# Epilogue: No Failure if There Is Learning
*This guide has been written from experience: arriving at learning through action. Insights emerge through action when we push the boundaries of what we already know. Sometimes called failure, we’ve followed the maxim credited to Nelson Mandela “I never fail: I either win or I learn.” This epilogue shares where the line was sometimes blurry; key* [*paradoxes*](/undp-accelerator-labs/references/glossary.md#paradox) *when we’ve tried – and not always succeeded – to balance between two opposing directions.*
## External versus Internal (Focus)
We started this large Network of innovation labs aware of the pitfalls of isolated teams mandated to push boundaries but disconnected from core business and the wider organization. Our strategy was designed to have the Lab teams start with deep immersion into ongoing programs and core business, and assist teams to assess coherence of their work across silos and with the external environment. The R\&D was meant to derive from this analysis of coherence, rather than chasing novelty. We made a conscious effort to embed the Lab teams into the wider organization – they ceremonially swore an oath of service,[\[1\]](#footnote-0) followed standard organizational procedures, and were encouraged to make themselves useful to the core business teams even if the scope of work didn’t meet the goals of collective R\&D.
“Are the Labs integrated (into the core organization)?” became an ongoing refrain. Regardless of how much we tried to encourage embedding, the myth that the Lab teams were somehow separate persisted. On the other hand, Lab teams reported that the more integrated they became into the wider organization, the less time and space they had to deliver on their R\&D mandates. It became clear over time that without deliberate protection of an R\&D mandate, the range of experiments would become narrower.
The internal-external paradox emerged: is lasting internal organizational change triggered by evidence of a new world outside[\[2\]](#footnote-1) or based on evolving internal strategy? Should the results of R\&D be handed over to teams inside the organization or is uptake among partners an easier lift? If inside the organization is the handover target, to what extent and how should it be monetized in order to sustain the R\&D capability?
Given the scope of this lab Network (115 countries at its largest), there wasn’t one definitive answer. We arrived at thinking about this as a paradox between exploration and implementation.
## Exploration versus Implementation (Risk)
This is a core paradox that resonated with many audiences, even if a common verdict was never drawn. What became clear after many blurred lines was that neither extreme is ideal for units which are meant to develop new knowledge, create new [value](/undp-accelerator-labs/references/glossary.md#value) and find entry points for system transformation. Even where R\&D teams succeed in establishing their legitimacy as discoverers of new possibilities, they risk becoming detached from the core business if they only pursue academic interests (extreme exploration). And at the same time, R\&D teams that veer too far into implementation and only execute pre-defined agendas risk not bringing in the type of insights that create new value, and thus cease to be what they were designed for (extreme implementation).
The learning sweet spot was most often found through trial and error: pushing too far in one direction or another. It appeared where R\&D teams helped core business teams identify emerging opportunity and problem spaces, test solutions, tap into uncommon data sources, and work with new partners to apply innovations. Most often, and over time in the large organization the Accelerator Labs were part of, the pull was towards implementation more than exploration.
## Holding Space versus Handing Over (Revenue)
When we started this effort, our primary backer compared the Accelerator Labs to a shot of steroids for the organization: enough to accelerate the way it works but not a lifelong health plan. The underlying assumption was that capabilities, [mindsets](/undp-accelerator-labs/references/glossary.md#mindset) and [skills](/undp-accelerator-labs/references/glossary.md#skill) could be transferred to the core business after initial exposure.
We never knew which innovation skills could be transferred, or to what extent, and we could have been more deliberate about figuring that out. Broadly speaking, the more technical a skill set, the harder it is to transfer to others. Data science was a good example (though the current generative AI boom is despecializing many such skills).
At the same time, it is not always about highly technical skills, sometimes even mindsets aren’t transferable. The audacity to lean in performatively to a foresight exercise can make or break how people approach futures, helping them feel what futures could bring. But foresight theatrics, where one convincingly acts as if it is the future, cannot always be taught or transferred from a lab to core business teams.
Once new ways of working are introduced, the question emerges: does the R\&D team hand this over to core teams and take on new agendas, or is it necessary to hold space in that new way of working, and if so, for how long? This is a paradox where both extremes were felt across teams globally.
The tension wasn’t only one born out of territorialism; in fact, that was rarely the problem since R\&D teams are typically keen to let go and find the next new thing. More often, R\&D teams felt they saw a future demand for something that the core business rejected, and we were sometimes stubborn in getting the message that the internal demand wasn’t there.
This tension in the handover was most acutely felt when it came to who would pay. Multiple times, R\&D teams in the network created prototypes which evolved into a new form of engagement with partners. But when it came to business development, they functioned as organizational public goods. They fed into business development but did not recover costs for their input. Beyond budgetary squabbles, this created serious consequences: it often meant that the skills to deliver an R\&D approach were not available after the specialized design phase. This dynamic created a paradox between holding space for the R\&D capability itself and assuming that once modelled, the core organization could take it up.
The paradox of who pays for collective R\&D is a difficult one to balance when working in public or social sectors. In the private sector or academic R\&D, the investments for R\&D come with clear dividends: new products and publications. In our experience, the dividends of R\&D in sustainable development are less obvious, but we found that new insights, partnerships, and perspectives emerged as valuable outcomes. Getting public sector leaders to see R\&D as an investment rather than an expendable budget line requires deliberate action to build legitimacy over time for collective R\&D results.
We came closest to finding the right balance in this paradox when we developed Collective R\&D service lines,[\[3\]](#footnote-2) which enabled some cost recovery for R\&D business development functions. If applied universally, these new functions would generate modest revenue, which would contribute to maintaining the open-ended R\&D capability, allowing it to engage in a continuous cycle of exploration and handover, keeping core business happy by generating new business, while also protecting the space for R\&D teams to do what they do best.
***
## Notes
1. All UN staff take an Oath of Office upon joining. We made this a formal component of the Accelerator Labs onboarding bootcamp given the high number of people joining the UN for the first time in these roles. See: [↑](#footnote-ref-0)
2. The often quoted Jack Welch (2005) "If the rate of change on the outside is faster than the rate of change on the inside, the end is in sight." [↑](#footnote-ref-1)
3. See our “ [R\&D Service Catalogue](https://drive.google.com/file/d/1L4CH0LP-YrFDYKNRINd3pPjIv8it-O9E/view)” (UNDP Accelerator Labs (2025a). [↑](#footnote-ref-2)
---
# Agent Instructions: Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter:
```
GET https://bas-leurss-personal-organization.gitbook.io/undp-accelerator-labs/advancing-r-and-d/epilogue-no-failure-if-there-is-learning.md?ask=
```
The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.