The "interaction pattern" refers to the basic configuration of how, when, and with whom SMEs interact with business support and knowledge exchange programmes. Done et al. (2011) posit that best practice interventions (BPI) typically have a mixture of preparation activities, workshops, and follow-up consultations. This approach was adopted by McGovern et al. (2017) and Jones et al. (2024) in their analysis of the diffusion of process improvement practices. Farrukh et al. (2019) identify the resource implications of these interaction patterns and stress the need for balance between extensive, lower-cost interactions, such as 1-2-many1 workshops versus the number of intensive of 1-2-12 consultations with individual firms. However, many studies of programme evaluations show that deeper interaction rather than broader intensity is much more effective (Bennett & Robson, 2004; Mole et al., 2009, 2011, 2017; Mole & Keogh, 2009; Rakićević et al., 2016; Suci et al., 2022). Cunningham and McGuire (2019) found that 1-2-1 interactions are more effective as the support professionals need to “enter into the world of the SME” (Devins & Gold, 2002, p. 23). Giebel (2024) suggests that U shaped interaction pattern is sometimes required where intensity pattern is high to low and high again, which is different than a more straight-line adjustment recommended by Jones et al. (2024) who suggest an intervention gradient is required, moving the intervention to more involvement and ownership from the client toward the end of the intervention window.

Diagnostic tools are widely used in business support programs (Rakićević et al., 2016), particularly in the SME context (Burgess & Wake, 2012; Sepulveda et al., 2010; St‐Pierre & Delisle, 2006). Diagnostic tools (Jones et al., 2024; Łobacz et al., 2016) serve as a template to help both the external consultant and the SME pinpoint areas for improvement and allow the business to be more time efficient in engaging with the business support programme (McGovern et al., 2017; Mole et al., 2011; Salazar-Elena & Guimón, 2019). Diagnostics also help raise awareness of strategic choices (Mazzarol et al., 2014) and provide an affordance for capability assessments (Kumar & Harris, 2021). Such tools also offer SMEs greater agency in shaping their future development trajectory (OECD, 2020). However, complex and extensive diagnostics are unsuitable for SMEs (Burgess & Wake, 2012; Foggin et al., 2004; Kumar & Harris, 2021; OECD, 2020), posing a problem in developing a diagnostic extensive enough to achieve its objectives, but which SMEs find accessible and actionable.

Many studies mention the role of a "change champion" within the SME—someone who acts as an intermediary and leads by example. However, Jones et al. (2020) found that SMEs often struggle to identify suitable champions with the right qualities. Additionally, SMEs tend to use employees in multiple roles (McAdam et al., 2007), which makes it harder to identify and develop specific experts. Nevertheless, Done et al. (2011) stressed that broader employee involvement is important and that middle management is often key to longer-term success. McAdam et al. (2007) and Done et al. (2011) found that employee involvement in SMEs is important for developing management practices, but that depth of engagement (i.e. the number and scope of employees involved in the support intervention) is often too low (Done et al., 2011). Jones et al. (2024) found that depth of engagement was one of the most critical factors in the success of business support interventions.

Developing ‘best’ management practices in SMEs requires both external intervention and internal action by the SME itself to make the changes permanent (Done et al., 2011; Jones et al., 2024). Essentially, this means that for the SME to develop different capabilities some form of individual learning (Kurdve et al., 2020) and organisational learning (Smith, 2009) needs to take place. These two learning processes are usually interlinked (McAdam et al., 2007). The resulting ‘socialisation’ (Kurdve et al., 2020) of these practices is termed ‘internalisation’ (Jones et al., 2024). Smith (2009) shows that this internalisation requires both action and reflection on the part of the SME participants in the business support programmes. Kurdve et al. (2020) and Done et al. (2011) also argue that both these learning processes, and resulting internalisation require a degree of increasing independence from the support programme. Hence, this component of business support should contain both aspects and this theme is best classified as ‘independent internalisation’ of the relevant knowledge and management practice.

Many effective support programs use expert consultants (Done et al., 2011; Jones et al., 2024; Klonowski, 2010; Łobacz et al., 2016; McGovern et al., 2017), who also act as coaches (Azizi et al., 2023; Farrukh et al., 2019; Mole, 2023; Vidal‐Salazar et al., 2012). Many evaluations show that the use of experts bolsters trust (Ardley et al., 2016; Łobacz et al., 2016; Ram & Trehan, 2010; Ward, 2016). Moreover, these experts provide highly codified knowledge (Vasconcelos et al., 2019), which is easier to transfer to SME clients (Oberschachtsiek & Scioch, 2015). However, Gray et al. (2011) identify coaching as effective for SME leaders in self-determination, unless the coaching is designed and directed to specific business practices, where it is less effective at producing business outcomes.

Concept translation (Galbraith et al., 2017) is the customisation (Cunningham & McGuire, 2019; Klonowski, 2010), adaptation and reconfiguration of ‘standard’ best practices (Łobacz et al., 2016), also referred to as ‘templates’ (McGovern et al., 2017), to the local business context. This is critical to the successful diffusion and development of practices (Done et al., 2011; Jones et al., 2024; McGovern et al., 2017). Key extensive studies, such as Henley (2024) and Bennett (2008), as well as intensive studies (Ardley et al., 2016; Jones et al., 2024), found that this ‘translation’ of concepts to each unique SME context is critical to the success of the intervention support.

R₁) Concept translation requires that there is some degree of high intensity within the interaction pattern and that multiple follow-up interactions are required (Jones et al., 2024; Kurdve et al., 2020). Concept translation has a tension between the fidelity (McGovern et al., 2017) of utilising a successful template or concept, and the high-intensity interaction required for tailoring the adoption of individual SMEs. Although the need for concept translation should drive interaction patterns, if the pattern is low in intensity, then less translation will occur, and therefore classed as a two-way relationship.

R₂) Kurdve et al. (2020) and Redien-Collot and Lefebvre (2015) found that frequent interactions with SMEs lead to more meaningful dialogue, while Farrukh et al. (2019) noted that firms need time between interactions to reflect. Jones et al. (2024) utilised these aspects and found that meaningful dialogue and reflection within the interaction pattern design led to independent internalisation. Similar to R₁, the goal of independent internalisation should drive the interaction pattern, but if there are no frequent interactions with reflective gaps, then less internalisation is more likely.

R₃) High-intensity one-on-one consultations with coaches are needed as this provides an opportunity for the individuals in the firm to understand the concepts (Kurdve et al., 2020) and for coaches to understand the context (Clarke et al., 2006). As they move forward, a gap-structured, high-intensity pattern helps coaches support firms reflect between events and interventions (Jones et al., 2024). This is a two-way relationship; coaching requires this type of interaction pattern to be effective; but high intensity interactions without coaching will be less effective.

R₄) Intensive support to an SME can often be provided at no extra cost requirement to many employees/teams within each SME [Logical inference], although it requires more commitment and resource from the SME (Jones et al., 2024; Tamošiūnas & Lukošius, 2009). Low-intensity 1-2 patterns, such as workshops (de Wit-de Vries et al., 2019), often limit the capacity for firms to send multiple employees. There is, therefore, a two-way relationship between interaction patterns and depth of engagement

R₅) The use of a diagnostic can help frame the interaction pattern and lead to efficient and relevant paths for expert advice (Mole et al., 2011). A diagnostic can help focus on where a mixture of intensity is part of the constraints of the programme and is critical for brokered types of intervention design (Henley, 2024). This relationship is likely to be mostly one-way – the interaction pattern does not affect the diagnostic deployment, although many diagnostics are likely to need to have a degree of 1-2-1 high intensity for them to be effective (r₅). Still, this can be overcome with self-assessment diagnostics (OECD, 2020).

R₆) The Diagnostic helps initiate a dialogue to facilitate the start of concept translation (Jones et al., 2024), and a diagnostic approach is an individually high-intensity means of customising the concepts and thus, aids concept translation (Mole et al., 2011). The need for concept translation could, in some circumstances, affect the deployment or design of a diagnostic (r₆).

r₇) Jones et al. (2024) also found that the diagnostic dialogue around the application of concepts helps lead to independent internalisation (organisational learning). The need for independent internalisation in some programme contexts could, in some circumstances, affect the deployment or design of a diagnostic.

r₈) Although it is the case that diagnostics are often completed within a singular setting with SME owner-managers (Mazzarol et al., 2009), diagnostics are more effective when deployed with multiple respondents and within teams (Jones et al., 2024). Moreover, the event of using the diagnostic could bolster the depth of engagement (Jones et al., 2020). The complexity of the diagnostics would be potentially affected by the need to deploy it in a group setting.

R₉) These expert coaches can help the translation process by customising best practices to suit the specific needs of the SME (Azizi et al., 2023; Farrukh et al., 2019; Mole, 2023; Oberschachtsiek & Scioch, 2015; Vidal‐Salazar et al., 2012) and manage the tension between concept fidelity and customisation (McGovern et al., 2017). Concept translation usually requires some form of expert coaching (Oberschachtsiek & Scioch, 2015).

r₁₀) Feedback from expert coaches also facilitated independent internalisation (Jones et al., 2024; Redien-Collot & Lefebvre, 2015). The need for specific independent learning could affect the deployment and focus on the expert coach, and the type of expertise required.

r₁₁) Expert coaches can assist in the deployment of the diagnostic, so it is done thoroughly with respect translating the relevant concepts and help the coaches understand the business and build trust with the clients (Mole et al., 2011; OECD, 2020). The diagnostic could require some degree of knowledge, but not necessarily expertise in the relevant area, or competence in coaching, but the level of knowledge of the support professional could constrain the diagnostic design.

r₁₂) The more employees engaged in the interventions, the more likely that internalisation will take place, from both a probability perspective and peer learning networks /pressure (Turianskyi et al., 2018). The socialisation of internalisation could increase the depth of engagement, as participants learn they wish to share this knowledge with others (Vasconcelos et al., 2019).

r₁₃) Depth of engagement might be assisted by coaching, if this were designated team coaching (Hackman & Wageman, 2005; Jones et al., 2024). Conversely, if depth of engagement was important, then this would affect the designated skill set of the coaches on the programme (Rousseau et al., 2013).

r₁₄) Depth of engagement could affect concept translation due to the socialisation involved in internalisation (Kurdve et al., 2020), whilst difficulties and requirements to translate certain concepts could drive a deeper and wider engagement with the support interventions (McAdam et al., 2007).

The analysis in section 3.1.2 reveals that all core components are linked in some way to all the other components, albeit with different strengths and different levels of evidence. In some cases, there is logical inference in some of the direction of the relationship rather than direct evidence. We can therefore consider these core components as a nexus arrangement (Jones et al., 2020), where there is an expectation that each component will have some form of interaction, effect or dependency with another component. There is not enough evidence to discriminate between R and r strength of relationships, given the lack of research into business support design components, as identified in the introduction. The resulting nexus is illustrated in Figure 2 below.

Several studies outline the importance of measuring and evaluating programme outcomes (Henley, 2024; Rakićević et al., 2016) and how this is gathered, not just what is measured (Singerman & Voytek, 2023). Henley (2024) and Singerman and Voytek (2023) both outline the importance of tangible firm financial performance data as well as satisfaction and self-assessment.

The Search and Selection element is a bi-directional process; firms seek knowledge transfer and collaboration for competitive advantage (Bishop et al., 2011), but SMEs struggle with finding the right collaboration for the relevant purpose (Kurdve et al., 2020). Many evaluations of business support programmes struggle with the separation of identifying the bias of self-selection of businesses (Bennett, 2008; Henley, 2024), which has led to the emphasis on Random Controlled Trials (RCT) for government-funded interventions and evaluations. Balog et al. (2022) found that overall effectiveness was reduced once a self-selection bias was removed. Henley (2024) argues that a selection process is required to identify the firms that would best benefit from business support, and Azizi et al. (2023) and Jones et al. (2024) show that the selection process is critical to retention and completion on business support programs. Arshed et al. (2021) identify an informal and formal process of capability assessment by support providers, which is beneficial to the overall support outcomes.

Previous experience of collaboration and greater levels of trust help lower collaboration barriers (Bruneel et al., 2010 cited in Knockaert et al., 2013) and Bennett (2008) identifies that a reliable partnership between funders and delivery providers is critical to success, as is a more stable policy and program landscape, particularly for SME’s accessing and utilising the relevant business programmes. Mole et al. (2014) show that even in the same programme across regions, the choices made by agents in the partnership affected the programme outcomes both positively and negatively.

Rakićević et al. (2016) in their best practice model of support programs show that a structured mobilisation design for implementation and mobilisation is key to the success of a business support program. Mole et al. (2017) showed that choices and project management capability around implementation are key differentials in program outcomes.

Henley (2024) highlights the high attrition rates on business support programmes, and Bennett (2008) and Jones et al. (2024) both identified the significance of a client management process once the SME has enrolled on the programme. This is to ensure retention and progression through the programme, particularly where the SME might find some aspects challenging and they need to devote resources or attempt new practices (Jones et al., 2024) and the development of trusted relationships between the providers and the client can directly assist the knowledge transfer process (Balog et al., 2022; Henley, 2024; Mole et al., 2017).