics' will smooth out any irregularities, with more 'statistics' leading to fewer variations." If this were true, would there be a need for sales agents at all? A “smoothing of statistics” will not work with numbers of tens of transactions per month (let alone smaller values).

Otherwise, it wouldn't be an issue. If we're talking about a higher number of transactions, then sales are probably not being organised effectively. If the aim is to manage the flow, why use agents at all?

b) A Customer is not visible behind such reasoning. When smoothed out, deviations are not analysed.

Better questions would be:

Who is this Customer?
Why are they acting this way?
Is this a one-off, or have there been similar cases?
If we collect all the “first small transactions” over a long period, won’t a pattern emerge?
Do Customers sampling our services share any characteristics?
Are there gaps in our product range that could explain this "strangeness"?
Why not interview Customers who are now placing large orders but started with small quantities?

In short, behind beautiful words about statistics, there may be a traditional "left to drift".

Incorrect answer 2: "Let's wait a few months and then take into account the subsequent deals."

This too is false. How long should we wait? How many deals should we consider? While in theory such an agreement is possible, what practical steps will we take when, for example, a large order follows two small ones? Should we count it? And if a significant deal comes after three small ones? The agent will say, "I secured this large order by effectively working with this customer before!" Should we accept this claim? After all, the same rationale could apply to any number of initial deals.

This approach may reduce the incentive to seek new customers, focusing agents on servicing existing ones to secure large orders. Over time, an agent's activity may decline as they rely on their existing client base, which “already feeds them fine".

This often leads to Incorrect Answer 3: So, let them focus on 'their' regular customers. Let them nurture these relationships.

It is known that "20% of clients bring in 80% of the net income…, etc."

Unfortunately, this apparently rational reasoning is wrong in this case. Servicing and expansion of relationships and orders from current customers are important, but it's not the same as seeking new clients. But we cannot give up the active search for new clients.

More precisely, we cannot solve a private motivational problem at the cost of reducing/curtailing work on finding new Clients.

The management and expansion of Customer relationships should be consistently and effectively handled by other specialists. We will discuss how to motivate them in our other articles.

C. A step towards the correct answer

We receive incorrect answers when we attempt to rectify the situation without first identifying the cause (the root) of the problem. What's the point of "fixing the consequences"? We fix one issue – another emerges.

Let's identify the cause. The reason for the problem is as follows: the reward does not correspond to the effort. In this case, it is disproportionately small.

Does this hint help? If it does not yet, we shall consider the reverse problem.

Problem 3: Why is big sometimes small?

A. Conditions

What do you do when you can count the number of Customers an agent has acquired in a month on the fingers of one hand, and on top of that, one Customer has placed an unexpectedly large order (so it is incorrect to use this large single order as a benchmark)?

In such cases, it's alarmingly easy to err by setting the sales benchmark two or even twenty (!) times too high. Surely, it's not right when monthly salaries to fluctuate very markedly for just such reasons, is it?

B. Traditional incorrect answers

Incorrect answer 1: ""Ignore this"

This response is often addressed in a manner akin to one of the misguided responses to Problem 2: "Ignore the number of Customers. Some months, the manager will receive 'extra'; in others will receive 'less', but overall, it will even out."

Clearly, this approach is flippant and flawed. We are compelled not only to reiterate the points made previously (refer to Problem 2) but also to suggest a careful re-reading of the problem: the number of deals is small by condition.

One could, of course, choose to avoid this issue by posing a new one: "How can we increase the number of new Customers to a level where the current issue becomes irrelevant?" However, this approach is fundamentally different from suggesting the number of Customers be ignored.

Incorrect answer 2: Turn the indicator down

An equally flippant suggestion often made is to diminish the importance of the "New Customers" performance indicator. The logic typically unfolds as follows:

If the total number of transactions (from both existing and new Customers) is significant, any variations will be mitigated: whether two or five new Customers are acquired in a month, the difference in new acquisitions will be absorbed by the overall income. The focus should be on a "final result" to ensure income growth.

This flawed argument is often supported by such "final result" sloghans.

The key flaw (rendering this advice ineffective) is as follows: precisely when the Customer base has expanded sufficiently, the influx of new Customers will diminish to an unsatisfactory level (in specific units). And along with the fluctuations, motivation will wane.

Why seek new Customers if there are so few per month (in relation to the total number and income from all current transactions), and the significance of this indicator is negligible, and one Customer differs from another, etc.. So the slight salary increase will be disproportionate to the large efforts.

It is easier to invoice for the amount than to "worry" about acquiring new Customers.

Incorrect Response 3: is the adoption of a regressive scale.

The greater the Customer's payment, the smaller the commission earned from the transaction.

The demotivating effect of this "method" is comparable only to the slashing of rates for overfulfillment of the plan in the period of socialist realism.

Moreover, this "strategy" encourages dishonesty and creates anxiety for the customer, complicating business processes for both the customer and the company. For instance, a manager might begin requesting customers to split transactions into multiple parts, etc.

A step towards the correct answer

Cause of the Problem: A significant change in the volume of a transaction (by several multiples) is not linearly proportional to the agent's efforts. Consequently, rewards do not align with efforts either. Incidentally, although Problems 1 and 2 are delineated separately here for convenience, they stem from the same cause.

To eliminate the cause of the problem, we need to remove the linear correlation between the amount of reward and the sum of the deal in the motivation formula.

Thus, we propose introducing ranges: let's set a "range" of income sizes for the first deal with each new Customer and give an estimate to the corresponding "amounts".

To achieve this objective:

Step 1: Identify the actual maximum value of a one-off transaction.

Suppose it is USD 14,000. Then let the range "higher than USD 14,000" be the last line in the table of ranges (see Table 4).

Step 2: Exchange the "Money to Opportunities"

Let’s say that someone proposes us to exchange the "Money to Opportunities": "I'll present you a one-off Customer, willing to spend 14,000 dollars for a single deal and for one time only. In return, you just introduce me to a single just one prospective Customer for many deals, regardless of the amount of the deal now".

Suppose a company (in the example considered) mentally trades a "supermarket" for an "instant half a million." (Should this exchange seem excessive, address the query: "What type of Customer would you mentally trade for a one-off big deal?").

This "supermarket" holds a value of = 4, as illustrated in the example concerning problem 1. This signifies that one score falls within the range "approximately" USD 3,500.

And let the biggest of the real initial deals fall in the penultimate range to have the motivation to overcome this threshold.

Like this ones:

The final entry may contain a ‘jump’, such as an immediate increase to seven scores; landing within this range represents a rare and record-breaking event. Therefore, it should include a bonus.

Naturally, this table can be expanded into a more detailed version.

For example in the following way:

Step 3: Benchmarking the number of deals

We introduce a field to represent the benchmark (standard) number of deals for each range and overall. By multiplying the value of each range by the target number of deals within that range and then summing these, we arrive at the benchmark for income (expressed in scores).

Here it is:

If we now examine the correlation between the actual and the benchmark income, we can derive the performance in terms of income.

Since we established a benchmark by considering the typical initial deals (that is, we do not link the award to the size of the deal but instead categorise the reward into ranges – see Table 6), the issue of disproportionality in specific deals is addressed.

The introduction of ranges and the lack of a direct correlation between the evaluation and the monetary amount enable us to smooth out the variations, even for a statistically small number of deals (for example, a dozen or fewer). This approach is not feasible when outcomes are measured purely in money or by the number of deals in units. This was precisely the significant challenge.

Now, the motivation is maintained, but the cause of problems 2 and 3 are eliminated.

Problem 4. How to ensure it is properly deferred

A. Conditions

What should be done in situations where it is crucial that the agent not only aims to secure a significant income from a newly found Customer, but also endeavours to negotiate a contract under more favourable terms for the company with deferred payment? Signing a contract with a delay of, for example, 30 days as opposed to 90 days, is not the same. Naturally, many favourable contracts have standard measures to prevent delays, but the agent attempts new strategies to convince the management to deviate from the norm and offer relief on the grounds the grounds that "the Customer wants to try", etc. The company, on the contrary, believes in the principle: "The first time – with prepayment".

B. Traditional incorrect answers

A wrong response involves efforts to combat the rational self-interest of the employee, to cultivate a "team spirit", and to introduce bonuses based on "the net financial results of the entire company" (often with the bonus portion of the salary divided into "individual" and "company-wide"), to create affinity for the company, etc.

The drawbacks of incentivisation based on "company-wide efforts" are well-documented. What is general is not personal. The employee's contribution becomes diluted in the "overall performance", which is influenced by a broad range of factors, many of which are beyond the employee's control. Furthermore, "overall performance" is often measured differently from the work performance of the employee, leading to a disconnect between their effort and their reward.

C. A step towards the correct answer

Let's identify the root cause of the problem. It is easier for the agent to secure a deal with a longer delay. To secure a deal with only a short delay, they must exert more effort, yet the value of the deal remains unchanged. This indicates that the root of the problem is the lack of additional rewards for securing deals with shorter delays. The criterion of "deal value" alone is insufficient. Let's address this issue.

In the table below:

The ranges of the allowed deferments are provided;
An evaluation of each range is made (the number of days deferred), expressed in scores.

The scores are attributed using the following principle:

а) The worst situation (for example, when the goods are simply taken for sale and/or the threshold beyond which there is only failure) is denoted with 0.

b) The level of deferment which is acceptable in that market is denoted with 1 point (i.e., it is so frequent that it is inevitable to consider).

c) For an immediate payment (without deferment) we denote the maximum number of points (as in the example 1,5)

d) Between these critical points are given the intermediate points. (You can set the level of differentiation (the size of the range) as you deem right.)

We add the field for the benchmark number of deferments for each range and as a whole.

Multiplying the range score by the benchmark number of deals (in the context of deferments) in that range and summing the results, we calculate the overall benchmark (in scores).

Here it is:

The correlation between the factual deferments (in scores) and their benchmark values demonstrates the performance of income timeliness.

One might ask, "Why have we aligned the deferments within the '0 to 1.5' range? These values appear markedly different compared to the 'Scores for Types of Customers' and 'Ranges of Deals'. Specifically, why was the maximum value set at 1.5 rather than, say, 10?"

The rationale is as follows: It is crucial that managers are not incentivised to prioritise a high volume of deals involving small amounts but with no deferments. However, if a deal without deferments were to be rated, for instance, at 10 scores, such bias would undoubtedly arise.

Our objective is to boost the number of profitable transactions involving payment deferments within reasonable limits, rather than to categorically reject deferments. For example, we would not decline a 45- or 60-day payment deferment for an international chain making bulk purchases in substantial quantities.

This explains why the importance of deferments, relative to other metrics, has been reduced.

Final decision and methodological summary

We shall combine the four challenges mentioned above into a single salary framework below.

1. The incorrect answers mentioned above arise from a cognitive error:

Rather than identifying and addressing the root cause of the problem, various "compensatory and palliative measures" are suggested. Needless to say, they are ineffective. For instance, if the problem is the absence of a linear relationship between an agent's efforts and the transaction price, then adjusting coefficients to "ameliorate the situation" or boosting "team morale" are futile endeavours. When the issue arises from multifactorial causes, it is impractical to disregard these factors in favour of a single, "unified" metric.

2. In all four cases, we undertook an identical conversion (applicable to different variables: "number of Customers", "income", "payment deferments"; there could be other variables as well): we established ranges, thereby transitioning from a linear system (which created the root of the problem) to a discrete one.

3. Let us now merge the criteria identified in section C of each task. These challenges all pertain to one role: the "agent engaged in active sales". We will develop a model to evaluate their performance, encompassing all the cases we have examined.

Below are our Reference Tables, which we have already populated.

Reference Tables

Customers

Income

Deferments

Let us assume that our agent has secured 14 deals with new customers in a month. We have recorded the result in the table below under "factual sales". For example:

In this manner, he totalled 88 scores, and his performance is 80,96%:

We will then determine the salary amount we are willing to pay for 100% performance.

According to F. Taylor’s rule (‘Reward for Performance’), it should be slightly higher than the average actual salary of the same specialists in the labour market of a given city. Let’s call it 'Salary Including Reward' or S.I.R. And let's remember that S.I.R. level is higher than average market salary.

Now we can calculate the actual salary of our agent like this: Salary = S.I.R. * P.

In our example, S.I.R. * 80.96%. This will be the agent’s salary in our example.

See also the 'Minimum Performance' article.

Source of examples

The model was based on the statistics and long-term work of the authors in developing sales departments:

in wholesale companies of various profiles;
in companies involved in equipment sales;
in companies providing services to corporate clients;
in companies delivering dental materials;
in companies engaged in the sales of souvenirs made from expensive materials and jewellery;
... the list can be continued ...

Designing 'smart salaries'

PreviousThe article about salary NextNine times you can't

Last updated 1 year ago

Was this helpful?