The first post in this series described how strategic abstraction becomes delivery avoidance. The second examined the practical and epistemological problems of starting - how designers defer entry into unfamiliar domains, and how invocations of "real needs" can function as a substitute for engagement with the actual. Both posts described patterns of behaviour. This post asks whether the field's own self-understanding - the ontological vagueness that allows "service designer" to function as an abstract signifier onto which incompatible expectations are projected - might structurally enable them.
What does service design design?
This sounds like a settled question. It isn't. Four decades of theoretical effort - from marketing (Shostack, 1977; 1982), through service management (Edvardsson and Olsson, 1996), to design research (Secomandi and Snelders, 2011; Blomkvist, 2015) - have produced a rich and irreconcilable set of answers. Secomandi and Snelders (2011) find some convergence: services exist within exchange relations, can be separated into interface and infrastructure, and the interface is material rather than intangible. But they do not specify a formal ontology of what the interface consists of. Subsequent work expanded the scope further - Vink and Koskela-Huotari (2021) extended the "design material" of service design to include institutional social structures; Suoheimo and Jones (2025) trace a progressive expansion from Shostack's tangible evidence to entire institutional arrangements. Each expansion is analytically powerful and practically paralysing.
The practical consequence is that the field lacks a common analytical vocabulary for specifying what has been designed. We can say we "improved the service" or "redesigned the touchpoints" - but we cannot specify, in a way another designer could independently verify, what entities the service acts upon, what events change their state, what conditions must hold for those events to occur, or what states the entities can be in. We can describe services in natural language. We cannot specify them formally.
This matters for the patterns this series has described, and for consequences that extend well beyond individual avoidance. If the field's own theoretical apparatus doesn't specify what the object of design is - if the most sophisticated position available is "an assemblage of socio-material arrangements" - then a designer who claims they can't start because the scope isn't clear, or the context isn't understood, or the "real needs" haven't been articulated, is making a claim that the field's own theory implicitly supports. If the material is everything, then insufficient understanding of anything becomes a legitimate reason for inaction.
The consequences compound. Bailey (2021), in their ethnography of design for government in the UK, found that the "set piece performances" of design workshops serve multiple functions - they deliver invigorating and reassuring experiences of creativity, enrol participating subjects into a cadre of people who believe they are engaged in innovation, and establish the sense of a coherent field of practice. But as I argued in Performance and Substance, performance is not the same as production. When the field cannot specify what production looks like, design theatre becomes indistinguishable from design practice - and organisations that recruit for "design capability" have no way to tell the difference. Alvesson and Spicer (2016) describe the organisational conditions under which this becomes self-sustaining: functional stupidity thrives in contexts where persuasion trumps substance, where critical thinking is discouraged because it threatens hard-earned certainties, and where the general positive vibe is more valued than the specific uncomfortable question. Graeber (2018) goes further: when an occupation cannot specify the relationship between its activities and its outcomes, the conditions exist for what he calls "bullshit jobs" - work that even the person performing it suspects may be pointless, but which persists because the organisational structures that sustain it are more robust than the feedback mechanisms that would test it.
This is not an abstract risk. It is the practical reality that design leaders in public services encounter when they try to recruit, retain, and develop designers on complex programmes. A field that differentiates through ethos rather than demonstrable expertise will attract, promote, and reward people who are good at performing the ethos - who talk holistically, who convene workshops, who invoke user needs with conviction - regardless of whether they produce concrete, evaluable design work. Andrews, Pritchett and Woolcock (2017), writing about institutional reform in development, describe the organisational version of this dynamic as the capability trap: a condition in which adopting new terminology, frameworks, and processes gives the appearance of capability without enacting any change in practice - "looks like" substitutes for "does". The capability trap is self-reinforcing, because the signals of compliance (the language, the workshops, the frameworks) satisfy the same institutional audiences that would need to detect the absence of production. The ontological vagueness of the field is not just intellectually unsatisfying. It is the structural condition that makes the capability trap sustainable and design theatre indistinguishable from design practice.
The ethos problem
But ontological vagueness is a symptom, not a root cause. The deeper issue is how the occupation constructs its mandate - and what that construction makes possible.
Fayard, Stigliani and Bechky (2016), in one of sharpest empirical studies of service design's occupational positioning I've encountered, show that many service designers differentiate themselves not primarily through technical expertise or domain-specific skill but through what they call an ethos: a set of values (holism, empathy, co-creation) enacted through material practices (design research, visualisation, prototyping). This ethos is the occupational mandate. It is how service designers construct their legitimacy vis-à-vis both traditional designers and management consultants; it is, in Hughes's (1958) terms, the internally shared understanding and externally perceived right to define proper conduct, values, beliefs, and ways of thinking.
The occupation emerged from a dissatisfaction with traditional design work - product, industrial, interaction, graphic - which many found too routinised and insufficiently strategic. Service designers wanted to move "upstream", to help organisations decide what to do rather than simply executing a brief. But this created a positioning problem: they occupied contested territory between traditional designers and management consultants, sharing design practices with the former and strategic aspirations with the latter. Their response was to construct a mandate grounded in values rather than skills. Abbott (1988) had argued that values play a decreasing role in legitimising occupations; Fayard et al. show that for nascent occupations where skills are not the main differentiator, values may be more important than ever.
Here is the problem, stated plainly: an occupation that differentiates through values rather than skills is an occupation whose mandate can be performed rather than practised.
This is not Fayard et al.'s conclusion. They treat the ethos as the genuine basis for the occupation's mandate, and their evidence supports this - when the ethos is enacted through material practice, it produces distinctive and valuable work. But their own analytical framework contains the seed of the dysfunction this series has described. If the mandate rests on holism, empathy, and co-creation rather than on domain expertise or technical skill, then the mandate can be credibly claimed by anyone who talks holistically, expresses empathy, and convenes co-creation workshops - regardless of whether they have produced anything concrete, engaged with the operational specifics of their domain, or delivered work that another practitioner could evaluate against explicit criteria.
The performed ethos and the enacted ethos
Fayard et al. make the crucial point that the ethos exists only as enacted - that values without material practices are abstract claims, and material practices without values are just tools anyone can learn. The distinction between performed and enacted is where the mandate lives or dies. But this distinction is, in practice, extraordinarily difficult to observe from the outside.
Consider the patterns described in the first and second posts. A designer who retreats from operational specifics to institutional critique - "everybody works in silos", "the basics aren't there", "people are unwilling to fix things" - is performing holism. They are taking a system view, identifying structural problems, diagnosing organisational dysfunction. It sounds like the ethos. It looks like someone applying the very values that Fayard et al. identify as constitutive of the occupation.
But it is holism without material practice. Empathy without specificity. System thinking without the entities. There are no artefacts. No maps. No specifications. No testable claims about what the service does. The ethos is being performed - credibly, perhaps even sincerely - but it is not being enacted through the material practices that would make it productive.
The same applies to the "real needs" deflection. Invoking user needs against organisational convenience is a form of empathy - the designer positions themselves as the voice of the user against the corrupting influence of internal politics. But if the designer hasn't done the user research, hasn't produced the artefacts, hasn't engaged with the operational specifics of the domain, then the empathy is abstract. It is the value claimed, not enacted. The ethos provides the rhetorical resources for the deflection precisely because the deflection sounds like the ethos.
This is the structural vulnerability the field has inadvertently constructed. By grounding the occupational mandate in values rather than in demonstrable domain expertise or formally specifiable outputs, service design has created conditions in which it is genuinely difficult to distinguish between a practitioner who is doing the work and a practitioner who is performing the stance. The tools for making that distinction - explicit ontological commitments, formally specifiable outputs, domain expertise that can be tested - are precisely the things the field has, for four decades, declined to develop.
Why this matters for leadership
Hay and Vink (2023) found that the "systemic turn" in service design research - the expansion from improving user experiences towards catalysing service ecosystem transformation - was "primarily conceptual". Practitioners they interviewed reported struggling not with conceptual scope but with navigating mandates, politics, and established professional territories. The expansion of what designers should think about happened without expanding what they can actually do. This is the same pattern at the occupational level: the field expanded its values (from empathy-for-users to holism-across-systems) without expanding the material practices through which those values could be enacted. The ethos grew. The practice didn't keep pace.
Hay and Vink (2024), in a companion study on why service designers struggle to address power dynamics, identify a revealing paradox: the designers they studied acknowledged that understanding professionals' emotional experiences was central to effective service design, yet they consistently struggled to address these dynamics in practice. The field's values included empathy for all stakeholders; the field's practices hadn't developed the tools to enact that empathy in contexts where power, politics, and professional territory were at stake.
Bailey and Drew (2017) make a related argument from practice. They describe the shift from skills to mindsets in public service design - the recognition that the contexts for design are changing faster than any fixed toolkit can accommodate, and that what designers need is not a better set of tools but a more adaptable orientation to complexity. This is right as far as it goes, but it also risks reinforcing the ethos-over-expertise pattern: if what matters is mindset rather than skill, then the mandate becomes even harder to evaluate from outside, and the gap between performance and practice becomes even wider.
The design leader's problem is this: how do you distinguish between a designer who is genuinely working through the difficulty of an unfamiliar domain - building domain knowledge, producing imperfect artefacts, iterating toward understanding - and a designer who is performing the ethos without enacting it? If the field's self-understanding privileges values over demonstrable outputs, the leader has no diagnostic criteria. Both designers can speak the same language. Both can claim holism, empathy, and co-creation. The difference is only visible in what they produce - and if the field has not developed the vocabulary for specifying what production looks like, the leader cannot name the gap.
What would change this
If the core difficulty is that ethos-over-expertise structurally enables performance-without-practice, then the remedy cannot be more ethos. It needs to operate at a different level.
The first move is making ontological commitments explicit. In the objects, events, and frames model I've been developing elsewhere in this series, I've argued for a structured discovery vocabulary: entities (objects with properties and states), events (state transitions with preconditions), and frames (compositional structures that relate entities and events into coherent wholes). This isn't the only possible vocabulary, but it is an explicit one - and explicitness is the point. A designer who can say "here is my context map, here are my traced events, here are the gaps, and here is what I need from you to fill them" is enacting the ethos through material practice. A designer who says "the basics aren't there" is performing the ethos without producing anything evaluable.
The second move is distinguishing levels of analysis. As Wastell (1999) argues, the anxiety of producing something imperfect is real - particularly for senior practitioners whose professional credibility is at stake. The artefact, in Wastell's framework, functions as a transitional object: something external to the individual that can be wrong without the person being wrong, that creates a surface for conversation rather than a performance to be evaluated. A brief that distinguishes levels explicitly - your job is to enumerate the entities, trace the events, identify the preconditions; when you encounter institutional constraints, document them and escalate them - gives the designer permission to not-know at the operational level (discovery is the work) while preventing retreat to the institutional level (that's a different job).
The third move is cultural, and it is the hardest. The occupation's self-understanding needs to shift from an ethos that must be performed at all times to one that can be honestly under construction. Edmondson's (1999) work on psychological safety is relevant here, but not sufficient. The issue is not just whether the team tolerates not-knowing; it is whether the occupation's own identity permits it. A service designer whose mandate rests on demonstrating holism, empathy, and co-creation needs to be able to say "I don't yet know this domain well enough to demonstrate those values in practice, and here is my plan for getting there". That sentence requires a shift from ethos-as-credential to ethos-as-method - from something you have to something you do, iteratively, in the specific context you're working in.
None of this will be resolved by better theory. It will be resolved, if at all, by design leaders who can name the distinction between performed and enacted ethos, who can create the conditions in which imperfect artefacts are safer than abstract critique, and who can hold their teams to a standard of production rather than a standard of values. The ethos is not the problem. Ethos without production is the problem. And the field's reluctance to specify what production looks like - to make explicit ontological commitments, to develop formally evaluable outputs, to ground the mandate in demonstrable expertise as well as values - is what keeps the gap open.
References
Abbott, A. (1988) The System of Professions: An Essay on the Division of Expert Labor. Chicago: University of Chicago Press.
Alvesson, M. and Spicer, A. (2016) The Stupidity Paradox: The Power and Pitfalls of Functional Stupidity at Work. London: Profile Books.
Andrews, M., Pritchett, L. and Woolcock, M. (2017) Building State Capability: Evidence, Analysis, Action. Oxford: Oxford University Press.
Bailey, J.A. (2021) Governmentality and Power in 'Design for Government' in the UK, 2008-2017: An Ethnography of an Emerging Field. PhD thesis, Lancaster University.
Bailey, J. and Drew, C. (2017) From Skills to Mindsets: Grappling with Complex Public Service Challenges. Touchpoint, 9(2).
Blomkvist, J. (2015) Ways Of Seeing Service: Surrogates For A Design Material. Nordes 2015: Design Ecologies.
Edvardsson, B. and Olsson, J. (1996) Key Concepts for New Service Development. The Service Industries Journal, 16(2), 140-164.
Edmondson, A. (1999) Psychological Safety and Learning Behavior in Work Teams. Administrative Science Quarterly, 44(2), 350-383.
Fayard, A., Stigliani, I. and Bechky, B.A. (2016) How Nascent Occupations Construct a Mandate: The Case of Service Designers' Ethos. Administrative Science Quarterly, 62(2), 270-303.
Graeber, D. (2018) Bullshit Jobs: A Theory. London: Allen Lane.
Hay, A.F. and Vink, J. (2023) The Emotional Neglect in Recent Service Design Developments. Design Studies, 89.
Hay, A.F. and Vink, J. (2024) Why Do Service Designers Struggle to Address Power Dynamics? Design Studies.
Hughes, E.C. (1958) Men and Their Work. Glencoe, IL: Free Press.
Secomandi, F. and Snelders, D. (2011) The Object of Service Design. Design Issues, 27(3), 20-34.
Shostack, G.L. (1977) Breaking Free from Product Marketing. Journal of Marketing, 41(2), 73-80.
Shostack, G.L. (1982) How to Design a Service. European Journal of Marketing, 16(1), 49-63.
Suoheimo, M. and Jones, P.H. (2025) Systemic Service Design. Routledge.
Vink, J. and Koskela-Huotari, K. (2021) Social structures as service design materials. In Designing for Service: Key Issues and New Directions (2nd ed.). Bloomsbury.
Wastell, D.G. (1999) Learning Dysfunctions in Information Systems Development: Overcoming the Social Defenses with Transitional Objects. MIS Quarterly, 23(4), 581-600.