The neglected organisation of failure: Activities,infrastructures and subjectivities

Abstract

The aim of this article is to explore a neglected issue – the organisation of failure in projects. The existing literature has primarily focused on explaining the causes, factors and circumstances that lead to failure, and how organisations account for and make sense of failure. However, this is only a partial explanation, particularly for organisations where decisions about project success and failure are routinely made. Drawing on a pragmatic approach that emphasises social agency as the ongoing determination of an inherently uncertain reality by knowledgeable actors, I describe the processes by which failure is managed in organisations so that it is both part of the productive workflow and recognisable and acceptable to all stakeholders. I highlight the critical role of evaluation infrastructures, allowing the assessment of whether a project's outcomes are in line with its objectives or the resources used, yet remaining largely invisible. The manipulation of these infrastructures allows sponsors to achieve the desired success/failure ratio of the projects they support without having to change their value objectives. The subjective dimension of the sponsor's organisation of failure has also been shown to play an important role, which can lead to failure being marginalised or made ambivalent. In developing the argument, I have drawn on both theoretical literature and empirical research in four different areas of organised labour – start-ups, platform work, development aid and environmental education.

The exploration of failure in organisation research is primarily concerned with explaining the causes, factors or circumstances that lead to failure, or how people explain and make sense of it. Failure is understood as something that happens to an organisation and needs to be addressed in order for the organisation to continue to thrive or at least survive (Habersang et al., 2019; Helmig et al., 2014; Kibler et al., 2017; Mellahi and Wilkinson, 2004, 2010). Much less attention is paid to failure as routine and intended outcome of organisational activity. For example, a food delivery platform needs to arrange for the failure of its ‘underperforming’ agents to be organisationally smooth and not disruptive to its operations; a charitable trust needs to ensure that the failure of the humanitarian projects it supports does not damage its reputation; an investment company is faced with the problem of how to legitimately organise the failure, that is, the termination, of a start-up in which it has invested substantial capital. In contrast to success, making failure happen organisationally is not trivial because the failed party is often damaged, they may not accept the decision, the shadow of failure may fall on all stakeholders, or the people in charge may not want to attribute failure if they do feel it is not deserved.

As the world of work becomes increasingly projectivised (Boltanski and Chiapello, 2007 [1999]; Lundin et al., 2017; Mats and Godenhjelm, 2023) and projects require evaluation, decisions about project success and failure become more frequent and routine. Although many activities are being transferred to projects, the arrangements for the declaration of project success or failure are, with few exceptions, the responsibility of the permanent organisations. Furthermore, it is not only a matter of conducting an evaluation. Someone in the organisation has to decide if, when and how the evaluation will take place, who will carry it out and what the results will be. The result must then be communicated to the stakeholders and they must be made to acknowledge and accept the failure. Therefore, as I will show later in the paper, it can sometimes be advantageous for an organisation to leave the evaluation to the project team or not to carry it out at all and consider all projects implemented as successful.

This paper aims to fill a gap in organisation research by focusing on how legitimate project failure is coordinately produced in the context of project-based work. It examines the ways in which failure is organised not only because of the growing importance of projects as a form of ‘temporary organisation’ (Burke and Morley, 2016; Lundin and Söderholm, 1995) but also because projects are prone to failure or, as Malpas and Wickham argue, always fail in some way (1995: 39). Research to date has produced a great deal of descriptive knowledge about the specific causes and consequences of failure in different fields, or guidelines for its prevention, but has struggled in its theorisation (Gupta et al., 2019: 282; Hughes et al., 2016: 3; Lindahl and Rehn, 2007; Pinto and Mantel, 1990; Vanderheiden and Mayer, 2020). The contribution of this text will be an argued impulse to shift the attention from the description of more and more causes and consequences of failure to the common features of failure organisation in projects, which will allow a better theoretical and practical understanding of the failure process. ¹

Theoretically, I draw on pragmatic sociology, which emphasises social agency as the constant determination of an inherently uncertain reality by knowledgeable actors (Barthe et al., 2013; Simpson and den Hond, 2022). Failure, like success, is achieved through the coordinated activities of agents within an organisation, and this organisational outcome is never guaranteed in advance. I combine the classical pragmatic approach with ideas from the French (neo-)pragmatic school that explains project-based work organisation and the key role of tests in assessment and evaluation (Boltanski and Thévenot, 2006 [1991]). In developing my argument, I have drawn primarily on the existing literature and partly on the results of empirical research in four different areas of project-based work – start-ups, platform work, development aid and environmental education, which will be used to abductively suggest a frame of reference for analysing how project failure is organised. ²

The structure of the paper will be as follows: First, I will trace the organisation of project failure in the existing literature and propose how to build on it. I will then briefly introduce the pragmatic approach to the projective world and what project failure means in this context. I then develop my argument by outlining the coordinated activities that create a project and the key role of evaluation infrastructures in the organisation of project failure. Finally, I focus on the intersubjective dimension of the organisation of failure, particularly on the part of project sponsors, because it has important implications for the organisation of failure, and also because this dimension is so far missing in the literature. I conclude by highlighting the contribution of failure organisation analysis to sociological and practitioner audiences.

Tracing the organisation of project failure in the literature

The conventional definition of a project as ‘a temporary endeavour undertaken to create a unique product, service or result’ (PMI, 2017) leads most organisation and project scholarship to analyse how the success of these temporary endeavours is achieved, judged, attributed or understood. Although the nature of project success tends to be under-defined theoretically (Ika, 2009; Pinto et al., 2022), success is seen as a natural goal of projects, and knowing how to attain it safely is therefore highly valued by scholars and especially practitioners (Ika and Pinto, 2022; PMI, 2017; but see Malpas and Wickham, 1995 and Lindahl and Rehn, 2007 for a heterodox perspective). Consequently, this means researching the various obstacles, including organisational ones, that complicate or directly prevent projects from achieving success. When failure becomes a research topic, it is also with the aim of identifying its causes, so that lessons can be learned and failures avoided in the future (see Gupta et al., 2019 or Schmidt, 2022 for a review). Schmidt lists thirty definitions of project failure in the literature between 1975 and 2020 and, based on his review, attempts to ‘to address the question of which performance criteria for a certain type of project, a certain type of stakeholder will associate with an ex-post attribution of failure’ (Schmidt, 2022: 257). He thus understands project failure as stakeholder attribution based on performance criteria. However, the attribution process itself, the organisational link between performance, criteria and stakeholders, is taken for granted and not explored at all. In determining failure, as with success, the literature is concerned with establishing both universal and specific criteria, ideally, such that failure can be anticipated and prevented early in the project. The organisational environment of projects is typically viewed as a socio-cultural context that, while influencing project success or failure, is rarely the subject of research in its own right (Engwall, 2003; Helanummi-Cole and Jalan, 2023).

Systematic theoretical and empirical research that goes beyond understanding the project as a purely outcome-oriented activity and conceptualises project failure more broadly than failing to achieve a given outcome is still lacking. However, there is a discernible thread running through several texts on failure that directly addresses the role of organisational routine in project failure. In their widely cited book, Meyer and Zucker (1989) developed a theory of permanently failing organisations. These organisations do not meet performance expectations over a long period of time but continue to operate for organisational reasons. In the spirit of the argument developed here, they can be seen as organisations in which failure has not been successfully organised. The explanation for this failure to deliver failure lies not in the same causes as their underperformance but in the diverging interests of two groups of organisational stakeholders: owners have an interest in the performance of the firm, while dependent actors have an interest in its survival. Whereas in times of prosperity, the interests of the two groups are aligned, in times of adversity and declining performance the owners want to close down the organisation, but face the united opposition of the dependent actors, for whom the continuity of the organisation is an existential interest (Meyer and Zucker, 1989: 91–107). This theory is inspiring for research on the organisation of failure in projects because in projects there are also two groups of actors, project sponsors and project contractors, whose interests coincide when the project is successful but tend to diverge when problems arise. Thus, Meyer and Zucker's argument justifies the assumptions that (a) the failure of a project is an organisationally distinct phenomenon from its success, (b) organising failure is not a trivial matter and may pose a challenge to the organisation and (c) the organisational relationships between project sponsors and project contractors translate into the likelihood of failure.

Almost at the same time, Pinto and Mantel (1990), in their seminal article on the causes of project failure, point out that failure depends not so much on the results of the project but primarily on the evaluation method chosen by the organisation. Thus, they advance the idea of failure as a result of organisational relationships and negotiations between project stakeholders. They are therefore sceptical about finding universal causes and criteria for failure that apply across organisations. What is perceived as a project success in one organisation may, in their view, be judged as a failure in another. They argue that there is a need for research into how managers define failure and how the parent organisations make judgements about it (Pinto and Mantel, 1990: 275). While Pinto and Mantel are correct in their assertion that different criteria are used in different organisations, these criteria are not arbitrary and organisationally unique; as I will argue below, certain organisations typically resort to specific types of criteria and project evaluations and thus corresponding definitions of failure.

The issue of the non-triviality of organising project failure was addressed by Dilts and Pence (2006), who examined the role of decision makers in organising project failure. They observed a difference in the evaluation of projects by project managers and sponsor representatives (executives) and found their different, albeit generally negative, motivations for initiating the failure process – while the project manager is not interested in terminating the project because it affects them personally, the executive is discouraged from initiating the termination because of the negative impact on their reputation as they were the ones who initiated the project (Dilts and Pence, 2006: 382), The finding thus extends Meyer and Zucker's (1989) theory of persistently failing organisations in that, in the case of projects, the sponsoring organisation may have an interest in terminating underperforming projects, but the individuals responsible for the decision do not; their interest is therefore aligned with that of the dependent actor, in this case the project manager. Nevertheless, ‘project managers feel a more clear and present danger of repercussions for project termination than do executives’ (Dilts and Pence, 2006: 394). The authors, like Pinto and Mantel (1990), are sceptical about the search for general criteria capable of distinguishing between successful and unsuccessful projects, pointing out that any criteria for success and failure must be ‘very site and situation specific’. However, this statement implicitly assumes that general patterns of failure can be found – in fact, they suggest one – but they are to be sought in the organisational accomplishment of failure, not in the assessment itself.

The translation of organisational relationships between project sponsors and project contractors into the propensity to fail has been analysed by Zarei (2017). Although his approach is conventional and primarily interested in whether power asymmetries have a negative or positive impact on project success, his work has provided insights relevant to the study of the organisation of failure. In the context of public infrastructure projects, he distinguishes between central agencies, which play the role of project sponsors and therefore have ‘the authority to start, change or stop a project’ (Zarei, 2017: 186), and delivery agencies, which implement projects and therefore play the role of dependent actors. Akin to Meyer and Zucker (1989), Zarei is interested in how, through information asymmetry, delivery agencies acquire informal authority and thus the ability to subvert power asymmetries in their favour, and thus the power to influence organisational decisions about the success or failure of a project: ‘In projects, informal authority may originate from information, expertise, or reputational power of an agency to manipulate other organisations’ behaviour.’ (Zarei, 2017: 189). Given the observed reluctance of central and delivery agencies to participate in organising the failure of projects in which they are involved, he suggests outsourcing project evaluation to external consultants who do not have organisational biases as a strategic solution. In doing so, he identifies one of the organisational failure arrangements that resolve not only the power asymmetry but also the subjective barriers to the implementation of project failure.

Tracing the routine organisational aspects of project failure in the literature has shown that although most of the literature on organisational project failure is devoted to identifying its performative or evaluative causes, we were able to find relevant research that provided important insights into the organisational aspects of failure. Considering projects as a form of temporary organisation, we can build on Meyer and Zucker's (1989) theory of permanently failing organisations, which could be reinterpreted here as the theory of ‘difficult-to-fail projects’. This hyperbole alludes to the fact that the organisation of project failure is not easy to routinise and can pose a challenge to the organisation, mainly for reasons associated with the relationship between project sponsors and contractors.

The importance of project failure in the projectified world

In order to understand the organisation of failure, it is necessary to consider the broader organisational context in which failure is routinely achieved. In contemporary organisational sociology, the projectified society prevails as a general interpretive framework, in which permanent and temporary organisations coexist symbiotically (Hodgson et al., 2019; Lundin et al., 2017; Mats and Godenhjelm, 2023; Sydow and Windeler, 2020). Because I build my argument on a pragmatic understanding of the project, I will adopt neo-pragmatic theory of the ‘projective world’ introduced by Luc Boltanski and his colleagues (Boltanski and Thévenot, 2006 [1991]; Boltanski and Chiapello, 2007 [1999]; Barondeau and Hobbs, 2019 ). They argue that projective world has emerged as a consequence of ‘the profound transformation in the managerial discourse and justifications of the development of capitalism since the mid-1970s’ and envisages ‘the emergence of new representations of society, of novel ways of putting people and things to the test, and therewith of new ways of succeeding or failing’ (Boltanski and Chiapello, 2007 [1999] : 4). It is the emphasis on new ways of testing projects and thus new ways of organising success and failure in a projectified society that makes this theory particularly valuable in the context of this study. Nonetheless, I will show that their notion of the projective world treats project failure only formally, as a network property rather than an organisational achievement, and therefore does not adequately explain how failure is organised in projects and how this affects its distribution.

In the networked projective world, projects form ‘temporary pockets of accumulation’ that have the capacity to create value (Boltanski and Chiapello, 2007 [1999]: 103–129). People and organisations are therefore motivated to initiate or join them. Since projects appear and disappear like bubbles in boiling water, the crux of the matter is how to get into projects, what to do in them and how to exit them. In a pragmatic vein, Boltanski and Chiapello argue that the failure of a project becomes significant only in its consequences, as a reduced chance of future connections, i.e. involvement in new projects (2007 [1999]: 126). The only sanction usually available to sponsors is a ban on participation in further competitions. It follows that in projectified work the substantive fulfilment of the project is not important in itself; the main purpose of carrying out the project is to secure participation in the next project. To some extent, this is supported by research on platform work, which shows that for workers, success is more about securing a steady influx of profitable contracts than about high scores in ratings (Cameron, 2022, Rahman, 2021). The startup mantra of ‘fail fast, fail often’ also reminds founders that their goal is not to bring every project to a successful conclusion but to keep trying projects until one sticks (Pontefract, 2018). In the non-profit sector, the professionalisation of project work can lead to ensuring project success at the expense of substantive fulfilment (Dolšak and Prakash, 2022). In this concept of project organisation, projects are successful in a neo-Darwinian way: only the project that increases its participants’ chances of getting another project is successful. In my view, this is an inadequate sociological understanding, because awarding and closing of projects is not a blind evolutionary mechanism but a deliberate strategy of value creation and governance, especially for the projectified organisation of work.

It is therefore important to consider that projects are initiated by reflexive actors to create value, for example, a product, a service, community empowerment or environmental awareness. They include goals to be achieved through the activity and, at the end of the project, more or less comprehensive evaluation procedures assess whether the goals have been achieved. Clearly, these evaluation processes also serve to determine the value of the project participants and their reputation. There is a noticeable concentration of power in these processes, which, as Boltanski and Chiapello (2007 [1999]) argue, transform performance into legitimate status (Ballesteros, 2021; Smillie, 2013). But that does not mean that projects are meaningful to their participants only in the context of potential future projects. As Lundin et al. remind us, ‘[w]ork in projects, which promises to be interesting, could simultaneously cause stress from time limits and responsibilities laid on the individuals concerning both business outcomes in the specific projects and the need to constantly perform to ensure their own future employability’ (2017: 133). Thus, in explaining the failure of project work, we need to consider both the substantive nature of the activities derived from the project objectives and the awareness of the temporality of the project in terms of the need for further projects.

Boltanski and Chiapello's theory of the projective world provides a deep insight into the contemporary organisation of work as a network of projects, still we need to look more closely at how and who creates projects, what happens in projects and how this relates to their success or failure. We are not seeking to resolve the question, so preoccupying French pragmatists, of who is a ‘great person’ in the projective world and who is not. Rather, we are concerned with finding out what specific actions and intersubjective experiences of actors within a project lead to its being declared a success or a failure or something else. For example, it makes a difference to an NGO if its project stops because the government that funds it runs out of money, because the evaluator finds that the organisation has not carried out the planned activities, or because it receives feedback that its activities are not meaningful to clients. Nor does the network model explain why in some areas projects are awarded to virtually everyone who applies, while in others they are highly selective; why, for example, only a minimum of international development projects are formally evaluated as failures, while at the same time, there is a general critical view that their substantive contribution to development is negligible. There is a need to unpack the black box of Boltanski and Chiapello's notion of the project and to examine the activities of the participants in relation to the goals.

Project-generating activities and project failures

From a management perspective, a project is defined as a short-term attempt to achieve a unique outcome (PMI, 2017), or organisationally, as a form of temporary organisation describable by notions of time, task, team and transition (Lundin and Söderholm, 1995). From the pragmatic point of view we are developing here, the overarching concepts are activity, emergence and relationality (Mead, 1934: 329): a project emerges as a recognisable set of interrelated activities. However, this is too general a definition that applies to most social objects. What distinguishes a project is the kind of activities that constitute it. From what we know so far about projects, we can identify the following as the most basic: providing goals and resources, carrying out the work to achieve the goals and evaluating whether the outcomes meet the goals. The activities may be performed by the same or different actors and may be linked to other activities such as accounting or reporting (cf. Scott, 2023). Importantly, a project emerges when actors relate these three core activities to each other. For example, entrepreneurship is not commonly understood as a project but as an activity in its own right. When investors conclude that the entrepreneur they have supported has failed, they assume that the person in question (a) knew or at least should have known the expectations associated with the business, (b) had sufficient material and symbolic resources to undertake the business and (c) that they are qualified to judge the outcome of the business.

Therefore, in the pragmatic approach, we do not understand a project as a temporary organised effort to achieve a result but a project as a social object emerges and persists through action when and to the extent that actors intersubjectively recognise the activities of provision, carrying out and evaluation in their concrete interconnectedness. In analysing failure, we will accordingly focus on the activities – interactions and transactions (Simpson and Den Hond, 2022) – that practically and symbolically constitute the project as a particular social act (or joint action, cf. Blumer, 1966). The roles of the actors involved in the projects are determined by their activities; it is not the reverse, as is commonly believed. The provision of goals and resources creates project sponsors (donors or funders); the carrying out of project work makes the contractor (individual, project team or organisation); the activity of evaluation establishes the evaluator. By observing how participants in project-generating transactions make sense of their actions, and how they interpret the actions of others in light of those actions, we can reveal how failure is co-organised and intersubjectively ratified as a reality.

This new perspective on projects allows us to rethink what has been observed so far about the organisation of failure: (a) general patterns of project failure can be identified if we focus on its organisational aspects, (b) producing legitimate failure is not a trivial matter and can pose a challenge to the organisation and (c) the relationship between project sponsors and project contractors is reflected in the propensity to fail (Dilts and Pence, 2006; Meyer and Zucker, 1989; Zarei, 2017).

There are two critical moments and corresponding tests that make the possibility of project failure a reality: project acceptance, that is, an agreement between the prospective sponsor and the contractor to collaborate in order to achieve the stated objectives, and final assessment, when the sponsor decides whether the contractor has achieved the stated objectives. In general, failure is more common in the project acceptance test than in the project implementation test. This is true of a platform driver's gigs, a start-up's efforts to attract an investor, or a non-profit's pursuit of a project. The point is not to say that it requires more effort to design a project than to execute it but that the chances of success are usually lower for the former than for the latter. Leaving aside the obvious limitations of the resources available to the sponsor, we can see a simple logic behind this asymmetry – it is in the sponsor's interest to support only those entities that it reasonably believes will be successful in delivering the project. At the same time, it is not in the sponsor's interest to undermine the credibility of its previous selection by a strict evaluation at the end of the project and, moreover, by making it clear that a significant proportion of the funds it has provided for the project have not been used productively and should be written off. It is also important to realise that in projectified work, the sponsor of a project is usually also the contractor who bills its own sponsor for its activities (cf. Manning, 2023). For example, a shared housing platform acts as a market access provider in the projects of people or organisations, that is, the contractors who deliver the service, but at the same time, the platform is a contractor in the project of its shareholders or investors who provide it with funds and evaluate its performance as a contractor. Similarly, sponsors of development projects are themselves contractors to their donors, or projects distributing state agencies are contractors to the government. Because of this interdependence, the approach to entry and exit testing is different, and so is the way in which they produce successes and failures.

The asymmetry between success rates in entry and exit tests, as well as differences in project failure rates in different areas of work, for example, between the for-profit and non-profit sectors, can be explained by distinct interpretations of failure by sponsors and contractors. As noted by Meyer and Zucker (1989) and Zarei (2017), project failure can have negative consequences for sponsors if it is also perceived as their failure. However, in the start-up environment, sponsors themselves emphasise that failures are normal for projects (Koskinen, 2023). To understand these differences, we can use the concept of ideologies of failure proposed by Berenice Fisher in her analysis of educational systems (Fisher, 1972). She distinguished between non-failure theories ‘which argue that under proper teaching conditions students do not fail or should not be labelled as failed’ and failure theories ‘which argue that even under the best teaching students fail and should be made aware of the fact’ (Fisher, 1972: 322). Arguably, the concept of failure ideologies can be generalised to all areas of organisations where systematic and routine decisions about success and failure are made, including projectified work. In the non-failure ideology, failure of a contractor implies a partial failure of the sponsor as well, whereas in the failure doctrine, the failure of some contractors proves the quality of others, that is, their successful projects (accepted or completed), and thus manifests the sponsor's success. Importantly, the ideologies of failure are unrelated to the goals of the projects, which may remain the same in both cases. In this way, Fisher argues, ‘sponsors have used and continue to use both failure and non-failure theories in their attempts to structure the social world’ (Fisher, 1972: 332). What allows sponsors to tailor the success rate of their projects to their ideology of failure is the calibration of evaluation infrastructures. For example, sponsors of development or education projects, where the ideology of non-failure prevails, often allow contractors to self-evaluate, which obviously leads to fewer failures than external evaluation (Kelly, 2021; Marsden and Oakley, 1990 ). In the start-up world, on the other hand, the ideology of failure is emphasised, and the decision to reject a project or terminate funding is entirely at the discretion of the investors, who do not disclose the technical details of the evaluation, thereby concealing the idiosyncrasy of their decision (Warnick et al., 2018; Zacharakis and Meyer, 1998; Zunino et al., 2022).

The role of evaluation infrastructures in organising failure

Despite their different success rates, all project assessments share a common feature that plays an essential role in determining their results, namely the distinction between the manifest values associated with the project objectives and the underlying infrastructures (Star, 1999) that provide the technical and procedural support for the assessment. This sets these tests apart from other organisational tests that are not overtly linked to value-based objectives, such as safety or competence tests. An example of the project's manifest values is the slogan ‘We care about safety wherever you go’, which a transport platform (a sponsor) presents on its website. Platform drivers (the contractors) are therefore expected to pay particular attention to passenger safety if they are to be successful in securing their micro-projects (‘rides’). On the other hand, driver rating criteria, that is, performance evaluation infrastructures that are not disclosed by the platform and can only be inferred retrospectively from the drivers’ evaluation experience, prioritise other aspects of the job, especially those that are suitable for automated monitoring and increase the provider's profit, such as promptness in accepting orders, sustained driving in the city, customer ratings or cancellation of orders by customers (Möhlmann and Zalmanson, 2017). The duality of goals and infrastructures in project evaluation therefore has important implications for project organisation, project work and the distribution of failures (see Hayes et al., 2018 for the NGO sector). Among other things, it creates a phenomenon of reactivity, whereby those contractors who are knowledgeable about the evaluation infrastructures and focus on them rather than on the stated project objectives are more likely to succeed than those for whom this knowledge is not available (Espeland and Sauder, 2007).

In projects, evaluation infrastructures become a kind of boundary object that allows sponsors and contractors to work together without consensus on certain issues, such as the evaluation system or the nature of the outputs (Star, 2010: 602). Contractors may find some of the evaluation infrastructures used to assess their projects inappropriate, for example too formalistic (KPIs or numbers of participants) or, on the contrary, subjective (panel decisions, self-assessment) or not well understood (algorithms), but they respect them and try to adapt project outputs to them. Sponsors may also have reservations about the evaluation method, in particular, that it does not force contractors to be more accountable, but they are not in a position to provide a better solution themselves. Thus, even without a consensus on evaluation issues, it is possible for projects to be jointly produced, and for some to legitimately fail, thanks to not-quite-visible or conspicuously inconspicuous evaluation infrastructures.

The evaluation of different types of projects may require different infrastructures and, conversely, some infrastructures may be common to various activities. Certain productivity indicators are universal to business, academic and NGO projects, such as tracking completed sub-activities, cost per project participant or performance assessment by a peer review or expert panel. Other infrastructures are very specific, such as the above-mentioned tracking of the speed of order acceptance in platforms, community meetings or the self-assessment of lecturers. Evaluation infrastructures in different fields are often not identical but share a similar logic, such as productivity measures (input-output ratio) or past performance (past performance as a predictor of future performance). This is crucial because it implies that infrastructures are not merely technical transformations of value structures in a project, but have their own procedural logic (Star and Griesemer, 1989). This is evident, for instance, in the gradual adoption of business evaluation infrastructures for evaluating non-profit projects (Hayes et al., 2018).

A well-known characteristic of infrastructure is that it is difficult for ordinary users, in this case contractors and to some extent sponsors, to understand, either because it is hidden or too complex for them (Star, 1999). The reason for making evaluation infrastructures invisible to contractors may be an ‘aesthetic’, aka legitimising one: in the same way that a mess of pipes and wires is hidden behind decorative plaster, evaluators keep the idiosyncrasy of their evaluation to themselves and only third party research can reveal it (see Parmentier and Le Loarne-Lemaire, 2021, for the evaluation of start-up pitches). Secrecy can also prevent contractors from being driven by assessment infrastructures rather than project objectives. Especially in highly competitive fields, knowledge of rating infrastructures is a significant advantage. Platform workers are interested in the inner workings of their evaluations in order to achieve higher personal scores and win more clients than hundreds of their competitors (Rahman, 2021). However, the mere fact that more informed contractors have an advantage over less informed ones would not be a major problem for the sponsor, as it would not undermine the legitimacy or effectiveness of the assessment process. An example of this is the emphasis on participatory evaluation and self-evaluation in the appraisal of development projects (Marsden et al., 1990). On the other hand, if knowledge of infrastructure leads to reactive efforts to one-sidedly conform to or even outwit evaluation, that is, to modify the work of contractors so as to prioritise evaluation criteria and sideline the objectives of project activities, sponsors need to intervene. In such a case, sponsors make the criteria invisible or – as is also common – simply convert the evaluation criteria into explicit project objectives (see Griesemer, 2020).

The intersubjective dimension of failure organisation

If there is something we take for granted about failure, it is its subjective dimension. In the literature, research on the experience of failure in projects has typically focused on studying the negative emotional consequences of failure for contractors (Lindgren et al., 2014; Shepherd et al., 2011, 2014; Todt et al., 2018; Välikangas et al., 2009). Here, we are more interested in the emotions associated with the organisation of project failure than in the consequences, specifically, the affects associated with the activities related to project contracting and evaluation, which are always relational. Failure is thus more accurately described as an intersubjective phenomenon, since we assume that individual experience is created by taking the attitude of the other towards oneself and by becoming aware of one's own role in the social organisation (Mead, 1934: 138ff.). Sponsors take the attitude of contractors into account when selecting and providing project goals and resources, for example, a driving platform demonstrates its knowledge of the attitude of drivers by stating on its website ‘Drive when it suits you and earn what you need’. Sponsors organise joint meetings or seminars to introduce contractors to their attitudes towards project implementation. The contractors, in turn, are aware of the sponsors’ attitudes and try to implement the project in a way that is considered successful by the sponsors. The experiences and emotions associated with failure must therefore be seen as relational, intersubjective phenomena.

For sponsors, depending on the ideology of failure to which they adhere (Fisher 1972), the success or failure of the projects they decide on is in part their professional success or failure. In addition, they are aware of the emotional pressure that failure exerts on project contractors, which indirectly affects them when they decide to fail. Finally, we must not overlook the important role played by the lack of clarity or inadequacy of criteria and evaluation infrastructures, which create uncertainty among decision-makers as to whether they have actually made the right decision. In the following section, I describe how intersubjective experiences are involved in the organisation of failure in different areas of project work.

In platform organisations, project awarding relies on the automated assessment infrastructure applied by the platform (Kornberger et al., 2017). Although the sponsors as individuals are not directly subjectively involved in the evaluation, their awareness of the attitudes and actions of the contractors on the platform leads them to continually develop and modify the evaluation infrastructures so that the success/failure ratio matches the company's strategy. Typically, projects are rated by the platform in aggregate, in the form of an overall numerical rating (0–5, 0–10, 0–100) of the contractor (Rahman, 2021). What is remarkable from an affective point of view is that the rating algorithm is biased towards the maximum, that is, most contractors have good and excellent ratings, which gives them a sense of achievement. However, the ratings are not directly related to the amount of the payout, which tends to be low by default. Here, we encounter a phenomenon similar to that observed by Fisher (1972) as a consequence of the non-failure ideology and referred to by Appadurai and Alexander (2020) as the denial of failure, where the conditions for the declaration of failure are strategically removed. It is not easy for the worker to evaluate and decide whether their work on the platform is as successful as the evaluation would suggest or not, which corresponds to the amount of their earnings.

In contrast to the prevalence of ‘micro successes’ among platform contractors, it is commonly reported in the start-up field that nine out of ten start-ups fail (Griffith, 2014). While it is not easy to define the failure of a start-up, it is rare for start-ups to meet the high expectations placed on them (Neumann, 2017 : 32). Despite the plethora of methods and metrics designed to determine their performance, the practice of start-up evaluation is complex and has a very strong subjective component (Köhn, 2018). From the pragmatic perspective of failure organisation, it is important to identify who is providing the resources for the start-up, whether it is the founder, their friends and family or an external investor. In the first case, the identification and confirmation of the failure is left to the founder and their close circle, and it is clear that the evaluation and decision-making is (inter)subjective. The founder (sponsor, evaluator and contractor in one person) may consider securing an interesting position in the corporate sector as an unintended but successful outcome of the start-up, even though the business plan itself did not materialise and the start-up as such failed (Lusoli, 2021). The decision of an investor to support a business plan is often a very personal one, in which intuition plays an important role and sometimes it is enough to reject a project if the investor and the applicant do not ‘get on’ personally or the ‘chemistry’ is not right (Zacharakis and Meyer, 1998). Cope with colleagues analysed venture capitalists’ narratives of their investment decisions and found no set of formal criteria but a highly personal approach, such as ‘personally I would prefer to back a failed entrepreneur, subject to seeing what the failure was … than a new starter’ (2004: 161). Evaluation infrastructures are embodied and inseparable from the providers and/or evaluators. And while standard infrastructure based on quantitative measures such as employment growth, rate of return and productivity are important in assessing project performance, many investors or their representatives are reluctant to rely solely on them because of their personal relationship with the founders (see ‘personal chemistry’ above). When a start-up gets into existential trouble, as is usually the case, they know that the cause may not be the founders’ mistakes (Cope et al., 2004), and it is subjectively and intersubjectively easier for them to push the contractors to transform (pivot) the business than to close it down (Bajwa et al., 2017).

For the non-profit sector, such as development aid and environmental education, there is virtually no research on sponsors’ decision-making in project selection and awarding; as Barman argues, ‘scholars typically focus not on these institutional funders’ own characteristics but instead on those of their potential recipients in order to explain foundations’ patterns of giving’ (2017: 278–279). Only the perspective of contractors – NGOs or fundraisers – is covered (e.g. Krause, 2014). The data from our research suggests that sponsor representatives subjectively see themselves as different from the ‘passionate’ and ‘quite colourful characters’ of non-profit workers; they see themselves more as managers or officials who view projects from a rational, goal-oriented perspective. They claim that they do not have to rely on intuition to do their work, as investors do, and although they often know their contractors because NGOs and providers form a project network organisation (Brinkerhoff and Brinkerhoff, 2004), unlike investors they do not see these relationships as a helpful factor in their decision-making. A possible reason for this reported distance may be that they are accountable to their superiors or to the donors whose funds they distribute and must be able to justify their decisions in purely rational terms (cf. Lundin et al., 2017; Manning, 2019). Due to the prevalence of the non-failure theory in the sector, the decision not to fund a project is to some extent perceived as a failure by the donors’ staff. Unlike investors, sponsors and evaluators in the not-for-profit sector do not subscribe to the importance of subjective involvement, but the increasing number of unsupported or abandoned projects suggests to them that they are not serving a good enough cause, and they may experience this affectively as their own failure.

Looking at project failures in the non-profit sector, most projects are judged successful by their sponsors (governments, international organisations, donors). However, this contrasts with public discourse, which often questions the real impact of development aid and environmental education projects. Their failure is not a result of donor evaluation, but of the seemingly unimproving global state of development and the environment (Saylan and Blumstein, 2011; Schaaf, 2013: 23). One reason for this is that, like for venture capitalists, it is subjectively uncomfortable for NGO project sponsors to declare a project a failure because it means that the resources invested on the basis of their decision have been wasted and no benefit has been achieved. Furthermore, the third sector is associated with the predominantly value- rather than goal-oriented actions of its actors. As a result, the general strategy is to require internal rather than external evaluation. This means that contractors are responsible for quantitatively or qualitatively assessing their own performance against project objectives (Campos et al., 2011). This distinguishes the NGO sector from others; because contractors are familiar with the evaluation infrastructure, they are able to design their projects to avoid its formal failure (Krause, 2014). This system of formal success might work well if it did not create a subjective sense among contractors that, even if they are successful in their projects, they are failing in their organisation's mission (Dolsak and Prakash, 2022 ). So again we have a form of denial of failure, where formal, institutionally ratified failure is effectively sidelined and replaced by subjective failure.

Conclusion

In this article, I set out to challenge the conventional definition of a project as an organisational structure with well-defined characteristics, which implies project failure as an error caused by either poor team management or team underperformance. As an alternative, I presented a different conceptualisation of project failure as an organisational accomplishment. I drew on a pragmatic understanding of the project as an emergent social object. In this perspective, project failure refers to either the failure of the project to emerge or the declaration of the project as a failure by the sponsoring actor. Both are important in the context of project-based work, where projects are seen as sites of value generation, and where project creation and the transition from one project to another signify individual and organisational success.

I suggest that both the failure to create a project and its poor assessment are the result of transactions between actors who provide the project's goals and the means to achieve them (sponsors), do the work necessary to fulfil the goals (contractors) and evaluate that work (evaluators). An exploration of project creation and evaluation in four different areas of projectified work revealed that evaluation infrastructures are instrumental in determining failure. Sponsors and evaluators use evaluation infrastructures as a tool to determine which potential contractors are most likely to score well at the end of the project and should therefore be supported, and whether project performance during and at the end is in line with the stated objectives. By setting up evaluation infrastructures, sponsors can determine the ‘rigour’ of the evaluation and thereby calibrate the ratio of successful and unsuccessful projects. In addition, by regulating contractors’ awareness of specific evaluation methods, sponsors influence contractors’ ability to tailor their project performance to their evaluation. Both of these strategies create a power advantage for sponsors over contractors in project-organised work.

The second common feature of the organisation of failure is the intersubjective involvement of sponsors. The exception is platform work, where evaluation is algorithmic and hence impersonal. In general, the more personally connected sponsors are with contractors in projects, the more difficult it is for sponsors to organise failure, and this is reflected in the ratio of successful to failed projects. This finding is complementary to the theory of permanently failing organisations, where failure is prevented by differential consequences of failure for owners and dependent actors in the firm (Meyer and Zucker, 1989). The third finding relates to the phenomenon of denial of failure, originally described by Fisher (1972), whereby platform evaluations of gigs and sponsor evaluations of non-profit projects are both biased towards success, yet from the perspective of the contractors, their work in the project may be substantially unsuccessful.

The overall aim of this article is to turn attention to project sponsorship, that is, the activity of setting goals and providing resources, which, with few exceptions, has been neglected in the field. In the project-organised world of work, project sponsorship is a site of enormous concentration of power because it determines who does what work and how successful they will be in doing so. More attention should therefore be paid to how sponsors define project goals and whether they can be achieved with the resources provided. It would also be valuable to find out whether there is an entity with a legitimate mandate to assess the adequacy of the resources provided to achieve the stated objectives. Clearly, if the resources provided by the sponsor are barely sufficient to achieve the objectives, the likelihood of ‘deserved’ project failure is high, regardless of how good the contractors’ work is.

For practitioners, the perspective developed in this article is useful in that it sheds new light on the importance of evaluation infrastructures for the production of legitimate failure. The fact that they are presented by evaluators as the embodiment of standards in a particular field of practice (Star, 1999: 381), and therefore unworthy of serious attention by ordinary users, makes them an effective tool for fulfilling the theory of failure embraced by the sponsors (Fisher, 1972). Contractors who rely on vaguely stated goals, the achievement of which is assessed by elaborate methodologies, are often left with the impression that they are unable to predict the success of their efforts. The solution is not to make evaluation infrastructures more sophisticated or public, as transparency alone will not solve the problem (Harvey et al., 2012), but to require sponsors to justify the appropriateness of the infrastructures used to assess the achievement of project objectives. Sponsors should also be more explicit about what they consider to be desirable success rates for projects – both accepted and completed – which will help potential contractors in deciding whether to engage in project work with sponsors and what expectations they should have of the outcome of project evaluation.