How to train supply managers – Necessary and sufficient purchasing skills leading to success

.


Introductiona lack of understanding on skills necessary for purchasers to fulfil their different objectives
The PSM (purchasing and supply management) function has evolved into a human-centric discipline in which human capital input is of strategic importance to firms (Bals et al., 2019;Hohenstein et al., 2014).Given the importance of the human aspect, it is not surprising that the contributions of PSM to a firm's effectiveness are considered to be strongly dependent on individual PSM professionals and their "skills" (Feisel et al., 2011;Knight et al., 2014;Tassabehji and Moorhouse, 2008).(In general, the PSM literature mainly uses "skill" or "skills" to refer to the separate competence construct items, which, according to Delamare-Le Deist and Winterton (2005), include knowledge, professional skills, interpersonal skills, intrapersonal traits, behaviours and attitudes.For brevity, this study also uses the term "skill" in cases referring to different items of the competence construct.) As the field of PSM is evolving and professionalising, it is also differentiating.Depending on the firm's strategy, PSM professionals have different targets, focuses or objectives (hereafter referred to as "objectives") (Schiele, 2019), such as lowering the costs of supplies (F.H. Hesping and Schiele, 2016), acquiring innovations from supplies (T.E. Johnsen, 2009) or improving sustainability (Schulze and Bals, 2020).
A distinction between PSM's objectives in reducing supply costs and creating long-time value by accessing the suppliers" innovations can also be found in other, primarily innovations-directed, PSM literature.Tracey and Neuhaus (2013) distinguish between innovation-focused ("facilitating") and operational ("regulatory") objectives of PSM professionals.This division is also seen by Calvi et al. (2018).They have highlighted that the PSM professionals who aim at innovation sourcing and implementation need to assess suppliers innovative capabilities rather than on low cost.According to Knight et al. (2014), both PSM types require deviating skills and share only a few common grounds.The recognition that purchasers might have very different objectives depending on their companies' strategies might suggest broad support in the literature on PSM skills for differentiated skillsets, at least one for cost and one for differentiation-oriented firms.Remarkably, this is not the case.Extant literature hardly distinguishes between the types of purchasers and their different objectives.
The outcome from the systematic review of PSM skills literature  performed for this study (see the theory section for further details) shows little unanimity.While a more significant majority of the studies agree upon the need for "PSM knowledge" (82 per cent), "business knowledge" (76 per cent) and "negotiation skills" (73 per cent), one-third of the studies introduce skills that the other studies did not consider, such as "creativity" (33 per cent), "industry knowledge" (33 per cent) and "legal knowledge" (30 per cent).
The possible deviations do not explain the unanimity within the set of PSM skills literature over time.The set of articles was divided into two categories: "old" (1987-2001; 16 articles) and "new" literature (2003-2020; 17 articles).Contentwise, both sets do not deviate significantly, as shown in the theory section.However, an explanation for the controversial findings could be that these various studies have presented the skillsets of different objectives and labelled them as something like a general skillset for a universal type of PSM professional (e.g.Bals et al., 2019;Giunipero et al., 2006;Giunipero and Pearcy, 2000;Tassabehji and Moorhouse, 2008).Nevertheless, as indicated by Knight et al. (2014), this general type might not exist.Depending on the empirical sample composition, different skills could emerge as being more or less important.Therefore, assuming that there is no "one fits all" type of purchaser, but differentiated types, the following research question arises: (RQ1) Which different skillsets do PSM professionals need following the different prevailing objectives of their organisation?
Further, are all different skills of equal importance?While the extant PSM literature is not unanimous in the PSM skillset design, the word choice detected in the PSM skills literature does attract attention and, more precisely, the use of the term "necessary" in combination with a particular skill.The PSM skills literature often assumes that "important" skills can be seen as "necessary" (conditions) for performing the PSM function (e.g.Bals et al., 2019;Giunipero and Pearcy, 2000;Tassabehji and Moorhouse, 2008; Von der Gracht, Giunipero and Schueller, 2016).
For instance, Bals et al. (2019) have used the term "necessary" in combination with "skills" six times, Giunipero and Pearcy (2000) seven times, Tassabehji and Moorhouse (2008) three times, and Schulze and Bals (2020) have used it twice.It is, however, questionable whether the presented skills are "sufficient" or "necessary" conditions for performing the PSM function.In the literature, the distinctions between "sufficient" or "necessary" conditions are often made in an implicit manner "using alternative formulations, such as X being critical or a pre-condition for Y" (Van der Valk, Sumo, Dul and Schroeder, 2016, p. 267).
Indeed, a necessary condition will prevent an outcome from occurring.There will be "no Y without X" in the necessary condition logic ( Van der Valk et al., 2016, p. 267).The sufficient condition logic is that the independent variable X leads to the outcome, i.e. the dependent variable Y. Van der Valk et al. (2016), who performed Necessary Condition Analysis (NCA) on "legal contracts" versus "trust" on innovation in buyer-supplier relations, concluded the following: "Thus, while a sufficient cause produces the outcome, a necessary cause allows the outcome to exist.Conversely, without the necessary cause, the outcome will not exist despite other factors being present" ( Van der Valk et al., 2016, p. 267).
The question of which skills are necessary and which are not is also essential for educators and trainers designing learning objectives, given the limitations of time available for education and hence the request to make choices which skills to teach first.The PSM skills literature does not differentiate between sufficient and necessary conditions.Consequently, it hinders the design of a prioritised set of learning objectives and further professionalising the PSM function.To support the development of learning and training objectives for PSM, this study, therefore, addresses the next research question: (RQ2) Which skills are necessary and which are sufficient for PSM professionals?
This study aimed to overcome these two gaps in the PSM skills literaturethe lack of differentiation in purchasers" objectives and the lack of differentiation between important and secondary skillsand followed a quantitative approach in surveying European PSM professionals (n = 366).The participants rated a set of 88 skills that were identified in the literature.Eventually, it became possible to condense a statistically relevant set of 15 factors, each representing a particular PSM skill category.As a next step, OLS (ordinary least square) regression was used to identify those skills that explained the results in a particular PSM objective and NCA (necessary condition analysis) to distil the primary or necessary conditions for job execution.
This investigation contributes in at least three ways: As the first contribution, this study adds to the literature by presenting an empirically validated PSM skills model and taxonomy, which is the first to be based on a sizeable multi-national sample, identifying 15 empirically backed skill factors.Instead of reporting on the availability of skills, only, this study uses the purchaser's success as the dependent variable.In this way, the importance of soft factors is unveiled, likewise the seldom reported relevance of "sellership", the need for purchasers to have the skill to promote their company and function.The empirical analysis also reveals a different ranking in the importance of purchasing skills than, on average, suggested by previous research.
Second, this research discusses and tests seven PSM objectives (focus on the cost objective, quality, delivery, sustainability, supplier satisfaction, innovation or competitive advantage).For each objective, it is shown that a different, adjusted skillset is necessary.This finding questions the validity of the traditional "one size fits all" approach, clearly indicating that there does not seem to be someone like "the" purchaser.Depending on corporate objectives, different skills are needed.By including different objectives organisations have, this research contributes by introducing a new analysis level, previously not considered in purchasing skill research.The empirical findings indicate internal inconsistencies in the standard one-size-fits-all approach, as some skills have an opposite effect on performance with different targets.
Third, being the first study to distinguish between necessary and sufficient conditions, this research can reposition the soft-skill vs.hardskill debate.Classical regression analysis identifies the differentiating factors (here: skills), explaining an outcome (here: purchaser's performance).However, the necessary condition analysis shows that the four identified soft skills, while having minor differentiating character, are essential antecedents to hard skills.Without a minimum level of soft skillsnetworking, result orientation, imagination and sellershippurchasers are not efficient in employing a series of the identified hard skills.
Finally, the findings challenge some of the common grounds in the literature review, such as alerting us about an over-emphasis on negotiation skills and revealing the pivotal importance of supplier relationship management and strategic sourcing management skills.Our research also contributes to PSM practice by providing a PSM skills model and taxonomy and by deriving a list of skills differentiating between the PSM objectives.It provides insight into those skills that can be associated with results in PSM.It contributes in practical terms to the practitioners" work regarding PSM curricula in higher education and company training programmes.
This paper is organised in the following manner.After introducing this study, the second section presents the literature on PSM skills and discusses PSM objectives.The subsequent methodology section describes how the literature review was performed, how an extensive European survey was set up, how the survey items were subjected to factor analyses and how the data, i.e. the factors, were analysed with OLS regression and with the newly developed NCA method (Dul, 2016).The sets of results of this study are presented in the fourth section.The final section includes the conclusions, a discussion on further research, the implications of the findings, and the study's limitations.

Systematic literature review indicating a lack of consensus over time
For this study, a keyword search in Scopus was applied, following the guidelines of Durach et al. (2017).It led to an initial set of 1007 articles.A list of PSM skills was gradually extracted (displayed partly in Appendix 1).The key search terms that were used were "buyer", "purchase", "purchaser", "purchasing", "procurement", "skill", "skills", "competence", "competency" and "competencies" and led to the search string "buyer OR purchas* OR procurement AND competenc* OR skill*".Articles regarding non-individual, organisational capabilities and those focused on consumer skills of family household purchasing budget use were discarded.Likewise, the minor citations that mention the search terms as "PSM skills" only incidentally were discarded.Eventually, only the skillsets referred to in 33 scientific studies published between 1987 and 2020 PSM skills could be extracted (see: Appendix 2).
Appendix 2 displays the PSM skills" frequencies of being mentioned at least once in the focal scientific PSM article and shows little unanimity.While a large majority of the studies agree upon the need for "PSM knowledge" (82 per cent), as well as "business knowledge" (76 per cent) and "negotiation skills" (73 per cent), about one-third of the studies introduce skills that the other studies did not consider, such as "creativity" (33 per cent), "industry knowledge" (33 per cent) or "legal knowledge" (30 per cent).Remarkably, whereas the PSM literature has distinguished different objectives within the function, the PSM skills literature has not made this differentiation.Instead, it presents general skillsets for a universal PSM professional type, with Knight et al. (2014) and Schulze and Bals (2020) as exceptions.
Appendix 1 displays the methods that were used in the researched articles.In approximately 20 per cent of the cases, these were qualitative (case studies or interviews); in 40 per cent of the cases, these were quantitative, survey-based, and in another 40 per cent of the cases, these were based on a mix of quantitative (surveys) and qualitative methods (case studies or interviews).Following Fawcett and Rutner (2014, p. 180), who stated that the PSM practices in firms have "evolved dramatically over the past generation", the data were subjected to further statistical analyses.
Therefore, the scientific PSM literature sample was divided into two categories: from 1987 to 2001 (16 articles) and 2003 to 2020 (17 articles).Next, an independent t-test was performed on the skills frequencies.The outcomes revealed insignificant p-values (p > .05).Therefore, the PSM skillsets, as proposed in the series of "old" (1987)(1988)(1989)(1990)(1991)(1992)(1993)(1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001) and "new" (2003-2020) literature, do not deviate significantly, and the non-unanimity within the PSM skills literature  is not explained with possible deviations over time.However, within the complete set of 33 articles, non-unanimity exists, which could be explained by the supposedly exploratory character of the research field or by the different objectives that could have been mixed up.The PSM skills literature is briefly described in the next sub-section.

PSM skills in literature: theoretical perspectives, distinguishing between professional skills and knowledge and personal skills and traits
Based upon the summaries of Tassabehji and Moorhouse (2008) and Bals et al. (2019), an extended summary of PSM skills in the literature is displayed in Table 1.Initially, the main focus has been on professional knowledge and skills: computer literacy skills, cost reduction and financial skills, quality management, analytical skills and strategic sourcing.Gradually, the attention for personal skills and traits emerged, especially for decision-making, communication, influencing and persuasion, teamwork, creativity and conflict resolution.
Traditionally, procurement was asked to provide the right product, at the right moment, with sufficient quality for the lowest price.However, in the last decades, PSM evolved from such a mainly clerical  function to more a strategic one, adding further objectives purchasers have to fulfil.From the 1990s, firms focused on make-or-buy decisions and started to define their core and non-core activities (Luzzini and Ronchi, 2016).The focus on core competences in organisations, as underlined by Prahalad and Hamel (1990) and the outsourcing of the rest had as a consequence, that the importance of supplier management, supply chain management and strategic decision-making increased, and so the scope of purchasing objectives (Luzzini and Ronchi, 2016).However, the attention for differentiated purchasing objectives is not reflected in the proposed skillsets as displayed in Table 1.These skills sets have a universal character and do not differentiate between purchasing practitioners' objectives.The study by Knight et al. (2014, p. 278) is an exception.This work was the first to detect different PSM objectives in their sample.Based on a survey in Taiwan (n = 72), strategic, tactical and routine purchasing skills profiles were found and were projected on the quadrants of Kraljic (1983).The most critical strategic competencies are a mix of personal and professional skills.They are: "influencing and persuasion", "decision-making skills", "coordinating skills", "ability to work in teams", "managing internal customers", "change management", "strategic thinking", "specification development", "understanding manufacturing terminology and processes".
Another exemption is the interview-based research (n = 46) of Schulze et al. (2019) that elaborated on skills specifically for corporate social responsibility (CSR, sometime also called ESG, as environmental, social and governance factors) and sustainable PSM and concluded that strategic skills are more effective than transactional, operational skills.This study is the first in this series to be dedicated to a particular PSM objectivein this case, sustainabilitywhile the bulk of the traditional literature on PSM skills is qualitative or conceptual and does not differentiate among PSM's diverse objectives might have to fulfil.
In line with the growing empirical and conceptual importance of the PSM function, the number of PSM objectives has increased, which has made the PSM function more complex.A solution to grasp the complexity comes from Tchokogué et al. (2017), who proposed a sequence of organisational objectives based on a systematic PSM literature review .This sequence expands from operational to strategic objectives.Moreover, the sequence in objectives seems to be hierarchical, i.e. the supply function must first master the operational corporate improvement objectives before it can support strategic competitive advantage objectives."Therefore, supply professionals should develop their ability to utilise the currently acknowledged strategic contributions better, while being ready to take advantage of the new types of contributions" (Tchokogué et al., 2017, p. 105).Thus, a distinction can be made between operational or tactical activities that involve an emphasis on efficient order placement and price savings and strategic activities "with an emphasis on total cost savings and value-added activities" (Giunipero et al., 2006, p. 824).
A sequence proposed by Tchokogué et al. (2017) at an organisational level has also been seen by Ahmed et al. (2012) to be applicable at an individual level.They have found evidence that soft skills (i.e.interpersonal skills and intrapersonal traits) are as important as professional skills (or hard skills).As is evident, the PSM skills literature also recognises personal and professional skills.In line with this skills literature, the educational literature distinguishes between "soft skills" and "hard skills" (e.g.Andrews and Higson, 2008;Bailly and Léné, 2013;Heckman and Kautz, 2012;Laker and Powell, 2011).According to Heckman and Kautz (2012, p. 451), soft skills are "personality traits, goals, motivations, and preferences".Laker and Powell (2011, p. 113 more expected to be the reason for ending a labour relationship than a lack of hard skills, and they add that soft skills not only are highly valued by employers but are also necessary to carry out professional tasks or hard skills (Ahmed et al., 2012).In line with that, Muir (2004, p. 99) has stated that "soft skills are necessary to actually implement-to articulate a vision; to enrol others in possibilities; and to communicate values, standards, and expectations".Socha, Razmov, and Davis (2003, p. 4) have added that "soft skills are necessary to apply the hard skills effectively".Hence, soft and hard skills need to be analysed and developed.

PSM objectives: operational and strategic requirements for PSM reflecting a broad scope of purchasing objectives
In terms of systematically describing the PSM objectives, Nollet et al. (2005) have noted that Doyle (1990) was the first to provide a PSM strategy reflecting five purchasing objectives: "quality improvement (variance reduction), velocity (concept-to-customer cycle time improvement), all-in-cost (total-cost purchasing practices), technology (access and active monitoring) and risk reduction (an activity-managed program approach)" (Nollet et al., 2005, p. 134).Moreover, PSM's task is to provide the most value throughout the value chain to contribute to sustained competitive advantage (Nollet et al., 2005).Schulze et al. (2019) have added to the list of these six objectives, of Nollet et al. (2005), a seventh objective owing to the "increasing resource scarcity challenges, stakeholder expectations and growing legal regulations regarding environmental and social concerns, PSM not only has to manage traditional objectives such as cost, time, quality and flexibility but is also faced with managing innovation and sustainability within supply chains" (Schulze et al., 2019, p. 3).Reflecting the understanding of purchasing as a strategic function, contributing to generating a competitive advantage and caring about the relationship towards suppliers and their satisfaction becomes a further objective purchasers have to comply to (Schiele, 2019).
As mentioned, the PSM function is mainly dependent on PSM professionals' skills that have to fulfil differentiated objectives that come forth from the firm targeting an industry or segment with either differentiation or a cost leader strategy (Porter, 1980).The approaches at the one hand in cost reductions and at the other hand in innovation sourcing require fundamentally different approaches, as is referred to by Legenvre and Gualandris (2018, p. 97), who found evidence that PSM professionals responsible for innovation sourcing "need to look beyond risks, cost, and product development to offer further competitive advantages".Hence, the objectives in PSM can differ, and the full PSM skill palette is broad, serving sheer operational objectives, as well as extremely strategic ones, leading to potentially contradictory skillsets.

Research design of the online European PSM skills survey
This study depended on the responses of 366 PSM professionals from the private sector in the PSM skills survey.Initially, about 3200 European PSM professionals from the public (n = 215) and private sector (n = 366) were personally invited via LinkedIn.comto take part in the online survey (n = 581; response rate > 18 per cent).The survey was pre-tested with six PSM professionals in different industries in both Germany and the Netherlands.The pre-test led to the clarification of the texts of the items and the introduction of missing items.
For this research, the respondents from public procurement were discarded as there was evidence that the procurement in the public sector differs significantly from the private sector (Stentoft Arlbjørn and Vagn Freytag, 2012;Telgen et al., 2007).Despite the similarities in the demands for supplies in public procurement and private purchasing, both deviate since public procurement is subject to external demands such as "transparency", "integrity", "accountability", and "exemplary behaviour" and to internal demands like serving multiple political goals for many stakeholders.Moreover, public procurement is budget-driven and has to comply with regulations (Telgen et al., 2007).
There was a standard spread across different age groups in the sample, and about one-quarter of the respondents were female, and three-quarters were male.The respondents were primarily Western European, predominantly Dutch.About 58 per cent of the respondents were of Dutch nationality, 14 per cent were French, and about 11 per cent were German.All items were measured on a five-point Likert scale (ranging from "fully disagree" to "fully agree").The survey consisted of different kinds of items.The participants self-rated 88 different competency levels and 22 self-rated statements on individual and organisational effectiveness in the seven objective areas aimed at improving (1) cost levels, (2) delivery performance, (3) quality and (4) sustainability levels, (5) achieving sustained competitive advantage, (6) supplier satisfaction and (7) innovation sourcing and implementation (see Appendix 3 for the formulation of the items and the sources).
The form and wording of the "cost reduction" and "innovation" statements were based on the studies by Azadegan and Dooley (2010), F. H. Hesping and Schiele (2016), Krause et al. (2001), Schiele et al. (2011) and Terpend et al. (2011).The quality, sustainability, delivery, competitive advantage and supplier satisfaction items were based on the literature (see the references in the listing in the theory section) and stated or rephrased in the exact wordings as the "cost reduction" and "innovation" statements.
The participants were invited to rate their competence level on 88 skills (see Appendix 4).After that, the participants were asked to rate statements on their effectiveness on the seven PSM objectives.All items were measured on a five-point Likert scale (ranging from "fully disagree" to "fully agree").Self-assessment on a Likert scale is a widely accepted method in the PSM skills literature (e.g.Giunipero, 2000;Giunipero and Pearcy, 2000;Knight et al., 2014).
The 88 skills items and the 22 statements on individual and organisational targets in the seven objectives were subjected to factor analyses.In both cases, Bartlett's tests of sphericity appeared to be significant, and the Kaiser-Mayer-Olkin measure of sampling adequacy (KMO) was above the threshold (KMO of skills: 0.963; KMO of statements: 0.818), and therefore, the factor analysis could be performed (Bartlett, 1950;Kaiser, 1970Kaiser, , 1974)).
Confirmatory factor analysis by applying principal component and Direct Oblimin rotation in SPSS was performed for the 22 statement items.The 22 items were found in the expected seven categories.Hence, the factor analysis confirmed that the three "cost" items belong in the "cost" factor, etcetera.The factor loadings were all above 0.625, and the Cronbach's Alpha for the seven constructs was assessed in all cases above 0.750, which is considered very satisfactory (Cronbach, 1951).Next, for each of the seven PSM objectives, the "effectiveness statements" were computed to a latent PSM effectiveness variable.
Multiple iterations of exploratory factor analyses were performed in SPSS by applying principal component and Equamax rotation.Equamax rotation was applied for this extensive set of 88 skills items of 366 respondents because it "combines Quartimax and Varimax criteria by simplifying both the variables and factors in the factor pattern matrix and spreading variances more equally across the factors" (Sass and Schmitt, 2010, p. 80).The eigenvalues of the selected factors were set above 1, which resulted in 16 factors.
The exploratory factor analyses were performed in an iterative process and limited to 12 to 16 factors.The outcomes were analysed and compared.The factor analysis was eventually confined to 15 factors that were selected as the most consistent ones.Increasing the number of factors resulted in "PSM knowledge" emerging as an additional single factor on its own, and the reason for this is difficult to interpret.Reducing the number would have merged soft skills into an undifferentiated, amorphous mega-factor.After that, the 15 factors of skill variables were transformed and computed into a new variable to construct the underlying items to perform OLS regression and NCA.
Moreover, the reliability of the items in the different factors was assessed.In total, 13 factors showed an Alpha of above 0.800, which is considered very satisfactory (Cronbach, 1951).A lower but satisfactory Alpha was calculated for two factors, i.e. 6 Forecasting skills (α = 0.729) and 15 Technical skills (α = 0.683).OLS regression was performed for the 15 PSM skills factors on each of the seven PSM objectives.The variance of inflation factors (VIF) did not exceed 4.2 and with tolerances higher than 0.2.Hence, the VIFs are under the maximum level for multicollinearity issues (Hair et al., 2010;Ringle et al., 2015).

NCAa new method for measuring necessary conditions for performance
The PSM skills literature often assumed that "important" skills could be seen as "necessary" (conditions) for the performance of the PSM function.Examples have been given of PSM skills studies in which the term "necessary" was used in combination with the term "skill", although it is questionable whether the presented skills were "sufficient" or "necessary" conditions for performing the PSM function.This finding is in line with Van der Valk et al. (2016, p. 267), who found that distinctions between "sufficient" or "necessary" conditions in the literature are often made in an implicit manner "using alternative formulations, such as X being critical or a pre-condition for Y".To systematically and statistically solve issues of this kind, the NCA method has been presented by Dul (Dul, 2016(Dul, , 2018b)).
The idea behind employing the NCA method is that, in general, business research is focused on factors that lead to specific outcomes.However, (underlying) factors necessary enablers for outcomes to occur are crucial in organisational decision-making ( Van der Valk et al., 2016).Such an enabling factor is a necessary condition "that must be present to enable a certain outcome; without the condition, the outcome will be absent" ( Van der Valk et al., 2016, p. 368).A necessary condition can be seen as a bottleneck that prevents the desired outcome's occurrence (Dul, 2016).According to Dul (2016), sufficient and necessary conditions can be seen as two different logical parts of the notion of causality.In the case of necessary conditions, the performance will not be achieved in their absence ("no Y without X") (Dul, 2016;Van der Valk et al., 2016).In conclusion, a sufficient condition produces the outcome, and a necessary condition allows the outcome to exist.Indeed, the absence of the necessary condition will lead to the absence of the outcome, regardless of other factors being present (Van der Valk et al., 2016).
According to Van der Valk et al. (2016, p. 267), the difference between necessary and sufficient condition logic is that in the necessary condition logic, there will be "no Y without X", and in the sufficient condition logic, "X leads to Y". Regression-based methods refer to suf- each input (…) is sufficient to increase the outcome, but not necessary: A lack of an input reduces the outcome, but it will not prevent the outcome if other inputs (…) compensate for it" (Hauff et al., 2019, p. 2).However, NCA is described in terms of Y = X 1 *X 2 *X 3 … * X n , in which the effect of a value appearing to be zero is more influential than in the regression-based methods."The dramatic sudden effect of zero values for necessary conditions fits many everyday experiences.A car stops moving if the fuel tank is empty; financial markets collapse if the trust is gone" (Dul, 2016, p. 11).Van der Valk et al. (2016) have concluded as follows: "Thus, while a sufficient cause produces the outcome, a necessary cause allows the outcome to exist.Conversely, without the necessary cause, the outcome will not exist despite other factors being present" ( Van der Valk et al., 2016, p. 267).
Although the NCA method is relatively recently developed, impressive research outcomes could be presented, of which three illustrative examples are given here.Van der Valk et al. (2016, p. 267) have performed NCA on "legal contracts" versus "trust" on different innovation levels in buyer-supplier relations, and their research shows "that successful relationships (…) must necessarily have contracts with at least medium levels of contractual detail, as well as the highest levels of trust".Leischnig et al. (2018, p. 385) have studied "the relationships between service employees" adaptive behaviours and customer satisfaction" and found "that interpersonal adaptive behaviour is a necessary condition (…) for high customer satisfaction".Fredrich et al. (2018, p. 862) have found that "absorptive learning capacity is a necessary condition of interfirm learning, while organisational slack resources and interdependence between firms are important additional conditions".
To analyse the data, the manual of Dul (2018b) was followed using the statistical functions of the open-source programming language R, complemented with the NCA R package provided by Dul (2018a) and using the R Studio application.The seven dependent variables were combined with the 15 independent variable constructs based on the abovementioned factor analysis.NCA was applied on the entire set for every single factor to find structures and internal logic in the set of 15 independent variable constructs.
The NCA method was performed concerning each of the 15 factors on the other 14 factors to test if soft skills would be necessary to carry out other skills.In the original logic, the 15 factors were projected as the independent X variables being necessary (or not) for the dependent Y variables, i.e. the seven different PSM objectives.However, calculations were made for each of the 15 factors.Each factor fulfilled the role as the dependent Y variable, intending to detect a possible dependency or skills hierarchy, which has been a unicum in (PSM) skills research as far as is known (see Appendix 5).
By using the statistical NCA R package, the effect sizes were calculated and presented."If the effect size is greater than zero, there is (…) an indication of the presence of a necessary condition.[…] It ranges from 0 to 1 (0 ≤ d ≤ 1).The effect size indicates to what extent the condition is necessary for the outcome.In other words: to what extent the condition constrains the outcome, and the outcome is constrained by the condition" (Dul, 2018b, p. 10).
Regarding the interpretation of the effect size, Dul (2016, p. 30) has stated that "an effect size can be valued as important or not, depending on the context.A given effect size can be small in one context and large in another".Dul has further stated that general qualifications for the effect size such as "small," "medium," or "large" are disputable.However, a general benchmark for necessary condition effect size is suggested as follows: "0 < d < 0.1 as a "small effect," 0.1 ≤ d < 0.3 as a "medium effect," 0.3 ≤ d < 0.5 as a "large effect," and d ≥ 0.5 as a "very large effect" (Dul, 2016, p. 30, p. 30).

PSM skills summarised in 15 factors
Factor analyses were performed at first to answer the research questions.Eventually, after an iterative process, 15 factors were distinguished, which are displayed in Table 2.In total, 85 of the 88 initial competence items could be found in one of the 15 factors.Appendix 4 shows the factors, the items and the factor loadings, which are PSM skills taxonomy.To recognise the factors, these have been numbered and labelled with a short title that briefly describes the factor content, a standard modus operandi in PSM skills literature (e.g.Giunipero and Pearcy, 2000;Tassabehji and Moorhouse, 2008).

Multiple sets of skills combinations supporting seven PSM objectives
The next step in the analysis was to address both the research questions that are as follows: Which different skillsets do PSM professionals need according to the different prevailing objectives of their PSM organisation?Which skills are necessary and which are sufficient for PSM professionals?Therefore, the 15 competence factors (vertical of Table 3) were combined with analyses combined with the seven PSM objectives (horizontal in Table 3).The analyses consisted of OLS regression and NCA, leading to a typology of PSM objectives as predictors for effectiveness related to different PSM objectives as an answer to the research questions.
Table 3 displays the sufficient (OLS) and necessary (NCA) conditions for the subsequent PSM objectives.The insignificant values are suppressed and displayed in grey font.Moreover, for the NCA, the significant but "small effects" with an effect size d smaller than 0.100 are also neglected and displayed in a grey font (Dul, 2018b, p. 10).Hence, Table 3 highlights the significant "medium/high effects" for the NCA and the significant p-values for the positive and negative Betas.Table 3 reads as follows: procurement performing in delivery objectives, cost focus is a necessary skill, while those purchasers with forecasting skills outperformed their colleagues, showing lower competence in this skill.Statistically, it was found that typically delivery specialists had fewer contracting skills, which are more important to fulfil other objectives.In education, delivery specialists would benefit from following cost reduction and forecast improvement training but sending them to contract law classes would not be efficient.

Theory implication: distinguishing 15 sufficient and necessary PSM skill factors for seven PSM objectives
The theoretical contribution is the novel approach in this study.Unlike preceding research on PSM competence, this study does not list the highest-ranked items or factors and is not focused on top-rankings or high scores directly provided by the respondents.This research is not centred on "important" high ranked competences as seen by a group or an individual.Preceding studies asked individual purchasers to give their view on which skills they considered important.However, they did not at the same time ask which targets these purchasers had to fulfil.In the end, mixing up the responses of someone working in operative procurement with a clear delivery focus and someone responsible for strategic sourcing in a hight-tech innovation commodity group, may level out the differences.This study, in contrast, distinguished different PSM objectives and asked the participants to state their effectiveness per PSM objective and associated the independent competence factors with these dependent PSM objective statement factors with OLS regression and NCA.This research aimed to associate necessary and sufficient skills items with different PSM objectives.Moreover, evidence for a skills hierarchy is shown.Results clearly demonstrate that future research would have to take up the different targets and different roles purchasers have and deepen the understanding of each of them, but not as a fictitious amalgamate of a single type of purchaser, which is not empirically present.
This research analysed which skillsets would support achieving particular objectives of purchasing in a firm (RQ1).Further, it distinguished between necessary skills to achieve these objectives and sufficient for that target (RQ2).Three main contributions strike out: first, an empirically backed 15-factor purchasing skills model emerges.Second, it was shown that depending on the purchasing objective different skills need to be emphasised.It can serve as a blueprint for curriculum and training design and will be discussed as "education implications".Third, this paper introduces the distinction between necessary and sufficient skills, which re-evaluates the role of soft skills as antecedents to hard skills.
For the seven PSM objectives, a distinction is made between the sufficient skills (X s ) that "produce" the high effectiveness in a specific PSM objective (Y) and necessary skills (X n ) that "allow" the effectiveness to existing in a particular PSM objective (Y) (see: Table 3).In the sufficient condition logic, "X s leads to Y", and in the necessary condition logic, there will be "no Y without X n " ( Van der Valk et al., 2016, p. 267).Applying this logic, a series of insights emerge.
Interestingly, the Cost reduction objective strongly shows a necessity for Organisational insight and governance (consisting of: Know how to add value to the organisation; Corporate Governance; Position of Purchasing in Organisation; Process and Project Management Skills; Optimisation of Purchasing Processes; and Team Ability Skills), only.The link between cost reduction and the need to have organisational insight is a remarkable outcome.In the literature, often a distinction is made between reducing supply cost and creating long-time value by accessing the suppliers" innovations.At the same time, cost reduction seems to be a process management issue.
A balance between innovation sourcing and cost reductions is also found in the objectives regarding Innovation sourcing, Quality Improvement and Sustainability.All show negative signs for Factor 7, i.e.Cost focus (e.g.Solicit offers; Making cost analyses; Cost Reduction Techniques).Here, the negative sign for the factor Cost focus could indicate that the innovation, quality and sustainability objectives have a nontransactional, non-operational, strategic character, and the remark of Legenvre and Gualandris (2018) to look beyond costs for competitive advantage is applicable.In contrast, a transactional, operational necessity for Cost focus is found in the Delivery improvement objective.This operational objective benefits from performing negotiations, cost analyses and cost reduction techniques.
Factors 9 and 10, i.e.Supplier relationship management and Innovation sourcing, respectively, show similarities.Both are associated with the strategic PSM objectives regarding improving sustainability, supplier satisfaction, competitive advantage innovation sourcing and quality improvements.The soft skills found in Factor 3 Imagination and Factor 4 Sellership skills are necessary conditions for innovation sourcing and achieving supplier satisfaction.
Altogether, based on an extensive survey conducted amongst 366 PSM professionals, the conclusion is that different skillsets do, indeed, exist in PSM.Further, contradictions between the objectives are primarily found in some objectives showing negative signs for the Factor 7 Cost focus.In contrast, Factor 13 Organisational insight divide the PSM objectives into those with an operational and strategic focus.This division in reducing supply cost and creating long-time value by accessing the suppliers' innovations is found in the literature, as mentioned in the introduction.Tracey and Neuhaus (2013) have distinguished between innovation-focused ("facilitators") and operational ("regulators") objectives of PSM professionals.This division is also noticed by Thomas E Johnsen et al. (2012), Legenvre and Gualandris (2018) and Knight et al. (2014).Here, a dividing line between PSM strategies focused on either differentiation or a cost leader strategy, as mentioned by Porter (1980), is seen.
Table 4 presents the necessity of skill factors for the seven PSM objectives.Interestingly, the Supplier satisfaction objective counts eight necessary factors, and the Delivery improvement objective counts only one factor.The difference between both depicts the distance between strategic PSM and operational PSM (e.g.Knight et al., 2014;Legenvre and Gualandris, 2018;Tracey and Neuhaus, 2013).Supplier satisfaction is one of the most challenging objectives to achieve, as it requires a large set of skills.Most other objectives rely on a narrower set of necessary skills (see Table 5).
Finally, another remarkable outcome is that the soft skills (factors 1 to 4) are mainly absent in OLS regression but prominently present in the NCA outcomes.NCA shows significant effects (p < .05),while with OLS, only insignificant Betas could be detected (except for imagination's influence on fulfilling the innovation target).Without NCA and just running a classical regression, it could have been concluded that soft skills are not essential.However, the above outcomes provide input to conduct a deepening investigation on the assumption that soft skills are necessary conditions to perform hard skills to analyse whether some of the 15 skills factors have a necessary influence on the other skill factors.

Deepening investigation: quantitative evidence for a skills hierarchy; soft skills are needed for hard skills
Based on the observation that hardly any soft skills show up in the regressions, we decided to conduct additional analysis to compare their influence on the other skill factors.Therefore, each of the 15 factors was the subject of NCA on the other 14 factors.In other words: in the subsequent analysis, we do not compare, like in the primary research, the influence of the 15 skill factors on the potential to fulfil each of the seven objectives, but we analyse the relationship among the 15 factors themselves.Could it be that some of the factors are a condition for others?
NCA provides an output showing the significance by providing pvalues and showing the effect sizes d.In Appendix 5, a table of effect sizes is presented, which formed the basis for the surface graph in Fig. 1, displaying the effect sizes d of the factors as necessary conditions for the other factors.In Fig. 1, "small effects" (0 < d < 0.1) are displayed in white colour, "medium effects" (0.1 ≤ d < 0.3) in grey colour and "large effects" (0.3 ≤ d < 0.5) in black colour (Dul, 2018b, p. 10).For instance, Factor 1 Networking shows "black peaks," i.e., it is necessary for all other factors; this factor shows high significant effect sizes.In contrast, Factor 14 Automation is mostly in the white "lower area" and shows minimal effect sizes on all other factors and is not necessary to carry out the skills in the other factors.This is also quite logical, for purchasing was also working before automation tools were introduced.Hence, the managing of them improves outcome but is not strictly required to buy something.
The factors in Fig. 1 are ranked based on their means, meaning that, for example, Factor 1 Networking has the most significant effect sizes on the other factors (see Appendix 5).The mean of the 14 effect sizes of this factor on the other factors is 0.415, which is on average a "large effect".Therefore, Factor 1 Networking is a necessary condition for all other skills factors.In contrast, Factor 14 Automation is not a necessary condition for the others as it shows a mean of 0.069, which is a "small effect".The remarkable outcome is that the "soft" factors networking, resultorientation, imagination and sellership score highest.It means that soft skills are essential to perform more hard skills coined skill factors, while the latter then make the difference between good and outperforming professionals.Note: See Appendix 4 for the complete overview with the descriptions of the skills, the references, the factor loadings and the Alphas.

Table 3
Regression and NCA outcomes for the seven objectives and the 15 factors.
Note: "Small effects" and insignificant outcomes of the NCA for the seven PSM achievements are suppressed and displayed in grey font (0 < d < 0.1 "small effect"; 0.1 ≤ d < 0.3 "medium effect"; 0.3 ≤ d < 0.5 "large effect"; d ≥ 0.5 "very large effect" (see paragraph 3.2 -Dul, 2018b, p. 10)).K. Stek and H. Schiele This finding goes in line with the idea introduced by Tchokogué et al. ( 2017) and Ahmed et al. (2012), Muir (2004) as well as Socha et al. (2003).Tchokogué et al. (2017) that organisational objectives are sequenced and hierarchical.According to them, PSM organisations must first master the operational corporate improvement objectives before strategic competitive advantage objectives can be supported.For individuals, Ahmed et al. (2012), Muir (2004) and Socha et al. (2003) have indicated that soft skills are necessary conditions to execute hard skills, which is exactly what our data support.
These outcomes of this study differ from previous research.The skills found in the literature review (Appendix 1) were categorised according to the new taxonomy in Table 2.In the literature, the most extensive attention was directed to items in Factor 7 Cost focus.On average, 77.2 per cent of the researched literature agrees on Factor 7, followed by Factor 14 Automation (60.1 per cent).The soft skills factors 1 Networking (24.5 per cent), 2 Result orientation (23.5 per cent), 3 Imagination (43.4 per cent) and 4 Sellership skills (27.2 per cent) are outside the top-5 according to the literature.This is in line with this study's "sufficient" OLS findings where the four factors are absent (with one exception).Nevertheless, with NCA, the importance and the prominent position of these four are made clear, which is another contribution of this research.

Implications for education: 15 skills as a blueprint for PSM training
The outcomes of this study have important implications for the design of PSM curricula in higher education, company training sessions and HRM practice in the field of PSM.The first implication is that sufficient and necessary skills for the seven PSM objectives are revealed.Educators, HRM officials, trainers and individual PSM professionals can benefit from these outcomes by adapting these and adjusting the learning objectives of courses, training and development plans.Consequently, firms establishing a purchasing training programme would benefit from, first, clearly establishing the target priorities their purchasing function would follow.This activity, which we did not find very strongly emphasised with the firms visited in the preparatory benchmark, has to be emphasised.Establishing the targets of the firm and the respective purchaser is the basis for a tailor-made skills assessment and development.Else, skills may be trained which do not contribute to success, while others are left out.On basis of a clear understanding of the targets, a tailor-made training programme could be offered.This programme is different from firm to firm.Likewise, universities or universities of applied sciences that teach purchasing would also have to define the direction of their targeted education and decide for respective course portfolios.It does not imply completely ignoring any of the 15 skills factors.It would be advisable to provide a basic skill level in all factors to enable trained persons to switch to organisations with different targets.However, the findings on the difference of effectiveness imply that choices may have to be made, and more emphasis may want to be placed depending on the chosen objective in education.
In the following, a short proposal based on the empirical findings of

Objective Skillset/Training
Delivery A particular "generic skill portfolio" and hence training set emerges for the more operationally oriented purchaser, who has to ensure prompt delivery at the right quality and the right sustainability level.The delivery expert adds solid analytical skills, forecasting, cost focus and also automation skills.Surprisingly, contract management shows a negative sign, indicating that, at least for the moment, legally well-trained delivery managers are not more successful.

Quality
The only target where technical skills make a significant difference is the quality target.In itself, this is an exciting finding supporting the challenge to the often expressed claim that purchasers should have strong technical skills (Cho et al., 2019).If purchasing, respectively the firm, is emphasising quality insurance from the supplier or the quality control is hosted in purchasing and not as a department of its own, technical know-how is required.Cross-functional skills are necessary, as well as SRM and analytical skills.Not surprisingly, a negative relation to cost-cutting skills is evidenced, whereas contracting skills are a weak differentiator.

Sustainability
Contracting, including the collaboration with legal departement is necessary for targeted sustainability, reflecting the regulation-driven character of sustainability considerations (Schulze and Bals, 2020;Schulze et al., 2019).Also, in the factor analysis, the item "corporate social responsibility" grouped itself to the contracting cluster, emphasising the solid legal connotation of sustainability.Likewise, sellership skills are necessary (as the sustainability targets are not always complementary to corporates' business requirements).Interestingly, cost skills show a robust and highly negative sign in the regression.Very cost reduction, global sourcing, and negotiation-oriented purchasers, on average, do not achieve sustainability targets better.

Cost
For a purchaser to be successful at cost reduction, an "overlap" target relevant for more strategic or more operationally oriented profiles requires a series of necessary skills.It is not necessary to be soft skills in networking and imagination, neither automation nor technical skills.Interestingly, successfully cost-cutting purchasers did or did not have particular contracting skills.A particular training in cost reduction methods such as, for instance, lever analysis requirement (Schiele et al., 2011) is a medium-strong differentiator.

Competitive advantage
Process and sellership skills are complemented by SRM and, as differentiator, advanced sourcing skills.Market analysis, category strategy planning and the diverse innovation-oriented subskills are grouped.Interestingly, in terms of soft skills being "result driven", i.e. showing self-assurance, poise, proactivity, are characteristics to such persons.The training programme for innovation purchasers and strategic purchasers will show considerable overlap.Likewise, the soft skills profile and hence training is similar, though with a slightly different focus on the one hand on imagination and on the other hand on what Hofstede calls "masculinity" (Hofstede, 1983).

Supplier satisfaction
A fascinating skills profile emerges for the target "supplier satisfaction".It is part of the strategic block, as the precondition for both innovation and strategic competitive advantage by privileged supplier access, is supplier satisfaction, which may lead to preferred customer status, the ultimate strategic target purchasing can achieve (Niels J Pulles et al., 2019).With this target, however, 12 out of 15 skills have the character of necessary conditions.A purchaser who achieves satisfied suppliers has a good level in almost all skills.No single target is as demanding.Effectively this means that supplier satisfaction seems to be more a question of the purchaser's maturity level (Schiele, 2007).Or, explainethe other way around: a poorly trained purchaser is likely to generate unsatisfied suppliers.

Innovation
The innovation purchaser shows a very similar profile as the purchaser can achieve a competitive advantage for his company.Hence, the innovation purchaser may also be the CPO of the future, as he can contribute to the entire firm's strategic positioning, thus fulfilling the new strategic role of purchasing.Purchasers successful in the objective "innovation" from and with suppliers typically rely, as necessary conditions, on a good Objective Skillset/Training level of salesmanship, process & project skills and the soft skill "imagination", which is also a differentiator.The latter is not surprising, as this skill comprises elements like creativity, inventiveness and holistic thinking.Business innovation is a very project and process-driven activity; firms typically having implemented any variant of a stage-gate model in innovation management (Cooper, 2019).Hence, purchasers need to acquire process & project skills to succeed in innovation.
Another necessary skill is salesmanship, for any novelty needs to be "sold" in the organisation.The more SRM and advanced sourcing skills a purchaser has, the better he performs.Part of the advanced skills is innovation sourcing and innovation implementation planning.Interestingly, a negative relation between cost and innovation skills has been found.
our study can be provided.Depending on the requested target, different training plans can be conceived: Another important implication is that soft skills form the basis for effective performance, being necessary conditions to many other skills (Fig. 1).Soft skills are appreciated by employers (Ahmed et al., 2012).Nevertheless, explicit soft skills learning objectives are mainly absent in PSM courses in higher education (Birou et al., 2016;Hoidn, 2017).
The outcomes of this research show the necessity for PSM knowledge, along with hard and soft skills.Institutions of higher education and training profit from anticipating this in their curricula to facilitate their graduates being able to possess 1) knowledge, 2) professional and interpersonal skills, and 3) intrapersonal traits following the definition of Delamare-Le Deist and Winterton (2005) of competency, i.e. "being competent".The findings are also in line with the 2009 declaration of the ministers of education from European countries, which states that fostering innovation and creativity in society is a task for higher education (Leuven/Louvain-la-Neuve Declaration, 2009).
An often-heard question is whether entrepreneurial behaviour, creativity, inventiveness and strategic or holistic thinking can be developed.Kiratli, Rozemeijer, Hilken, de Ruyter, and de Jong ( 2016) have confirmed and provided evidence that establishing a team creative climate in purchasing departments is supportive.Laker and Powell (2011) also acknowledge that soft skills can be influenced.Yet, they argue that compared with the learning of hard skills, soft skills learning encounters higher resistance levels from both the trainees and the management.In their study, Feisel et al. (2011) have found that these intrapersonal skills of experienced professionals and their strategic behaviour are difficult to influence.However, in PSM higher education, Scholten and Dubois (2017) have found evidence that their student-centred method is developing cognitive and professional and interpersonal and intrapersonal skills."Our results show that students who are actively involved in the learning process by taking responsibility, engaging in collaborative learning and by taking the chance to learn from practice are able to develop higher-order learning in relation to content as well as skills that are needed in today's job environment" (Scholten andDubois, 2017, p. 1696).
Therefore, further (longitudinal) research per industry is suggested on which particular competences are a sufficient or a necessary condition for effectiveness.Research is needed on how the necessary interpersonal skills and intrapersonal traits can be trained within the context  in Appendix 5).Legend: Displayed in white colour "small effects" (0 < d < 0.1); in grey colour "medium effects"(0.1 ≤ d < 0.3); and in black colour "large effects" (0.3 ≤ d < 0.5); d (Dul, 2018b, p. 10).
of PSM.Hence, empirical research on incorporating benchmark tools and training methods in firms and higher education is laudable.

Managerial implication: a dedicated purchasing skills management system
This study's ultimate target in business administration is to ensure that a firm's purchasing establishment becomes effective here by putting a skilled team of PSM professionals to achieve the firm's targets.Several managerial recommendations can be derived from the findings: Management recommendation 1: Identify and prioritise the seven PSM targets as input for a training plan.A first contribution of this research is identifying and empirical verifying the seven different purchasing objectives and their linkages with dedicated skills models.This study provided empirical evidence that a universal PSM professional does seem to exist.However, whether the PSM professionals' skills profiles match their PSM organisations' objectives or not has substantial implications for firms.The corporate actors need to know which objectives they should follow in which order.The PSM strategy benefits from being aligned with the corporate strategy as suggested, for instance, by Porter (1980).When the PSM department, respectively the entire organisation, does not have a clear strategy, it would be difficult to (1) set requirements for the staff, (2) establish development or learning objectives and (3) train (or recruit) people who can support the achievement of those objectives.Training personnel without having clear objectives means that people acquire skills they cannot all productively employ.In the worst-case such over and wrong qualified persons may become "frustrated purchasers" (Tassabehji and Moorhouse, 2008).It also implies that a firm that did not define its objectives and prioritise them lacks the maturity to train its personnel effectively.PSM functions will, first, have to define and reflect a firm's strategy clearly and prioritise and combine the seven possible PSM objectives accordingly even as they break these objectives down to individual positions that may have specific target priorities (Frank Henrik Hesping and Schiele, 2015).It can then serve as an input for the skills development plan.
Management recommendation 2: Establish a purchasing skills training programme relying on soft skills and hard skills.This study's second main contribution is the identification of an empirically-backed skills model comprising 15 skill factors.As a managerial implication, the PSM department's management could establish a training programme oriented and these 15 skills factors.Depending on the firm's size, either a company training programme or an external PSM skills training programme is required.For the latter, the training system would be matched to the skill system developed here in this study and can be applied to train employees according to the firm's targets.The importance of soft skills also reflects on careful personnel selection, whose soft skills should be tested.Future research is needed to develop test banks on how to best assess candidate's soft skills, such as networking, results orientation, imagination and sellership skills, which form the backbone of successful application of the other skills.
Management recommendation 3: Establish a purchasing skill controlling system measuring the available training level and the 15 skill factors.Professional HR has been suggested to monitor its organisation's skill level, measuring it and using this input to plan training and recruiting.To apply such a model to purchasing, a comprehensive purchasing skills model is needed.The present research provides the necessary input.The 15 skill factor model can serve as a blueprint for a corporate skills development scheme as proposed in the HR literature (Mansfield (1996); Campion et al. (2011)).According to these models, employees" skills are regularly measured, compiled and compared to the strategically derived target skill portfolio, the gap indicating future training needs.Adding up the individual skills level can create a key performance indicator for a firm's total skills level and development control.

Limitations and future research
This research has its limitations.The effect sizes of the NCA of the skills factors on the seven PSM objectives are "medium" (Dul, 2018b).It means that there must be additional factors influencing PSM professionals" effectiveness.None of these skills or the organisational conditions and systems related to them could be covered in this study.Future research would benefit from exploring these relationships and directly control for organisation and management support as suggested, for instance, by Tassabehji and Moorhouse (2008).
Although the 15 factors identified show a consistent picture with reasonable VIFs and Cronbach's Alphas, the decision to not discard items with lower factor loadings led to lower AVEs (Average Variance Extracted).Similar factor loadings and AVEs are found with the factor analysis of Giunipero and Pearcy (2000).On the other hand, keeping the items in the list allows readers to make their own choice, for example in the moment of course design, by prioritising high loading items depending on the depth of the training unit/course they are designing.
An additional limitation is that the population that gets covered by this study can be described as predominantly Western European and is based primarily in the Netherlands (58 per cent), France (14 per cent) and German-speaking countries (11 per cent), which may have caused cultural biases.However, literature shows that these countries are representatives of three distinct cultural clusters (Kale, 1995).Future research might want to test for cultural aspects in purchasing skills.For instance, the prominence of the soft skills factor "result orientation" and there in particular the item self-assurance (being assertive) is quite strongly linked to Hofstede's masculinity index, which might not score so prominent in other cultural settings.
Another limitation is that the skills items and achievement statements have been self-assessed by the participants.The participants joined the anonymous survey out of their interest, and there is little indication of biased results.That said, in self-rated surveys, there is always a risk that the participants overstate their skills.Such overstatement is known as the Dunning-Kruger effect, i.e. unskilled professionals have the tendency "to hold overly favourable views of their abilities in many social and intellectual domains" (Kruger andDunning, 1999, p. 1121).On the contrary, skilled professionals tend "to underestimate their performance relative to their peers" (Kruger andDunning, 1999, p. 1126).If the possibility of this effect applied to PSM professionals, this study would be subject to underestimating the differences in performance.Indeed, negative signs are found in the OLS regression.Using peer assessment or a superior's evaluation could test for the relevance of the Dunning-Kruger effect.
From a methodology perspective, NCA's employment, a relatively young method, has its risks.Qualitative Comparative Analysis (QCA) is a broader spread method that could have served as an alternative for NCA.Recently, there has been some controversy in this regard which suggests that QCA would be more suited to studies like the present one because of NCA's limited ability to identify sets of necessary conditions, i.e. combinations of conditions that need to be tested rather than testing individual conditions (Dul et al., 2018;Thiem, 2018).That said, it should also be noted that this assertion has been refuted by Dul et al. (2018), advising that the two methods have a different target.This study's target was to test if particular individual skillsand not combinations of skills are necessary.Therefore, NCA was applied in this research.There is evidence that even with skewed predictors, "NCA can give large necessity effects" (Sorjonen and Melin, 2019, p. 2).
Employing a QCA that could handle the model's complexity could identify new combinations of effectiveness factors that our method could not distil.QCA is not as applicable for larger sets as in this study.The outcomes of 366 respondents, 15 competence factors and seven objectives on five-point scales would lead to an immeasurable QCA outcome of about 40,000 data points.Moreover, it uses other methods for the calculation than the necessary conditions analysis in QCA.The "ceiling line" in QCA is a static diagonal, whereas this line depends on Customer-oriented (Erpenbeck and Scharnhorst, 2005;Heyse et al., 2004;Schiele, 2007) .429 Factor 2 Result orientation α = .904 Self-assurance -Being assertive and having self-esteem - (Erpenbeck and Scharnhorst, 2005;Heyse et al., 2004) .688Poise -Being (self)confident - (Erpenbeck and Scharnhorst, 2005;Heyse et al., 2004) .658Result-orientated action-taking -Aiming on effectiveness - (Erpenbeck and Scharnhorst, 2005;Heyse et al., 2004) .514Ability to Resolve Conflicts -Being able to avoid and resolve conflicts - (Erpenbeck and Scharnhorst, 2005;Heyse et al., 2004) .501Proactivity -Being anticipatory, change-oriented and self-initiated behaviour in situations - (Erpenbeck and Scharnhorst, 2005;Heyse et al., 2004) .488Power of Persuasion -Having influential skills without using power - (Erpenbeck and Scharnhorst, 2005;Heyse et al., 2004) .487Willingness to take risks - (Erpenbeck and Scharnhorst, 2005;Heyse et al., 2004) .465Capacity to Advice -Having consultancy skills - (Erpenbeck and Scharnhorst, 2005;Heyse et al., 2004) .459 Factor 3 Imagination α = .815 Creativity -Being creative in professional life (Erpenbeck and Scharnhorst, 2005;Heyse et al., 2004) .646Inventiveness -Being imaginativeness (Erpenbeck and Scharnhorst, 2005;Heyse et al., 2004) .605Willingness to Learn -Being professionally curious, motivation to learn continuously (Erpenbeck and Scharnhorst, 2005;Heyse et al., 2004) .529Holistic Thinking -Ability to think and act holistically (Erpenbeck and Scharnhorst, 2005;Heyse et al., 2004) .494 Factor 4 Sellership α = .825 Salesperson skills -Having acquisition strength and having canvassing ability (Erpenbeck and Scharnhorst, 2005;Heyse et al., 2004) .706Sellership Skills -Having the drive to sell and establish trust with a customer - (Schiele, 2007) .661Personality Characteristics Development -The ability to develop the own soft skills continuously, qualities and traits (e.g.persuasiveness, creativeness, entrepreneurial, adaptability) - (Schiele, 2007) .536 Cross-cultural Awareness -The ability to become aware of cultural values, beliefs and perceptions of the own and other one's cultures - (Schiele, 2007) .457Customer orientation -Being focussed on the internal customer or internal user group - (Giunipero and Pearcy, 2000;Mulder et al., 2005;Tassabehji and Moorhouse, 2008) .419 Factor 5 Cross-functional cooperation α = .834 Cooperating with department Research & Development -Knowing basics about Research & Development and knowing how to establish/maintain relationships - (Schiele, 2007) .673 Cooperating with the department Quality Management -Knowing basics about Quality Management and knowing how to establish/maintain relationships - (Schiele, 2007) .622 Cooperating with the department Marketing Management -Knowing basics about Marketing Management (or Public Relations) and knowing how to establish/ maintain relationships - (Schiele, 2007) .542 Cooperating with the Logistics and Storage departments -Knowing basics about Logistics and Storage and knowing how to establish/maintain relationships - (Schiele, 2007) .539 Cooperating with the department Production/Operations -Knowing basics about Production/Operations and knowing how to establish/maintain relationships - (Schiele, 2007) .513 Factor 6 Forecasting skills α = .729 Forecasting of the demand -Planning of annual demands, based on the sales forecast and experience as input for annual negotiations - (Schiele, 2007;Tassabehji and Moorhouse, 2008) .749 Enterprise Resource Planning/Material Requirements Planning/Advanced Planning and Scheduling -Material Requirements Planning/Advanced Planning and Scheduling/IT skills necessary to extract planning data from employed ERP system - (Schiele, 2007) .717 Supply Chain Analysis and Planning -Analysis and planning not only of the immediate supply market but consideration of the entire supply chain - (Giunipero and Pearcy, 2000;Mulder et al., 2005;Schiele, 2007;Tassabehji and Moorhouse, 2008) .422 Factor 7 Cost focus α = .863 Solicit Offers (RfQ/RfP/RfI) -Request for Quotation (RfQ)/Proposal (RfP)/Information (RfI) -Inviting suppliers to submit a bid, which meets the requirements as laid down in the request - (Schiele, 2007) .603 Making cost analyses -E.g. the calculation of the total costs of ownership or other cost calculations - (Giunipero and Pearcy, 2000;Mulder et al., 2005;Tassabehji and Moorhouse, 2008) .532 Cost Reduction Techniques -Act of cutting costs to improve profitability (e.g. by analysis and statistics) - (Schiele, 2007) .519Purchasing knowledge -Knowledge on purchasing in an organisation, e.g.Systems, organisational agility and best practice knowledge - (Schiele, 2007) .515Global Sourcing/Supplier Acquisition -Sourcing materials, processes, designs, technologies and suppliers from global market/acquiring new global suppliers - (Schiele, 2007) .507 Negotiation the specific terms -Negotiating considers all commercial and legal terms that need to be settled satisfactorily with the supplierpre-study outcome - (Giunipero and Pearcy, 2000;Mulder et al., 2005;Schiele, 2007;Tassabehji and Moorhouse, 2008) .481 Factor 8 Contracting skills α = .874 Working together with the Legal department -Knowing basics about legal aspects, like contracts, compliance and legal consequences and knowing how to establish/ maintain the relationships - (Giunipero, 2000) .682 Contract Development -Designing of contracts, application of standard vs customised contracts - (Giunipero and Pearcy, 2000;Mulder et al., 2005;Schiele, 2007;Tassabehji and Moorhouse, 2008) .681 Contract Management -Monitoring and enforcing the contracts after they have been signed - (Schiele, 2007) .532Claims Management -Claims management deals with opportunistic suppliers who tend to increase their margin with extra work apart from the contract.The negotiated price is too low, and the suppliers" margin comes from subsequent extra charges - (Schiele, 2007) .514 Evaluate Offers & Supplier Selection -Evaluating the offers and selecting a supplier (or more) with whom the delivery of the product will be negotiated, including the tendering process (diversity, ethical issues, and cost/price) - (Schiele, 2007;Van Weele, 2009) .514 Corporate Social Responsibility -A business model that requires active compliance with the spirit of the law, ethical standards and national or international norms - (Schiele, 2007) .450 Developing specifications for supplies -Specifying the requirements and needs for supplies - (Giunipero and Pearcy, 2000) .403 (continued on next page) K. Stek and H. Schiele Supplier Development -Cooperation with suppliers to improve their processes and product capabilities - (Schiele, 2007) .656Supply Risk management -The implementation of strategies to manage both every day and exceptional risks along the supply chain to deal with risks and uncertainties caused by, or impact on, logistics related resources or activities in the supply chain - (Schiele, 2007;Wieland and Wallenburg, 2012) .595 Supplier Evaluation -Process of measuring and monitoring the performance of current suppliers - (Schiele, 2007) .577Supplier Relationship Management -The ongoing management of the suppliers after contracting/strategically planning for, and managing, all interactions with suppliers - (Schiele, 2007) .576 Early Supplier Involvement -Involving the supplier in the new product development process from a very early stage - (Schiele, 2007) .556Strategic Business Partner -The process of becoming a preferred strategic business partner with your supplier - (Schiele, 2007) .537Sustainability -Sustainable purchasing: considering environmental, social, ethical and economic issues in the management of the organisation's external resources - (Miemczyk et al., 2012;Schulze and Bals, 2020;Schulze et al., 2019) .437 Factor 10 Innovation sourcing α = .855 Pooling Planning & Organising -to bundle the entire demands of the organisation(s).Pooling requires careful planning, demand identification and the application of organisational solutions (lead buyer concept, centralisation, purchasing councils) - (Schiele, 2007) .664 Supply Market Analysis -Analysing the supply market.Who are the possible suppliers?What is the competition level between suppliers, and what is their market power?- (Giunipero and Pearcy, 2000;Mulder et al., 2005;Schiele, 2007;Tassabehji and Moorhouse, 2008) .559 Innovation Sourcing -a systematic scan of the solutions available on the supply market - (Giunipero and Pearcy, 2000) .541Category Strategy Development -Development of the sourcing strategy for a category or family of purchasing goods, including strategic analysis and category classification/portfolio management (e.g.Kraljic) - (Schiele, 2007) .503 Stakeholder Relationship Management -Being focussed on all internal and external stakeholders - (Tate et al., 2010) .464Innovation implementation -Implementing suppliers innovations in the own organisation - (Luzzini and Ronchi, 2011) .422 Factor 11 Analytics α = .835 Statistical Analyses -Applying statistical analyses methods - (Schiele, 2007) .704Big Data Analyses -Uncovering hidden patterns, correlations and other insights from large amounts of data using specific statistical big data analyses methods - (Waller and Fawcett, 2013) .668 Performance Measurement and Follow-up -Continuous monitoring of target achievement, incl.project controlling (Performance of suppliers is part of supplier evaluation) - (Schiele, 2007) .606 Set key performance indicators (KPI ′′ s) -Defining key performance indicators/objectives/targets and implementing them - (Schiele, 2007) .578Portfolio Analysis Support -Analysing the purchasing portfolio, describing a classification of purchases, and supporting the portfolio management - (Schiele, 2007) .494 Factor 12 Leadership and Personnel α = .936 Employee Performance Measurement -Monitoring and evaluating employees: knowledge on employee performance measurement - (Schiele, 2007) .746Employee Integration and Development Plan -Knowledge about the HR process of employee integration and its development plan - (Schiele, 2007) .739Personnel Selection Process -Knowledge about the HR process of selecting new personnel - (Schiele, 2007) .729Training personnel -Actively giving structured training and education to improve the knowledge and skills of colleagues - (Schiele, 2007) .624Leadership/managing personnel -Managing employees in teams - (Schiele, 2007) .611Purchasing Roles and Job Profiles -Defining and designing different roles and job profiles of purchasers - (Schiele, 2007) .560Managing change processes -The ability to lead a team or group successfully through a change process - (Giunipero and Pearcy, 2000;Tassabehji and Moorhouse, 2008) .470 Working together with the Department of Human Resources Management -Knowing basics about Human Resources Management and knowing how to establish/ maintain relationships - (Schiele, 2007) .465 Factor 13 Organisational insight and governance α = .878 Project Management Skills -Initiating, planning, executing, controlling and closing the work of a project team - (Schiele, 2007) .526Add value to the organisation -Knowledge on the added value of purchasing to the organisation and the importance of purchasing to the organisation - (Schiele, 2007) .488 Process Management -The design of processes and the updating as well as reading and understanding processes - (Schiele, 2007) .485Team Ability Skills -Having the ability to cooperating with others in a team - (Schiele, 2007) .485Corporate Governance -Knowledge on how organisations are governed, including board, the role of an advisory board, stakeholders et cetera - (Schiele, 2007) .453Optimisation of Purchasing Processes -Purchasing Process Improvement - (Schiele, 2007) .452Position of Purchasing in Organisation -Knowledge on how to ensure that purchasing plays an adequate role in the organisation - (Schiele, 2007) .427 Factor 14 Automation α = .812 Automation -Working on the automation of purchasing processes - (Schiele, 2007) .773Procurement IT Systems/e-procurement applications -Knowing the working of a computerized designed e-procurement system - (Schiele, 2007) .772 Factor 15 Technical skills α = .683 Technical knowledge of products and production systems -Understanding the technical aspects of the own products and production processes - (Giunipero and Pearcy, 2000;Tassabehji and Moorhouse, 2008) .644 Technology Planning -knowledge on the technological requirements of its own company - (Giunipero and Pearcy, 2000;Schiele, 2007) .586Commodity and Domain-Specific Knowledge -Knowledge on a particular purchasing domain, e.g.different industries, services, construction, purchasing of health, et cetera.- (Giunipero and Pearcy, 2000;Mulder et al., 2005;Schiele, 2007;Tassabehji and Moorhouse, 2008) .526 Note: α is Cronbach's Alpha for the construct.Appendix 6 shows the effect size of the subsequent factors on the other 14 factors.(0 < d < 0.1 "small effect"; 0.1 ≤ d < 0.3 "medium effect"; 0.3 ≤ d < 0.5 "large effect"; d ≥ 0.5 "very large effect" (Dul, 2018b)).The factors are ranked based on their means (right column), meaning that, for example, "Factor 1 -Networking" has the largest effect sizes on the other factors.The mean is 0.415, which is on average a "large effect"."Factor 14 -Automation" is not a necessary condition, showing a mean of 0.069 ("small effect").Moreover, significant relations are displayed in bold.For the smallest effect sizes, no significance was found (factors 3, 10 and 5).Appendix 6 is the underlying dataset for the surface graph in Fig. 1.
(continued on next page) K.Stek and H. Schiele

Table 2 -
Outcomes of the factor analysis -A comprehensive PSM skills taxonomy.

Table 4 -
Matrix showing the significant effect sizes of the direct necessary conditions for each PSM objective.

Table 5 -
Training and education focus according to objective.