What is key account management?

What is key account management?

It is what a supplier does differently for the customers it treats as most important, and who does it. Workman et al. (2003)³ define it this way: “We define key account management (KAM) as the performance of additional activities and/or designation of special personnel directed at an organization’s most important customers.” The definition has two conditions: some identification of the most important customers, and extra activities, special personnel or both directed at them. KAM is therefore absent in two cases, when no customer is treated as more important than the others, and when some are but nothing different is done for them. A formal programme is not one of the conditions: “If the CEO, senior managers, or managing partners are responsible for the firm’s most important customers, they are involved in KAM.” On this definition, the authors add, most firms are involved in KAM “whether they formally recognize it or not”.

The practice goes by several names. Homburg et al. (2002)² count as KAM the approaches to the most important customers that earlier work discussed as key account selling, national account management, national account selling, strategic account management, major account management and global account management, and they note that “KAM appears to be the most accepted term” in recent publications. Among the US firms in the survey reported below, the labels most often used for the most important customers were key accounts (46 per cent), major accounts (14 per cent), strategic accounts (12 per cent) and national accounts (11 per cent).

In this guide, key account management means the extra activities a supplier performs and the people it designates for the customers it treats as most important, together with the way it organises both. A key account programme is the formalized form of it, and a key account manager is one of the people it can designate. Two things are left out of it. One is how the key accounts are chosen: Workman et al.³ note that their definition “does not include the process or criteria by which firms select their most important customers”. The account score essay argues that a key-account score allocates scarce attention, service and access, and should show the resource it allocates, the buyer evidence behind it, the strategic exception that applies and who can challenge it; this guide starts once the accounts are chosen. The other is a single customer relationship seen on its own: Homburg et al.² set their level of analysis at “the KAM approach (which encompasses relationships with several important customers)”, not at one buyer and one seller.

Does key account management need a formal programme?

Not on these definitions. Homburg et al. (2002)² observe that earlier conceptual and empirical work had looked mainly at formal key account programmes in Fortune 500 companies, and they treat the choice as a design question in its own right: “We claim that to formalize or not to formalize the key account approach represents a decision dimension of its own.” Their advice to managers is explicit: “We particularly emphasize that managing key accounts does not necessarily require setting up a formal key account program.” Formal programmes were nonetheless common in their data: according to a note in Workman et al. (2003)³, more than half of the randomly sampled firms in the same survey had one.

The underlying problem is organisational. In Homburg et al.²’s account, the demands of important customers raise an organisational design problem for many suppliers: internal structures often hamper coordinated account management, and the set of activities for complex customers “cannot be handled by the sales function alone but requires” participation from other functions. Citing Spekman and Johnston, they also name one fundamental problem for sales managers: to “obtain the cooperation of other organizational members without having formal authority over them”. The account-based marketing guide covers a related go-to-market model, in which marketing, sales and executives treat a tier of high-value accounts as individual markets; this guide is about how a supplier organises itself for the customers it treats as most important.

What does a supplier decide when it designs KAM?

Four things, in Homburg et al.²’s conceptualization. Earlier research, they write, can be summarised in three questions, “(1) What is done? (2) Who does it? and (3) With whom is it done?”, and they add a fourth: “(4) How formalized is it?” They call the four dimensions activities, actors, resources and formalization, and define a small set of constructs for each, which their survey measured on seven-point scales:

• Activities. Activity intensity is “the extent to which the supplier does more for key accounts than for average accounts”, in products, services, prices, logistics, information sharing and promotion; activity proactiveness is “the extent to which activities are initiated by the supplier”.
• Actors. Use of teams is “the extent to which teams are formed to coordinate activities for key accounts”; top-management involvement is “the extent to which senior management participates in KAM”.
• Resources. Access to marketing and sales resources, and to nonmarketing and nonsales resources, is the extent to which a key account manager can obtain needed contributions from those groups; the esprit de corps of the selling center is “the extent to which selling center participants feel obliged to common goals and to each other”.
• Formalization. It is “the extent to which the treatment of the most important customers is governed by formal rules and standard procedures”.

They also name the key account coordinator, “the person who is mainly responsible for coordinating activities related to key accounts”, who may hold a dedicated full-time position or carry the task part-time, and they define the outcome against the supplier’s own customers: KAM effectiveness is “the extent to which an organization achieves better relationship outcomes for its key accounts than for its average accounts”. Respondents rated it, against average accounts, on mutual trust, information sharing, a reputation of fairness, investments in the relationship, long-term relationships, fewer conflicts and sales targets.

Which forms of KAM did the survey find?

Homburg et al. (2002)² mailed a questionnaire to the head of the sales organisation at a random sample of 1,000 US and 1,000 German firms in five business-to-business sectors: chemicals and pharmaceuticals, machinery, computers and electronics, banks and insurance, and food and packaged goods. Respondents answered for their most important set of business customers, whatever they called them. After 174 US and 171 German firms were found unsuitable, 264 German and 121 US firms responded, “for effective response rates of 31.8% and 14.6%, respectively”. A cluster analysis on the eight constructs above sorted 375 of the firms into eight approaches:

• Top-management KAM (37 firms): highly formalized, with the most top-management involvement and extensive use of teams, run from headquarters; the authors write that it truly deserves the name programme.
• Middle-management KAM (76): what the authors read as a formal programme run at middle-management level, with high access to marketing and sales resources, rather weak esprit de corps and medium access to other functions.
• Operating-level KAM (57): many activities and standardized procedures, lower top-management involvement, and the highest share of dedicated sales managers as coordinators (70.8 per cent).
• Cross-functional, dominant KAM (44): the highest values on nearly all variables; in the authors’ words, “It seems that, in these companies, customer management is virtually identical with KAM”.
• Unstructured KAM (38): no programme and no special structures, with activities more reactive than proactive; the authors read it as firms “pursuing KAM on an ad hoc basis”.
• Isolated KAM (40): midrange values, 44.4 per cent of coordinators based locally, and low esprit de corps and access to functions outside marketing and sales; the authors’ overall picture is that “KAM is a rather isolated, local sales effort”.
• Country-club KAM (37): high top-management involvement with low values on most other variables; the authors take this to suggest that KAM there is “little more than representation by senior managers”.
• No KAM (46): the lowest values on nearly all variables; the authors read it as firms that do not manage their key accounts, or have only begun to.

Two of these, unstructured and isolated KAM, had not been described in earlier typologies; the authors write that they “involve a considerable number of activities for key accounts but do not require formalization of the approach”. The data also cut against the idea of one ladder from basic to advanced KAM, along which activities, teams and top-management involvement rise together: “high degrees of top-management involvement occur in combination with both high and low degrees of activity intensity”, and with both high and low use of teams.

Did the form of KAM make a difference with key accounts?

Among six of the eight forms, not by the authors’ test; the other two rated lower. Figure 1 shows mean KAM effectiveness by approach. Six approaches, from the highly formalized top-management KAM to the least formalized country-club KAM, were not significantly different from one another; isolated KAM and no KAM were lower. The authors conclude that “the same level of performance can be accomplished through different approaches”, and that on the key-account and the organisation-level outcomes alike, “the no-KAM and the isolated KAM approaches perform the worst”.

The differences between approaches on all four outcomes remained significant when the authors controlled for market dynamism and competitive intensity. Against competitors, firms using cross-functional KAM gave themselves the highest mean ratings on performance in the market, where three other approaches shared their band, and on adaptiveness, where none did; top-management KAM had the highest mean profitability, a pre-tax margin band reported by the same respondent. The authors note that the most effective approaches were not the most profitable ones and suggest that some approaches may involve higher costs as well as higher revenues. On profitability the bands overlap more: only top-management KAM was significantly above no KAM, and cross-functional, unstructured and isolated KAM could not be separated from either end. Table 6 therefore bears out the authors’ “perform the worst” on KAM effectiveness, performance in the market and adaptiveness, where no KAM and isolated KAM have the two lowest means, but not on profitability, where isolated KAM (5.23) sat above four other approaches.

The authors call the no-KAM result “the most comprehensive empirical demonstration so far that suppliers benefit from managing their key accounts”, and they take the results for no KAM and isolated KAM to suggest that failing to achieve access to and commitment of cross-functional resources “seems to play a critical role”. On this guide’s reading, the design supports a narrower statement. Each firm’s ratings come from one manager, at one point in time, comparing the firm’s key accounts with its own average accounts, and the profitability band comes from the same person, although the authors treat it as an objective measure that a key-informant bias should not reach. The survey shows which approaches went with higher ratings, not that adopting one would raise them.

Which features went with higher ratings?

Workman et al. (2003)³ used the same survey, with the same 264 German and 121 US responses, to ask which design dimensions go with KAM effectiveness, in a structural equation model estimated on 230 firms. The completely standardized coefficients in their Figure 2 are 0.55 for the esprit de corps of the key account management team, 0.27 for access to marketing and sales resources, 0.25 for activity intensity and 0.18 for activity proactiveness, all significant at the 1 per cent level; 0.11 for top-management involvement, significant only at the 10 per cent level; no significant coefficient for use of teams; and minus 0.13 for approach formalization, significant at the 5 per cent level. The model accounts for 68 per cent of the variance in KAM effectiveness.

On teams the authors write: “We fail to find any statistically significant relationship between use of teams and KAM effectiveness.” Their reading is that the team spirit of the people involved counts for more than the extent of team use, but they also note that they “did not measure whether key accounts are managed through permanent or nonpermanent teams”, and that the null result may reflect different kinds of team in the sample. On formalization the coefficient ran opposite to their hypothesis. Their explanation is that formalizing the approach to key accounts “may lead to bureaucracy and impede flexibility in responding to the demands of different key accounts”; a reviewer’s further explanation, reported in a note, is that “it may be the case that KAM programs are more formalized when powerful customers demand such programs”, and that, if such customers are also less profitable, this may be an additional reason for the finding.

The direction of all these associations is open, and the authors say so: “We are not able to establish direction of causality.” In their words, “It is possible that success with key accounts leads to designation of special personnel, additional resources, and a higher esprit de corps”, and customer power may be a missing factor. In the same model, KAM effectiveness went with performance in the market (0.64), and performance in the market with profitability (0.23).

The two articles look at formalization from two angles. In the taxonomy, a programme was not needed to reach the upper band of KAM effectiveness: the six approaches there include the two most formalized (cross-functional KAM at 5.64 and top-management KAM at 5.48 on the seven-point formalization scale) and the least formalized (country-club KAM at 2.12). But every approach that Homburg et al.² rate rather high to very high on formalization sat in that band, and both lower-band approaches were below the sample mean of 4.15 (isolated KAM 3.64, no KAM 2.72), so the taxonomy does not show that formalization made no difference. In the model, which holds the other dimensions constant, more formalization went with slightly lower ratings. On this guide’s reading, the two agree only that a formal programme was not a condition for higher ratings of key-account results in these data. Neither shows what installing or dismantling a programme would do. The two lower-band approaches also had the lowest selling center esprit de corps (isolated KAM 3.93, no KAM 3.82), measured with the same items as the variable that has the largest coefficient in Workman et al.³’s model.

What part does top management play?

On this guide’s reading of these data, a smaller one than the KAM literature had claimed for it. Workman et al.³ note that managerial and conceptual articles on KAM “have consistently called for top management involvement”, and they quote the director of the Strategic Account Management Association, as reported by Napolitano, calling top management sponsorship and involvement “the most critical indicator of success”. Homburg et al.² measure top-management involvement by agreement with three statements: that even small matters about key accounts are referred to someone higher up, that very few decisions about them are made without senior managers, and that top management often deals with KAM; the authors note that the construct is “close to the centralization construct” of organisation theory. In Workman et al.’s model it had the smallest coefficient that reached significance, and only at the 10 per cent level. Workman et al. read the same coefficient differently: they count it as support for their hypothesis, though “smaller in magnitude than the other statistically significant coefficients”, and advise that “managers should involve the top managers in their firm in KAM”, which they justify by the coefficient and by the signal top-management involvement sends to the rest of the organisation. They also write that, of the four dimensions in their model, “activities and resources are more important than actors and formalization”. In the taxonomy, high involvement appears in top-management KAM, which had the highest mean profitability, in cross-functional KAM, whose mean profitability lay between the two ends, and in country-club KAM, which shared the lowest profitability band with no KAM. Guesalaga and Johnston (2010)¹ summarise Homburg et al. as finding that approaches with high top-management involvement “tend to have higher profitability”; Table 6 bears that out for top-management KAM but not for country-club KAM.

Guesalaga and Johnston¹ content-analysed 64 academic articles on KAM published in 17 journals from 1979 to 2009 and 90 practitioner articles from Velocity, the magazine of the Strategic Account Management Association, from 2004 to 2008. The two sets differed significantly in their topics: practitioners wrote most about organizing for KAM and adapting KAM approaches to different customers, which academics had covered less. Two topics stood out as important to managers and little studied: “the role of senior management in KAM, and the importance of internal alignment in determining KAM success”. Internal alignment, the topic the practitioner articles discussed most, is the one they connect to Workman et al.³’s esprit de corps. On senior management, they find the survey measure too narrow: “the dominant conceptualization of top management involvement (see Homburg et al., 2002)² is one-dimensional”, reflecting only participation in decisions about key customers. The practitioner articles, they write, suggest that problems occur when top executives decide about a customer without enough knowledge of how it does business, or when senior managers “try to overtake the account during negotiations while diminishing the key account managers’ role as a primary contact”. These are observations from practitioners’ writing that the authors cite, one of them from a 2001 issue outside their 2004 to 2008 sample; they were not tested.

What should a KAM design record hold?

Six fields, in one row per group of key accounts handled the same way: the extra activities, who initiates them, the coordinator and what top management does, the contributions needed from other functions and whether the coordinator can obtain them, the procedures that are written down, and how results with these accounts compare with average accounts. The fields follow Homburg et al.²’s four questions, their activity proactiveness and their definition of KAM effectiveness; the contributions field also reflects access to marketing and sales resources, which had the second-largest coefficient in Workman et al.³’s model; that model measured access to field sales, customer service and product management only, and access to other functions appears in Homburg et al.’s taxonomy but not in it.

An illustration, with no company data: a B2B software firm treats its fifteen largest subscription customers as key accounts. Its record lists the extra activities (a quarterly review of planned use with product management, a named support engineer, one set of contract terms for all of a customer’s subsidiaries) and notes which of them the firm proposed and which the customers asked for. It names the coordinator and what the executive sponsor actually does, lists the contributions needed from product, support and finance and whether the coordinator can get them without escalating, and records which procedures are written down and why. It compares the fifteen with average accounts on the relationship outcomes Homburg et al.² measured and on renewal. It does not assume that a written programme, a standing team or more executive time will raise the results: in Workman et al.³’s model, formalization had a small negative coefficient, use of teams no significant one, and top-management involvement a small one, significant only at the 10 per cent level. These are the author’s examples and have not been tested.

What would a software seller still have to test?

Most of the numbers. Both empirical articles report one mail survey: one informant per firm, one point in time, firms in five sectors in Germany and the United States, and an overall response rate of 23.3 per cent. KAM effectiveness is the respondent’s rating of key accounts against the firm’s own average accounts, and profitability is a self-reported margin band. Homburg et al.² name “the static design of our study” and a single informant who sees only the seller’s side of each relationship as limitations; Workman et al.³ add that an unmeasured factor such as customer power may affect both sides of their model, and that success with key accounts may come first. Guesalaga and Johnston¹ count topics in what academics and practitioners wrote, using one association’s magazine for the practitioners’ view, and they acknowledge that its topics are not exhaustive; their review shows what was discussed, not what works.

On this guide’s reading, two things carry over to a software firm with some confidence: KAM is defined by what a supplier does differently and by whom, not by whether it has a programme, and its design involves four questions that can be answered separately. Whether the eight approaches, the small negative association of formalization or the weight of esprit de corps hold for software firms, subscription renewals and buying committees, the sources do not say.

References

1. Guesalaga, R., & Johnston, W. (2010). What's next in key account management research? Building the bridge between the academic literature and the practitioners' priorities. Industrial Marketing Management, 39(7), 1063–1068. https://doi.org/10.1016/j.indmarman.2009.12.008
2. Homburg, C., Workman, J. P., Jr., & Jensen, O. (2002). A configurational perspective on key account management. Journal of Marketing, 66(2), 38–60. https://doi.org/10.1509/jmkg.66.2.38.18471
3. Workman, J. P., Jr., Homburg, C., & Jensen, O. (2003). Intraorganizational determinants of key account management effectiveness. Journal of the Academy of Marketing Science, 31(1), 3–21. https://doi.org/10.1177/0092070302238599