There was a very interesting article in last Tuesday’s (November 10) edition of The Wall Street Journal. It is entitled, It Just Isn’t Working? Some File for Customer Divorce, and it relates how some small businesses are eliminating problematic customers.

In the paper edition, the article appeared under the “Small Business’ banner, and it now resides under a similarly-labeled section of the web site; however, it applies equally well to businesses of all sizes.

We must admit, though, that the topic is especially poignant for small businesses because the decision-maker and the implementer (executioner) are often one-and-the-same person, and the customer may be a friend or acquaintance. There’s no, “my boss told me that I have to do it” excuse when you are the boss. Note, however, that many seemingly-independent, entrepreneurial men still use their wives as reasons why something can’t be bought or sold at a given price. We would never admit to such behavior, unless our Chairman, Jill, permits us.

The main point of the article is that despite generally tough economic times, a few small businesses are finding that it is not worthwhile to deal with certain clients because those clients disproportionately consume time and resources given the revenue that they generate.

Discipline! Discipline!

Though we sound like Colonel Hathi from The Jungle Book, there are a couple of obvious ways to minimize the probability of enabling such problems:

1. Don’t appear too desperate or too needy for the business. (In that sense, it is a lot like dating.)
2. Evaluate the client, and have a sliding scale of prices ranging from lower prices for easy clients to higher prices for pains-in-the-butt or annoying ones, and stick to it.

Admittedly, such tactics may not be feasible with all products and services. They generally work better for services than products, and for customized services, rather than uniform or generic ones, and for short-term projects rather than long-term ones. However, unless there is a substantial benefit that can be derived from either existing or prospective clients because of the relationship with a tiresome one, the tiresome one should pay for the irritation, aggravation, discomfort, and effort that they cause. (We’ll qualify this a bit below.)

If such customers are not willing to pay, then the supplier or vendor is, in fact, subsidizing the opportunistic behavior. Moreover, if the supplier is not willing to sever the relationship (or take actions that will change the other party’s behavior and profitability), then we humbly recommend that they not complain to everyone around them (thereby making their employees’ and families’ lives more miserable than need be).

That may seem a bit harsh, but it is consistent with our response when someone complains about a spouse: divorce them, kill them, or shut-up. Revealed preference says that despite the complaints, if the vendor is unwilling to sever the relationship, then he or she must find the arrangement to be acceptable. Sometimes it is necessary to do unpleasant things–not immoral, unethical, or illegal, just unpleasant ones–to get what one wants out of life.

So, be a good stoic, and live with it in silence (or turn it into a blog post).

We prefer to maintain pricing discipline because at the margin, one’s willingness to lower the price or fee is often viewed as a sign of weakness. So, clients who make that inference then often request more time and/or resources. That means that a sign of weakness harms both revenue and resource-usage, especially for services that are provided over time. (Note that this problem can be mitigated somewhat by offering a lower-quality or less-robust substitute whenever someone wants a lower price, i.e., “we’ll lower the price, but you’re not getting…”)

By the way, another option is to write very, very detailed contracts, but those preparation costs only add to the resource consumed by such customers, and covering those incremental, transaction-related costs increase the necessary price to make the transaction “worthwhile.” (You can see why many folks open franchises where such development and preparation costs can be shared nationally or even globally.)

Two Costs to Consider

Before firing a customer who is thought to be ”unprofitable” be sure to consider a few factors to determine if the customer is truly unworthy.

Clearly the choice of firing or retaining a customer is a decision, or the selection (and implementation) of one possible alternative from several available alternatives. (Note that “do nothing” is often an available course of action.) So, rather than be concerned with the reporting of financial results,  or accounting, per se, we are interested in the costs and benefits that vary across alternatives, the relevant costs and benefits.

In particular, we are interested in

1. Expected Relevant  Benefits and Costs
2. Expected Opportunity Costs

Note that these costs may vary through time as circumstances and capacity utilization change. Technically, opportunity costs, which we’ll define below, can be categorized as relevant costs, but we think that they are worthy of their own place on the list.

1. Expected Relevant Costs (& Benefits)

In a decision, which involves the selection of one possible course of actions from several available options, the relevant costs of an alternative, like firing a customer, are the resources sacrificed by selecting and implementing that course of action. Relevant revenue (and/or benefits) is(are) the resources received by taking that course of action.

Generally, one thinks of such costs and benefits as being measured financially and reported in terms of dollars or some other currency, but they need not be. That’s true for small businesses where many sacrifices, including, say, staying awake an extra hour to work on a project, are difficult to measure financially.¹

For items that can be measured financially, the difference between relevant revenues and relevant costs is the contribution, and the contribution of an alternative can be very, very different than its reported profit. That’s because: (1) certain relevant items may not appear on an income statement the shows reported revenue and expenses and (2) certain reported revenue and expenses that do appear on an income statement may be irrelevant for the decision at hand.

So the total net benefit of an alternative is some measure of the difference between relevant revenues (and other non-financial benefits) and the relevant costs (and other non-financial sacrifices). We are ignoring the problem of combining dollar-denominated revenues and costs with non-financial and psychic benefits and costs since it is very subjective and goes far beyond the scope of this already-long essay. We would have to introduce a long recitation about utility functions and expected utility.)

Prospective, not Historical

Note that these costs are incurred prospectively–i.e., in the future, not in the past, and that means that they are uncertain or imprecisely known. That’s why we titled this section “Expected Relevant Costs:” to indicate that randomness. However for simplicity we’ll immediately drop the uncertain aspect and assume that we can precisely measure such resource sacrifices.²

Despite our various caveats and qualifications, we’ll try to be clear:  by “relevant” we mean the real incremental costs that are incurred by making a decision. For example, if the decision is deliver a good or perform a service,  then costs of rent, computers, software, and most non-manufacturing-related utilities usually don’t change whether an additional customer is served or not. Paper and ink or toner costs might change–if there is a lot of printing involved. Monthly health-care premiums won’t unless more employees are added and retained. Transportation costs will but only if more travel or transportation is required.

Arbitrariness of Reported Costs

By “relevant” we don’t mean necessarily mean the reported or accounting costs that are assigned or allocated to a customer or project on some arbitrary basis.  An accounting system, regardless of its cost, serves many purposes, e.g., to satisfy the IRS, for performance measurement, etc. When there are jointly-used resources and/or resources that last more than one accounting period, then all accounting systems require many, many assumptions and conventions. Those assumptions and conventions permit simplifications so that addition, subtraction, multiplication and division can be used to determine reported costs, but such accounting systems are very particular models (or abstractions or distortions) of history–of the reality that has happened, and do not consider (much) of the future.

Irrelevant Costs

For example, by definition, the allocation of fixed and/or sunk costs, e.g., depreciation, of shared resources is arbitrary. For long-lived assets, the allocation of its cost across time is arbitrary. For shared resources, the allocation across objects is arbitrary. That means that the reported cost depends upon the accounting methods applied and is an artifact of simple arithmetic calculations and not a reflection of behavior. (They’re fixed for Pete’s sake.)

So, while purchase costs of such shared resources can often be measured with high degrees of precision, there is no notion of accuracy when allocated to particular customers or products. There are generally an infinite number of reasonable and justifiable ways to make such assignments.

There are good reasons–behavior, tax, rhetorical–for selecting one allocation method over another, but none provides more or less information. Different methods may provide greater or fewer details and may provide different assignments across objects–different reported costs. That may or may not be useful for those non-informative purposes, e.g., minimizing taxes, but by its very nature, the arbitrariness of the method makes it uninformative. In other words, unless there is only one customer (in one accounting cycle), then shared-and-fixed costs can be assigned to customers in a variety of reasonable and appropriate ways, and one method isn’t better than another unless it helps attain a goal. Moreover, some allocation methods may make  a customer appear to be “profitable” or not–despite the fact that there would be no change in resource consumption if the customer is dropped or kept.

In the previous paragraph, we covered the equivalent of five or six sessions of a good managerial accounting course–like the kind we used to teach. The lesson for small businesses–and for all for that matter–is simple. Don’t try to be too sophisticated and believe that the output of your accounting system, the reported costs accumulated from past transactions or event, describe reality for any particular decision (set of alternatives). If there are shared resources–shared across objects or time–then there is no notion of true profitability.

That means that one shouldn’t imagine that there is some type of hidden accounting magic that permits vast simplification via various assumptions and conventions, yet somehow cancels out in just the right way to accurately inform, regardless of the decision. Now, if that sentence doesn’t make sense to the dear reader, and he or she is in charge of such decisions, then the reader should consider getting help because it is a very important notion.

Your accounting system output may be useful to help determine relevant costs of a decision, but when shared resources have fixed costs, know that your system is one of many equally-valid and simplified representations of reality–analogous to Picasso’s cubism or a two-dimensional cartoon animation of yesterday’s commute or the size and color variety of the pixels of a digital photo. With different assumptions, the reported costs of various objects will change; otherwise, the different assumptions would really be different. However, note that the underlying reality doesn’t change.

Costs reported in accounting systems can help identify some categories of relevant costs, but accounting systems capture only past transactions and don’t capture future sacrifices (and how those sacrifices change among alternative (future) courses of action). In addition, cost reports ignore opportunity costs, or benefits foregone by taking one action that precludes taking another, mutually-exclusive one. (But, we won’t ignore, them: see below.)

The consideration or inclusion of irrelevant costs–particularly the poorly-reasoned desire–to recover or recoup the allocated expenses related to fixed or sunk shared resources often leads to disaster, e.g., downward death spirals of demand. In such cases, the elimination of one “unprofitable” customer creates additional unprofitable ones in the future because they the costs may be much more rigid than perceived or reported.

Be sure that reported costs behave the way that you think they do–and are, in fact, relevant–before eliminating seemingly unprofitable clients.

2. Expected Opportunity Costs

In a decision, the opportunity cost of the selected course of action, is the benefit forsaken by not taking the next best alternative. It is non-negative, and if there is nothing better to do, it might be zero.

Unlike in our discussion of relevant costs, we’ll keep the modifier “expected,” because the next best alternative may involve attempting to gain business and revenue from new customers, and the outcomes of such activities are usually uncertain.

Technically, and non-financially, there is an opportunity cost to any course of action, and through time, as constraints vary, the next best thing to do changes frequently. (Often, however, that sacrifice has zero monetary value.)

Perhaps the reader visits this site from their office but never visits it from home. Why? As much as it hurts to admit it, perhaps the reader has estimated–possibly erroneously–that there may be more valuable or enjoyable actions to take at home than reading our opinions and expositions. That might not be for certain, but there may be a high enough likelihood and high enough pay-off to justify ignoring us in the evening.

Aside: given the various constraints that come and go through time, those readers with a bit of mathematical aptitude should be able to imagine the difficulty of modeling and quantifying these various opportunities…a seemingly infinite horizon, multiple decisions that affect resources and opportunities, the option to quit… ad infinitum. It gets ugly very quickly.

Expected to Operate at Capacity

So, if a business plans to operate at capacity and that capacity cannot be changed, then the expected opportunity cost of accepting a new customer is the financial contribution (and other benefits) associated with the next best alternative, i.e., the lost financial contribution(s) and benefits associated with getting rid of the customer(s) to free-up sufficient capacity to serve the marginal customer.³

Expected to Operating below Capacity

If a business plans to operate below capacity, then the expected opportunity cost is what the firm is likely to forsake by not pursing other customers. That requires the assessment of the likelihood of gaining new customers and the “distribution” of contributions from those new customers. For our mathematical and argumentative readers, note that we are using “expected” and “distribution” very loosely. Moreover, depending upon one’s risk preferences, the expected opportunity cost may be of little interest. (Strictly, the expected opportunity cost is interesting only to risk-neutral decision-makers.)

So, when deciding to drop a customer, the decision-maker should compare the financial contribution of that customer (and any other non-financial benefits) to his or her assessment of the probable and (the appropriately-weighted) improbable benefits of finding new customers during the time period of interest.

If there is a strong chance of gaining new, profitable business elsewhere, then drop the client and start your marketing drive. Part of that calculation should include the less-obvious, indirect benefits of each alternative. We refer to them as ancillary concerns, although their magnitude could be quite large.

Ancillary Concerns:

Side Benefits of Difficult Customers

1. Reputational Effects on Existing or Potential Customers
2. Learning and Experimental Learning

1.Reputational Effects on Existing or Potential Clients

Consider how severing the relationship with one customer may affect others. Is the client particularly prestigious or well-connected so that an association with them generates additional side business or benefits? Will severing the relationship harm that business or the prospect of gaining new clients? Or will it be easier to maintain higher prices with existing clients if they know of your pricing discipline and examples of dropped customers?

No action is taken in isolation; however, not every action generates observable or measurable implications. So, specify the various indirect outcomes of firing a customer and estimate the short- and long-term contributions and net benefits of such actions.

2. Learning

Note that there is an indirect advantage to dealing with and maintaining difficult customers. They provide valuable information about how to communicate with other, existing customers; how to market to prospects; and how to protect one’s time, sanity, and organization. They teach what to promise and what not to promise.

Similar to the notion that nothing is foolproof because fools are so ingenious, difficult clients expose the weaknesses in standard operating procedures and business methods and practices.

So, there is an informational value to the association, which may be maximized if the supplier behaves opportunistically, too. By that we mean experiment with actions and reactions to see how the difficult customer responds. Slow delivery, slow the return of telephone calls and e-mail messages, mention additional service charges for anything not specified by the contract, and consider the response.

We don’t recommend such tactics to be vindictive or vengeful, or passive-aggressive, or any other anti-social reason. We recommend it simply to discover new ways to use the relationship advantageously. If there is no financial advantage, then seek other advantages, and experimenting with a real, live (though unprofitable) client is one way to get valuable feedback that may help manage the expectations and relationships (and therefore the profitability) of other customers.

Alternatively, it may be the case that such customers are much more pliable than previously thought and that they are (or were) simply better negotiators.

In addition, the predicaments they cause that must be solved and those solutions may provide valuable technical know-how that can be applied elsewhere (or sold for profit).

We’ll likely add or edit this post in the near, future. If you would like to know more, please contact us.

Copyright © 2009 Spero Consulting.

---

Footnotes: