With this chapter, the initiative journey reaches its most demanding stage: moving a solution that succeeded in a pilot into broad, real-world use. We examine the gap between adopting an innovation and implementing it, the factors that make some innovations spread faster than others, and the concerns of those who are asked to change, before turning to scaling readiness and the transfer of ownership.
The center for teaching excellence at a university developed a model for innovation project assignments: students identify a real problem in their field and test a solution to it with actual beneficiaries. The model was piloted in three courses taught by enthusiastic faculty members, and the results were striking: serious projects, high engagement, and students who took their projects on to the business accelerator. Once the results were presented to the university council, a circular was issued requiring the model to be applied in sixty courses the following semester. A year later, the model was “implemented” in every course on paper: most faculty members had simply renamed their old assignments to match it, and what students actually did had not changed.
The model succeeded when it was adopted by those who chose it, and it faltered when it was imposed on people who had never been consulted and were given neither the time nor the support to understand it. This story recurs across organizations in many forms, and it is our entry point into understanding why the move from limited success to wide use is the hardest stage of the journey.
After studying this chapter, the reader should be able to:
The implementation literature distinguishes between the adoption of an innovation, that is, the organization’s decision to use it, and its implementation, that is, its actual and consistent use by the intended people. Klein and Sorra showed that many innovations are adopted and then never implemented, or are implemented only superficially. In their view, implementation effectiveness depends on two things: the implementation climate, meaning the extent to which people perceive that using the innovation is expected, supported and rewarded; and the innovation’s fit with their values, meaning the extent to which they see it as consistent with what they believe in their work (Klein & Sorra, 1996). When implementation is expected and required but fit is weak, people comply on the surface without being convinced, which is exactly what happened in the opening case.
Klein and Knight observe that many innovations fail not because they are ineffective, but because they were never implemented with enough quality and consistency for their effectiveness to show (Klein & Knight, 2005). An organization may therefore pronounce a good innovation a failure when, in truth, it never implemented it at all. This is why scaling demands management as careful as that devoted to development.
Chapter 1 presented Rogers’s definition of innovation. Among his most important contributions is the identification of five characteristics of an innovation, as perceived by adopters, that explain much of the variation in how quickly innovations are adopted (Rogers, 2003). Table 15.1 sets them out, together with the question each one poses for the scaling team.
| Characteristic | Meaning | Design question |
|---|---|---|
| Relative advantage | The degree to which people perceive the innovation as better than what they currently use | What does the user personally gain, beyond what the organization gains? |
| Compatibility | Its consistency with their values, experience, needs and ways of working | What must the user give up in order to adopt it? |
| Complexity | How difficult it is to understand and use | What is the simplest form in which it can start? |
| Trialability | Whether it can be tried on a limited scale before commitment | Can users try it without risking much? |
| Observability | How visible its results are to others | How will others see its results among their colleagues? |
These characteristics underline that adoption turns on adopters’ perceptions of an innovation’s attributes. From the center’s perspective, the model in the opening case was undoubtedly better; to many faculty members, however, it looked like a complicated additional burden that did not fit courses designed around transmission and testing, and they were given no chance to try it before committing to it. Greenhalgh and colleagues’ review extended these characteristics to the context of service organizations, adding factors such as organizational readiness for change and adopters’ ability to adapt the innovation to their own context (Greenhalgh et al., 2004).
Where Chapter 7 dealt with the needs of the innovator, this section turns to the needs of those asked to adopt someone else’s innovation. Adoption is a change in a person’s work, and it raises legitimate questions and concerns. Working in the context of implementing educational innovations, Hall and Hord developed a model of the stages of concern that people pass through when asked to adopt an innovation. They showed that these concerns tend to shift from one level to another, that they may overlap within the same person, and that effective support responds to the stage the person is in (Hall & Hord, 2015). Figure 15.1 presents these stages in grouped form.
What is this? What will change for me? Can I do it?Clear information, dialogue about the effect on work and time, and honest reassurance
How do I organize my time? How do I deal with this practical problem?Hands-on training, ready-made tools, and close support during implementation
Does this benefit the beneficiaries? How can I improve it?Data on results, exchange of experience with colleagues, and room for development
A common mistake is to address people with impact questions while they are still preoccupied with self questions: presenting the innovation’s results and its benefits to the organization while they are asking how it will affect their time, their work and how others see them. In the opening case, faculty members were shown the results of the three courses, but no one asked them about the extra time that supervising real student projects would require, or about the support they would need. Their superficial compliance was therefore less a rejection of the innovation than a response to concerns that no one had answered.
This is consistent with what Chapter 8 said about culture: an administrative circular operates at the level of espoused values, while the faculty member’s basic assumption remains that the annual evaluation and promotion do not count this effort. Changing that assumption is beyond the reach of a circular; it requires culture-embedding mechanisms.
Before rolling out an innovation to your colleagues, sit down with five of them who did not take part in the pilot and ask: What worries you most about it? Their answers will tell you which stage of concern they are in, and what support they need before any presentation of results.
Chapter 14 cautioned that a pilot run under ideal conditions says nothing about ordinary conditions. List calls this phenomenon the voltage effect: many programs that achieve strong results in their early trials lose much of their impact when they scale. The causes are largely predictable. Among them: the initial result was a false positive; those who took part in the pilot do not represent those who will later use the solution; pilot conditions differ from operating conditions; and costs rise with scale instead of falling (List, 2022).
Most of these causes converged in the opening case: the faculty members teaching the three courses were enthusiasts who had chosen the pilot, whereas those teaching the sixty courses had not; and the three courses enjoyed the center’s direct support, which the sixty did not. This is why readiness questions should be asked before scaling: Has the solution been tested under conditions that represent reality? Does it work with people who did not choose it? What does it actually cost without exceptional support? Who will own and operate it?
One of the most important things to settle before scaling is the distinction between a solution’s core and its periphery. Greenhalgh and colleagues’ review noted that innovations with a fixed, well-defined core and a flexible, adaptable periphery spread more easily than innovations that are either entirely rigid or entirely flexible (Greenhalgh et al., 2004). The core is what must remain unchanged for the value to survive; the periphery is what each unit can adapt to its own context. Figure 15.2 illustrates this distinction.
Had the core been defined this clearly in the opening case, each faculty member could have adapted the model to their course without losing its value, and genuine implementation could have been told apart from superficial implementation with a single question: Has the core survived?
It is common in organizations for an innovation team to develop a solution and then “hand it over” to the unit that will operate it, only for that unit not to adopt it. The opening case of Chapter 13 offered one version of this. The reason is that ownership is built during the journey; it is not handed over at the end. Building it requires:
This is why the “operations voice” is one of the five roles on an innovation team, and why “Who will operate it?” is among the questions at the scaling gate.
Scaling takes more forms than a single leap from one unit to the entire organization. Three pathways can be distinguished, differing in speed and risk:
At the university, we propose that embedding innovation in courses begin with a limited number of interdisciplinary courses and then scale through pull and phasing together: the model is offered to interested faculty members with support from the center for teaching excellence, their students’ projects are showcased at the annual innovation day, and supervising those projects counts toward the annual evaluation. When other faculty members see the results among their colleagues and find support and recognition, they adopt the model out of conviction. We return to this design in Chapter 20.
Administrative mandates produce compliance; conviction produces adoption. The quickest route to superficial compliance is to impose an innovation on people who were never consulted and never given what they needed to understand it.
Answer the following questions before any decision to scale, and give each dimension a rating: ready, partly ready, or not ready. A dimension that is not ready identifies what must be done before or during scaling, without necessarily requiring that scaling be halted.
| Dimension | Check question | Rating |
|---|---|---|
| Evidence | Has the solution been tested under conditions that represent reality, and with people who did not choose it? | |
| Core and periphery | Has it been defined what must remain fixed and what can be adapted? | |
| Perceived characteristics | Do adopters see its advantage, compatibility and simplicity? And can they try it out? | |
| Adopters’ concerns | Do we know their current concerns? And is support designed around them? | |
| Implementation climate | Is use of the solution expected, supported and recognized in the formal systems? | |
| Cost | What is the actual operating cost without exceptional support? And who bears it? | |
| Ownership | Who owns the solution? Did they take part in developing it? Has it been embedded in their budget and procedures? | |
| Pathway | Has the appropriate scaling pathway been chosen: phased, pull-based or comprehensive? | |
| Follow-up | How will we know that implementation is genuine rather than superficial? |