Abstract no: 002-0179

"Innovation Stickiness: Investigating a Complex Supply Network"

Second World Conference on POM and 15th Annual POM Conference,

Cancun, Mexico, April 30 - May 3, 2004

Abstract

This paper presents research carried out with a major UK utility provider whose supply network performance indicators included a specific focus upon innovation. The work presented here investigatedthe different innovations generated and how successfully they had been shared within the supply network in general.

Data logging the innovations generated was reviewed. Following this, a questionnaire was developed to explore the relative ease of sharing, or “stickiness”, of different innovation types, depending upon contextual elements of the innovation and the participants in the sharing activity.

1 Introduction

Innovation can be a difficult concept to grasp. The literature has many perspectives, from the ‘softer’ ideas around enabling conditions and communication, to the harder ‘output’ oriented topics of measured impacts and types. In this context a number of these perspectives are considered. First however, it is worth quoting a definition of innovation, taken from Rogers (1995). An innovation:

“..is an idea, practice or object that is perceived as new by an individual or other unit ofadoption”

i.e. the innovation does not have to be a ‘new’ idea, but merely an idea applied in a new context (Schroeder et al., 2000). This is important when considering individuals’ understanding of what ‘counts’ as an innovation.

The early innovation literature assumed a technological view, which has now expanded to include a variety of process-type innovations (Slappendel, 1996). Abernathy and Utterback (1978) have identified three main innovation types. These may be taken as product innovation (such as Dyson’s vacuum cleaner or Bayliss’ clockwork radio), process innovation i.e. the means of making or delivering the product (be that a physical good or service), and organizational innovation i.e. a wider business process or organizational philosophy perspective. Trott (1998) suggests many more.

Much work has been carried out on the idea of innovation diffusion, whereby innovations are driven by a particular need, researched and applied, communicated, shared and adopted in new circumstances (Rogers, 1995). A barrier to diffusion is the issue of ‘stickiness’ of knowledge transfer (as part of the innovation sharing process). Von Hippel (1994) describes knowledge stickiness as the incremental cost of transforming a given unit of information in to a form useable to the recipient. Suzulanski (2003) suggests that stickiness is a function of multiple factors, specifically:

  1. The knowledge transferred, including causal ambiguity (lessening the propensity to learn from a source)
  2. The source of knowledge, including the source’s motivation (or lack of) and perceived reliability of that source
  3. The recipient of knowledge, including the recipient’s motivation (or lack of) to accept new knowledge
  4. The context, such as a barren organizational context or arduous relationship, both of which inhibit communication and transfer.

Connell et al. (2003) propose that knowledge transfer is eased when two organizations knowledge processing systems and organizational cultures are similar. The literature on supply network relationships also clearly indicates trust, reciprocity and strength of relationship as being prerequisites for successful collaboration (Baker, 1991), (Dyer and Nobeoka, 2000).

Diffusion in its fullest sense may be relatively rare, especially when the innovation adoption is not driven by need. However, activities that potentially help sharing take place, such as the nature of innovations, pre-existing relationships and so on, all play a part. This research considers the nature of innovations.

2 Background

The research focuses upon a UK utility provider (UP) and its first-tier supply network. The UP operated a five year programme of work for which suppliers were employed to upgrade and maintain assets across a region of the UK, in line with Industry Regulator requirements and Environmental Legislation. The supply network structure was complex comprising four key supplier entities, each entity being a partnership or alliance between at least two companies (see Figure 1).

Each of the four suppliers was required to innovate with the specific objective of reducing costs in the project solutions they designed and executed. In addition, the suppliers were required to share those innovations with each other and the UP. To encourage this, measurement of the innovation activity was formally adopted by the UP as a Key Performance Indicator (KPI) within a broad framework of measures modelled on a balanced scorecard approach (see Kaplan and Norton, 1992). Each supplier submitted innovations on a quarterly basis, which were then assessed by the UP for suitability and relevance and comprised part of an overall performance rating.

The research presented here builds upon previous work which investigated the practice of knowledge and information sharing within the UP’s supply network structure. The previous work focused upon a number of features of the sharing initiative, including affects upon the suppliers’ competitive position, behaviours associated with performance measurement and factors facilitating and inhibiting sharing(see Shaw and Peel, 2003). Initial findings had indicated that suppliers were very willing to share innovations, in part because this requirement had been clearly articulated within the contract although other constraints were observable (in line with Dyer and Nobeoka's work (2000)). It was noted however, that the performance measure was shaping behaviour whereby suppliers were encouraged to submit as many innovations as possible; consequently some innovations could be perceived more as ‘common sense’. In addition, the nature of the different suppliers’ organizational contexts, together with the different innovation types, presented real difficulties for sharing particularly where adoption was not driven by need.

This earlier work has now been extended by introducing the concept of ‘stickiness’ (Von Hippel, 1994). It was expected that as well as an indicator of preferred innovation types (i.e. considering volume) an increased understanding of those types would also be gained. The“stickier” areas of innovation sharing would indicate the context in which knowledge and innovation sharing would requiregreater support from all the organizations involved, particularly the UP, in future work programmes.

The research had two main phases, the first exploring the innovation frequencies of the different innovation types and the second exploring the concept of stickiness.

3 Methodology

The research focused upon how the ‘stickiness’ associated withinformation sharing might vary across innovation types.

Access had been granted to an innovation log which contained details of all the innovations submitted to the UP by each supplier, each quarter. The log contained approximately one thousand innovations. Details held included the innovation owner (an individual), whether the innovation had resulted in a cost saving and a categorisation. The innovation categories had been identified by the UP and were a mixture of discipline areas and perceived likely innovation types. Whilst these categories did not map particularly well against those indicated in the literature, it was decided that they be maintained during data gathering in order to remain meaningful to the interviewees and respondents. Future work will map the UP’s innovation categories to those found in the literature.

The first phase of research attempted to identify simple patterns in the innovation data regarding frequencies of type, per supplier, across the programme of work. Graphs were generated to investigate the propositions that innovations would diminish over time and whether certain innovation types, such as “procedural”, would be fewer in number than, say, “technical”.

During the second phase of research a questionnaire was designed to investigate the sharing of a variety of innovations. This phase of work could not focus exclusively upon diffusion, as no formal record had been kept of innovations that had been adopted & / adapted by the suppliers. Attention was therefore placed on the characteristics of how different innovation types had been shared directly with another supplier. Again, this was investigated for the duration of the programme of work to date.

The questionnaire was adapted from Suzulanski's (2003) work investigating the sharing and implementation of best practice initiatives within large organisations. The questionnaire was applied both from the perspective of innovations shared and also innovations received. The former work applied to approximately fifty innovation cases per supplier (five innovation types across ten quarters). The latter was applied to innovations selected by interviewees, where adoption had taken place; this was the only source of information regarding innovations originating from an external source. As far as possible the adopted innovations were crosschecked with the innovation source.

4 Results and Issues

4.1 Phase 1

Graphs were constructed for each supplier and each innovation type across the programme of work. Surprisingly different patterns were obtained to those expected: no trends of declining rates of innovation generation were observed across suppliers and no single innovation type appearedparticularly dominant. Unanticipated observations included:

Volumes of Innovations logged per supplier:

  • sudden peaks occurred across all innovation types (see Figure 2), indicating the likelihood that a supplier had placed significant effort on innovation generation generally during one particularquarter
  • those peaks did not coincide across suppliers but often appeared consecutive. Subsequent informal discussions with suppliers indicated that a previously poor performance rating or particularly high rating of another supplier, encouraged more enthusiastic innovation activity during specific periods of time
  • initially the innovation log had not been circulated between suppliers and an observable peak occurred in all suppliers’ innovations post-circulation ( ~Q2 / 3 2002-3).

Supplier logs per innovation type:

  • an immediate decline in the volume of innovations was observable between the first and second quarters’ (2001-2), across all suppliers and innovation types. Beyond this one could predict that the rate of innovations would decline, particularly towards more recent reporting periods, as the majority of need-driven innovations would already have been identified. Interestingly this was not the case, with rates of innovation generation undulating with no apparent evidence of tailing off
  • of all the innovation types, ‘technical’ was the highest volume category, closely followed by ‘procedural’. The lowest volume categories were ‘safety’ and ‘environmental’. The high level of ‘procedural’ innovations is most likely indicative of the immediate difficulties multiple organizations would have faced in entering into a new partnership.

Whilst the volumes and frequencies of innovation generation and logging had presented somewhat unexpected results, the activity of sharing innovations was thought to be more predictable. This was explored in phase two of the research.

4.2 Phase 2

The questionnaire was constructed in line with Suzulanski's (2003) constructs of ‘ambiguity’, ‘motivation of source’, ‘motivation of recipient’ and ‘organisational context’ and in consultation with key supplier staff. The questionnaire was finalised into two versions: one with the interviewee supplier as innovation source and one as innovation recipient. The administration took place face to face. Working through the innovation log, innovations were selected by type and per quarter, to capture a temporal element of sharing in addition to innovation type.

Initial findings indicate that the contextual nature of the region of work (by demographic, geographic properties and so on) restricted the applicability of innovations to different suppliers. This relates most closely to ‘motivation of recipient'. Most ‘procedural’ innovations were also generated in very different contexts (i.e. organizational partnerships) and again were of limited use to other suppliers. Again, this relates most closely to ‘motivation of the recipient'. Regarding other innovation types, a strong sense of commonality around the legislatively-driven ‘safety’ and ‘environmental’ innovations reduced perceived stickiness overall.

5 Conclusion

The data will offer significantly greater insight and the analysis continues with a view to identifying more specific predictors of stickiness using techniques such as ANOVAs / regression. In addition, the significance of frequency of communication within and between suppliers, and regarding different innovation types, is expected to provide interesting results. The frequency of communication can be linked with both the ‘source motivation’, ‘recipient motivation’ and 'organizational context' constructs; this will be investigated using Social Network Analysis (see : for more information) as a new phase of work supporting innovation stickiness.

It is hoped that the results will contribute new theory linking knowledge stickiness and innovation typologies. It is also anticipated that the UP will gain considerable insight into how innovation should best be supported for the duration of their current programme of work, but also for the start-up of a new programme expected within the next two years.

Acknowledgements

Thanks go to all staff members at the UP and supplier organizations for their help and time.

References

ABERNATHY W. J. and UTTERBACK J., 1978, Patterns of Industrial Innovation. In Readings in the Management of Innovation, M. L. Tushman and W. L. Moore (HarperCollins).

BAKER W. E., 1991, The Network Organization in Theory and Practice. In Networks and Organizations: Structure, Form and Action, N. Nohria and R. G. Eccles (Harvard Business School Press).

CONNELL N., KLEIN J. H. and POWELL P., 2003, It's Tacit Knowledge But Not As We Know It: Redirecting the Search for Knowledge. Journal of the Operational Research Society, 140-152.

DYER J. H. and NOBEOKA K., 2000, Creating and Managing a High-performance Knowledge-sharing Network: The Toyota Case. Strategic Management Journal, 345-367.

KAPLAN R. S. and NORTON D. P., 1992, The Balanced Scorecard - Measures that Drive Performance. Harvard Business Review, 71-79.

ROGERS E., 1995, Diffusion of Innovations (Free Press).

SCHROEDER R. G., VAN DE VEN A. H., SCUDDER G. and POLLEY D., 2000, The Development of Innovation Ideas. In Research on the Management of Innovations: The Minnesota Studies, A. H. Van de Ven, H. L. Angle and M. S. Poole (University Oxford Press).

SHAW N. and PEEL A., 2003, "Review of a Knowledge and Innovation Sharing Initiative: The Case of a UK Utility Provider". Decision Sciences Institute 7th International Conference, Shanghai, China.

SLAPPENDEL C., 1996, Perspectives on Innovation in Organizations. Organization Studies, 107-129.

SUZULANSKI G., 2003, Sticky Knowledge: Barriers to Knowing the Firm (Sage Publications).

TROTT P., 1998, Innovation Management and New Product Development (Financial Times publishing).

VON HIPPEL E., 1994, Sticky Information and the Locus of Problem-Solving. Management Science, 429-439.

1