The Line Nobody Dropped
A manufacturer rarely learns that it has been dropped, partly because in most cases nothing has been dropped, and there is no letter, no clause invoked, and no meeting where anyone makes a decision that would have to be written down. What happens instead is a run of small reallocations that nobody in the distributorship experiences as a decision at all, beginning with a quotation that goes out carrying a competing part number in the first position and the brand’s product beneath it as the alternate, continuing through a stocking level trimmed at an inventory review and never restored, and arriving eventually at a training class cancelled for low enrollment after 2 people signed up, one of whom was the product manager who asked for it. The account stays open through every bit of that, the agreement is still signed, and the distributor, asked directly, will say the relationship is fine.
What has changed is nothing a contract would recognize. Allocation is what changed, meaning whose product gets offered first when a customer calls about a motor that failed on second shift and holds no particular preference about what replaces it, and by the time allocation shows up as a number, the regional manager has written two quarters off to a slow market. Lines die this way far more often than they die by decision, in an accumulation of small reallocations that nobody would defend individually and nobody is ever asked to.
The Clock Nobody Started
There is a rough five year clock in this business, more industry anecdote than measurement, and anyone who has appointed a distributor or taken on a brand recognizes its shape. A manufacturer appointing a distributor allows something like five years to learn whether the appointment worked. A distributor taking on a line, particularly one meant to displace an incumbent, allows about the same. Neither party writes the date down, neither says in advance what success at the end of it would have to look like, and neither schedules the review that would test it against anything.
So what arrives at year five is not an evaluation but a ratification, in which a decision that accumulated gets treated as a decision that was made, and the reasoning gets built backward from whatever the number happens to say. Path dependence works exactly this way, through an event that permanently changes the trajectory of a system at the expense of the adjacent destinations it closed off, and the closing happens long before anyone notices that a choice was available. The questions worth asking in this article are all answerable in year 2, and in my experience of watching both sides of these relationships they are rarely asked before year 6, by which point the answer has stopped being actionable.
Two Years Against Five
The reason the clock never gets checked shows up in the tenure of the people running it. In a hermeneutic study of industrial distribution leaders across three channel groups, distributors, independent rep agencies, and factory direct reps, I found that “the factory direct replication is 3 months to 2 years in measured time, whereas independent reps and distributors measure tenure in decades” (Tolbert, 2022). The subjects also settled on what happens at each turnover, which is “the abandonment of initiatives by the new replicant in a 2-year cadence,” driven by a narrow window in which a new brand-side leader has to produce something visible enough to move up.
Set a 2-year replication against a 5-year clock and most of what follows in this article stops being mysterious. Nobody on the brand side is present for a full clock, which means the clock is never reviewed by anyone who was there when it started, and the initiative that would have justified the appointment gets abandoned twice before the appointment comes up for judgment. A distributor who has spent decades in one market is negotiating with a counterpart whose horizon ends before the first specification cycle closes, and the two parties are not running the same relationship at all, even while both believe they are.
Most of what follows in this article is downstream of that asymmetry. A channel manager who needs a visible result inside 2 years appoints a second distributor, because appointing is fast and specification work is not. The line fossilizes into reorders, because new specification takes longer than any sponsor of it stays in the seat. A displacement splits and never finishes, because displacement needs the whole clock and its brand-side sponsorship turns over twice inside it. And the complaint that would have corrected any of this never gets made, because the person who would have received it has already gone.
Neglect Has a Name, and the Research Has Had It for 50 Years
Hirschman (1970) described two responses available to a member of a declining organization, exit and voice, with loyalty conditioning which one gets used. Rusbult, Zembrodt, and Gunn (1982) added a fourth that the original pair had left out, neglect, meaning the passive allowance of deterioration, in which the party stays in the relationship while quietly withdrawing effort from it. Ping (1993) carried the whole typology into marketing channels and tested it on retailers in a wholesale relationship, where the structure held, and Ping (1999) then examined the antecedents of exit directly and reported that exiting carries antecedents beyond satisfaction and the cost of leaving, specifically loyal behavior, voice, and relationship neglect. Firms in a channel relationship respond to dissatisfaction roughly the way people respond to it, which is that most of them neither leave nor complain.
My own findings name the same behavior from inside the channel rather than from the survey literature, and name it more precisely. Manifested agency loss, in the sense Bendickson et al. (2016) give the agency problem, results from the defection of one or more channel groups, where the defection is a failure to serve the distribution promise and the policies, real or perceived, that carry it (Tolbert, 2022), and neglect is that failure committed passively rather than committed outright. A “brand is a promise,” as one distributor in that study put it, which means a promise can be broken by nobody doing anything in particular, and that is the form of breach no agreement anticipates and no scorecard records. The commitment and trust literature, running from Anderson and Narus (1990) through Morgan and Hunt (1994), treats these relationships as investments that compound or decay rather than as transactions that clear, which is the frame the neglect finding requires and the frame an annual score quietly denies.
There is a second construct from that study worth holding onto here, because it governs the brand-side half of this article. Liminality, in a distribution context, is the gray zone in which a leadership decision is permissible under the agreements and is still experienced as a defection by the other party. Most of what wrecks these relationships lives in that zone rather than outside it, which is why both sides can behave defensibly all the way to the end.
What Neglect Looks Like From Inside a Distributorship
Neglect has no signature at the brand level, which is why the brand cannot see it, and a very clear signature one layer down in the ordinary conduct of people who are not thinking about the brand at all. It looks like an inside salesperson who has stopped offering the line by default, not out of grievance but because the configurator needs a login that expires while the competing brand’s price already sits in the system, and like an applications specialist who solves the difficult ones with whatever product returns an answer the same day, since the customer is on the phone now and the brand’s support queue is measured in business days. A branch stocks one size deep instead of three after two returns took a quarter to credit and the branch manager is measured on working capital. A quotation carries the brand, correctly priced, in the second position, which is a position nobody buys from.
Every one of those is a competent person working well inside the constraints they actually face, and that is the part manufacturers consistently misread. What looks like disloyalty from the brand’s side is local rationality in a building that carries 15 other brands and has no particular reason to spend its scarcest resource, the attention of the 3 or 4 people who genuinely influence what gets quoted, on the line that costs the most effort to transact. One rep agency principal in the study put the underlying condition more bluntly than any distributor would: “A distributor doesn’t have the brand loyalty. They live and die by the customer. They throw the vendor under the bus as fast as possible” (Tolbert, 2022). Whether that is fair matters less than that it describes the incentive correctly.
The Brand Usually Goes First, and Calls It Coverage
The article to this point has treated neglect as something the distributor does, and in a large share of cases the distributor is making the second move rather than the first. Brands withdraw the same way, quietly and without notification, and in my study the majority of agency loss was attributed by the subjects to brands and their factory direct reps, in decisions “prone to path dependence” whose unintended consequences go unconsidered “until they manifest in an often-negative eruption equated as a loss” (Tolbert, 2022).
The brand-side version of the same withdrawal has a recognizable inventory to it. A second distributor gets appointed in the territory and nobody calls to say so, or the competitor two exits down gets signed and the distributor learns about it when a customer mentions who else quoted the line. Factory leads that used to route to one partner start routing to another. The application engineer’s visits thin, the training allocation moves, the demo equipment and the co-op dollars follow it, and a national account carve-out removes the one piece of business that justified the inventory position in the first place. Every item on that list is permissible under the agreement, which is exactly what makes it liminal, and exclusivity has been close to a dead letter in this industry for years, with very little protection attached to the word.
What the brand experiences as coverage, the distributor experiences as being passed over, and the study’s subjects were considerably more emotional about that than about pricing or terms. One rep principal, describing a brand handed to somebody with no history in the category, said, “You got guys that you look at my line card. It’s very, very similar products… . ‘Why is he handling motion control?’ He knows nothing. And that [angrily sets] me off because I had, I had had the servo stuff for years before” (Tolbert, 2022). Hibbard, Kumar, and Stern (2001) studied exactly this class of event, supplier acts that reconfigure a channel, and found dealer responses conditioned by the quality of the relationship preceding the act, with consequences carrying into dealer performance afterward. The history is part of the mechanism rather than the background to it.
The distance between the parties makes all of this worse, and the study found that distance measurable in focus. The distributor sits closest to the customer and carries the least specific loyalty to any single brand. The factory direct rep sits farthest from the customer with singular focus on one. That gap produces genuine incomprehension in both directions, captured best in a story one rep principal told about riding back from a successful call with the distributor in the car: “And the German guy says, ‘So explain to me, why do we need the distributor?’ With the distributor in the back seat” (Tolbert, 2022). Nothing in that moment violated a policy, and it almost certainly cost the relationship more than any argument about rebate tiers ever did.
Once both sides have withdrawn, each reads the other’s withdrawal as the cause rather than the response. The brand sees soft volume and concludes the distributor stopped hunting. The distributor sees thinning support and concludes the brand has moved on. Both conclusions are reasonable from where they are standing, both are wrong about the sequence, and the loop reinforces without anyone raising it out loud.
The Line That Never Grows
The failure mode that defeats every instrument in this article is not the drop at all. It is the line whose sales are acceptable and have not moved in years, where the same 15 or 20 end users reorder because the product is already in their plants and nothing new has been specified in a long while. Nobody in that building is refusing to sell the brand, and what they are doing instead is taking orders on it, which looks identical from the brand’s side of the ledger and is a completely different activity.
Read that line with the tools most brands have and every reading comes back clean, starting with revenue that is flat but positive and so triggers nothing. The recommendation question comes back high, because the people answering genuinely think well of the brand and would tell a peer to carry it. Product quality comes back high for the same reason, and even ease of doing business can score acceptably, since reordering a known part number is the simplest transaction in the building. The condition is invisible to a satisfaction survey and invisible to a revenue report, and it shows up only in the composition of the revenue.
Composition is what detects it, and composition is not a survey question but a query against the revenue, answered by the age of the accounts behind a line, the concentration across them, and the direction both are moving. Where 80 percent of a line’s volume sits with the same accounts it sat with 3 years ago, and the concentration is tightening rather than loosening, the line is an annuity against an installed base and it is decaying at exactly the rate that base gets replaced. A brand can compute this wherever the distributor reports point-of-sale data, and most never look, because the total is fine. One brand-side subject in my study described a large national partner in exactly these terms, recalling a time when “the distributor really did impact sales” and observing that “the chain stores now don’t have an impact on sales. They’re essentially order takers,” even while the brand’s internal politics still treated that partner as the account that drove whether the region was up or down (Tolbert, 2022).
The Pareto structure behind it is well understood inside distribution and rarely applied to brands. One distributor principal in the study extended the customer concentration rule to the line card and drew a boundary from it: “if you allow a manufacturer to gain more than 10 or 15% market share within your mindshare, within your distributorship, then the manufacturer owns you because nobody has more than 10% net on a consistent basis” (Tolbert, 2022). Mindshare is the resource being allocated, not shelf space, and the distributor who said that was defending against capture rather than complaining about it.
What holds a fossilized line in place, finally, is not performance but belief. Brands accumulate what the study treated as memetic force, where the brand owns a position of meaning regardless of the accuracy or the age of the meme, and the subjects all recognized the old proposition that nobody ever got fired for buying the dominant brand in the category. An installed base protected by that belief keeps reordering long after the reason has expired, which is comfortable for the incumbent, fatal to any challenger, and invisible to both.
The Split That Kills Two Lines at Once
A distributor that takes on a new brand to displace an incumbent, and then never finishes the displacement, ends up carrying both. This is the worst of the four conditions in this article, not because the damage is larger but because leadership is being actively reassured by two numbers that are each defensible.
The incumbent holds flat, which reads as stability. What it has become by then is the annuity described above, the installed base reordering while nothing new is specified, wearing the appearance of a stable account. The challenger grows quickly in percentage terms, which reads as momentum. What it actually is, is a small absolute number off a zero base, growing at a rate that will never reach the volume where the line sustains itself without dedicated attention. Put both readings in front of a leadership team and the conclusion is that the transition is working and needs more time.
Underneath, the constraint is that displacement runs on the attention of very few people, usually the specialist and whoever in applications can be trusted with the hard ones, and that attention does not divide cleanly. Half of it aimed at each brand does not produce half the conversion on each. It produces something worse on both, because the conversions that decide a displacement are the ones where somebody learns a second configurator, defends an unfamiliar part number to a maintenance manager who has run the incumbent for 15 years, and personally absorbs the first warranty argument. Very few people will do that twice at the same time. Meanwhile the distributor carries two inventories, two price files, two sets of training obligations and two factory relationships to service, and none of that cost lands on either brand’s margin line, so none of it enters the review.
Then both manufacturers look at the same card, see a partner splitting attention with a competitor, and begin hedging on the same schedule, which starts the withdrawal loop on two fronts simultaneously. Year five arrives with two half lines, two brands appointing around the distributor, and a leadership team that still believes the transition is in progress. Nobody made an identifiably bad decision at any point, which is what makes it maddening and why it goes unexamined for so long.
Why the Satisfaction Survey Sees None of It
Most manufacturers do measure their channel, and the instrument is nearly always some version of the single recommendation question that Reichheld (2003) popularized, occasionally with a satisfaction item and a comment box attached, run annually and reported as one number. The academic record on that number has been unkind for a long time. Keiningham, Cooil, Andreassen, and Aksoy (2007) tested the growth claim longitudinally in the industries Reichheld had cited as exemplars and failed to replicate the superiority of net promoter over established satisfaction measures at predicting revenue growth, while Bendle, Bagga, and Nastasoiu (2019) revisited the whole episode as a study in what happens when a practitioner metric outruns its evidence. Jones and Sasser (1995) had already reported the result that should have settled the question before the metric existed, which is that satisfied customers defect at rates their satisfaction scores give no warning of. The measure survives because it is cheap, comparable, and easy to put on a slide.
In a channel the problem is sharper than the general critique, in four specific ways. The question asks for an opinion rather than a report of conduct, and whether a distributor would recommend a brand and whether that distributor’s inside sales group led with it last Tuesday are different facts about the world, only one of which shows up in the revenue. One number cannot separate dimensions moving in opposite directions, so a brand admired for its products and avoided in daily transaction produces a figure that stays stable while the thing underneath it changes shape. The survey also reaches the wrong seat, since it typically goes to a principal or whoever signs the agreement, and that person knows the relationship at the level of terms, rebates, and quarterly reviews rather than at the level where the quote gets built. Finally, the survey is distributed by the party being evaluated, and when a brand’s own channel management sends the link, the respondent understands exactly who reads the answers and what the rebate program is worth.
If the standard instrument reaches the wrong person, asks for a feeling instead of a behavior, and collapses opposing dimensions into a single figure gathered under conditions that discourage candor, what is a brand learning when the score holds steady? It is learning that the score held steady.
Level and Correlation Answer Different Questions
A technical objection gets raised whenever someone proposes measuring a channel on more than one dimension, and it deserves a direct answer, because the answer is where the diagnostic value lives. The objection is that the dimensions are redundant, since asking a group of distributors about product quality, support, pricing, ease of doing business, and willingness to recommend produces five items that correlate, and running the standard reduction returns a single general factor accounting for a large share of the variance, at which point the conventional conclusion is that four of the five items are not earning their place.
The conclusion confuses two different properties of the same items. Correlation describes whether the items rank respondents in the same order, while level describes where the scores actually sit, and those properties are independent. Two dimensions can rank the same firms in nearly the same order while sitting an enormous distance apart in absolute terms, one strongly positive and the other deeply negative across the same population, so the correlation reports that they behave alike while the level reports that one of them is a problem. Action follows the level, because the level is where the detractors are.
In an exploratory study of one manufacturer’s distributors that I conducted and that is now under review, the pattern ran exactly that way. Willingness to recommend and product quality both scored strongly positive. Pricing and ease of doing business both scored net negative among the same respondents, in the same administration, with ease of doing business the lowest of the five and a majority of respondents falling in the detractor band on it. Support scored negative as well, though its reading stops being distinguishable from zero once the nesting of respondents inside distributors is accounted for, which is the correction a small number of participating firms forces and which the paper reports. Ease of doing business was also the second strongest predictor, behind product quality, of whether a distributor would recommend the brand at all. The single recommendation question, run alone on that population, would have returned a healthy number and reported a healthy channel, and none of the negative dimensions would have appeared anywhere. The study covers one brand, one administration, and a selected group of its partners, so it demonstrates that the failure is possible rather than measuring how common it is.
That ease of doing business is simultaneously the weakest score and among the strongest drivers is consistent with what Dixon, Freeman, and Toman (2010) reported on the customer side, where reducing the effort required to transact predicted loyalty better than attempts to exceed expectations. A channel partner allocates toward the brand that is easy to sell, and the research on both sides of the question keeps arriving in the same place, which is not the place most channel dashboards are looking.
Whose Answer Is It
A second design failure compounds the first, since channel surveys almost always blend respondents, so owner, outside sales, inside sales, and applications go into one pool and the result gets reported as the distributor’s view, when there is no such thing as the distributor’s view. An owner judges a brand on margin, terms, rebate structure, and whether the relationship earns the inventory commitment it asks for, which is a different question from the one outside sales answers about whether the brand wins at the customer, and narrower still than what inside sales answers about whether a quotation can be produced without a phone call, or what applications answers about whether anyone picks up when the sizing gets difficult. Those assessments routinely point in different directions for the same brand, and averaging them yields a figure that describes nobody in the building.
The role split is a design judgment rather than a tested result, and the reasoning behind it is that neglect is role specific. A brand can stand in good order with ownership while quietly declining with inside sales, which is the most dangerous combination available, because the people who would say something are content and the people reallocating are not being asked. The fossilized line makes the clearest case for it. Inside sales will report that line as fine, since they touch it constantly and it works, while outside sales and applications will report something else, since neither has put it into a new specification in years and both know exactly why. When two roles diverge that far on one brand in one building, the gap is the signature, and a blended score erases it by construction. Splitting roles also demands discipline about small cells, because a role with 2 respondents in a branch is not a finding, and an honest report says so rather than dressing noise up as signal.
What Detection Requires
None of this argues for surveying more, and it argues instead for measuring a handful of things properly rather than one thing conveniently, and for pairing what the survey can see with what only the revenue can. Composition answers the annuity question, since the age and concentration of the accounts behind a line will separate an installed base reordering from a channel selling, and no instrument administered to people will substitute for that. Separating the roles puts the remaining questions where the decision gets made rather than where the agreement was signed. Reporting the dimensions at their own levels keeps a strong one from concealing a weak one, which is the exact failure a composite produces. Candor has to be structural rather than hoped for, through administration by a party with no stake in the answer and reporting rules that make individual responses unrecoverable, since a partner who depends on a brand has every reason to be diplomatic on that brand’s own survey. And because a single reading cannot distinguish a low score from a falling one, the whole exercise has to repeat on a schedule before neglect becomes visible as the trend it is.
That specification is what the Channel Health Index was built to. It separates the roles, reports the dimensions at their own levels rather than as a composite, and is administered by a party with no stake in the answer. The research behind it was not run that way, since the study’s links went out through the brand’s own channel management, and the paper names that as a limitation rather than offering it as a model. The instrument matters less than the specification behind it, and the specification exists because a brand that only measures whether its partners like it has no way of knowing whether they are still selling it, while a distributor that only reviews its line card when something goes wrong is managing a portfolio by exception.
In Finality
The distributor who stops selling your line is not angry with you, and in most cases has not thought about you much at all, which is the actual condition under discussion and the reason the satisfaction score stays healthy while the volume goes. Neglect asks nothing of the person doing it, requires no decision, generates no conversation, and leaves no record until the record is a revenue chart. The same is true in the other direction, where a brand reallocating leads and attention toward a newer partner is not punishing anyone and is simply solving for a 2-year window inside a 5-year commitment that nobody wrote down.
The bill arrives at a line review where the brand goes undefended because nobody in the room has a recent reason to defend it, or at a channel meeting where the appointment gets rationalized by a number that was never the point. By then the questions worth asking are 3 or 4 years old. They were answerable the entire time.
Talk to us
If you sell through distribution. The CHI network wave reads your channel across all 5 dimensions, split by the roles that decide what gets quoted, administered independently of your channel management. What it costs, and the terms.
If you carry the lines. The CHI Line Review runs the same instrument on your own top brands, with your people as the respondents, and puts the result in front of your leadership in a 2 hour session. If the question is whether to add a line rather than how the current ones are doing, the New Brand Analysis is built for the front of the clock rather than the end of it. For distributors, or for repair, service, and integration.
References
Anderson, J. C., & Narus, J. A. (1990). A model of distributor firm and manufacturer firm working partnerships. Journal of Marketing, 54(1), 42-58.
Bendickson, J., Muldoon, J., Liguori, E., & Davis, P. (2016). Agency theory: The times, they are a-changin’. Management Decision, 54(1), 174-193. https://doi.org/10.1108/MD-02-2015-0058
Bendle, N. T., Bagga, C. K., & Nastasoiu, A. (2019). Forging a stronger academic-practitioner partnership: The case of Net Promoter Score (NPS). Journal of Marketing Theory and Practice, 27(2), 210-226. https://doi.org/10.1080/10696679.2019.1577689
Dixon, M., Freeman, K., & Toman, N. (2010). Stop trying to delight your customers. Harvard Business Review, 88(7/8), 116-122.
Hibbard, J. D., Kumar, N., & Stern, L. W. (2001). Examining the impact of destructive acts in marketing channel relationships. Journal of Marketing Research, 38(1), 45-61.
Hirschman, A. O. (1970). Exit, voice, and loyalty: Responses to decline in firms, organizations, and states. Harvard University Press.
Jones, T. O., & Sasser, W. E. (1995). Why satisfied customers defect. Harvard Business Review, 73(6), 88-99.
Keiningham, T. L., Cooil, B., Andreassen, T. W., & Aksoy, L. (2007). A longitudinal examination of net promoter and firm revenue growth. Journal of Marketing, 71(3), 39-51.
Morgan, R. M., & Hunt, S. D. (1994). The commitment-trust theory of relationship marketing. Journal of Marketing, 58(3), 20-38.
Ping, R. A. (1993). The effects of satisfaction and structural constraints on retailer exiting, voice, loyalty, opportunism, and neglect. Journal of Retailing, 69(3), 320-352. https://doi.org/10.1016/0022-4359(93)90010-G
Ping, R. A. (1999). Unexplored antecedents of exiting in a marketing channel. Journal of Retailing, 75(2), 218-241.
Reichheld, F. F. (2003). The one number you need to grow. Harvard Business Review, 81(12), 46-54.
Rusbult, C. E., Zembrodt, I. M., & Gunn, L. K. (1982). Exit, voice, loyalty, and neglect: Responses to dissatisfaction in romantic involvements. Journal of Personality and Social Psychology, 43(6), 1230-1242. https://doi.org/10.1037/0022-3514.43.6.1230
Tolbert, C. L. (2022). A hermeneutic study of industrial distribution: The nuanced understanding of organizational fitness in the context of complex systems and memetic culture [Doctoral dissertation, Columbia International University].
The instrument this argument produced. Channel Health Index