Scaling Mechanisms

Faster-growing firms are almost twice as likely to innovate as slow-growing ones. The direction of that innovation matters as much as its presence. Nagji and Tuff's Harvard Business Review paper "Managing Your Innovation Portfolio" - drawing on research across industrial, technology, and consumer goods sectors - found that outperforming firms typically allocate approximately 70 percent of innovation resources to core improvements, 20 percent to adjacent initiatives, and 10 percent to transformational efforts, but the long-term cumulative return on innovation investment runs in the opposite direction: transformational initiatives generate roughly 70 percent of returns, adjacent 20 percent, and core 10 percent.

The Nagji-Tuff finding sits inside the March exploration/exploitation tension introduced earlier. Core improvements are exploitation; transformational initiatives are exploration; the systematic underweighting of exploration that March identified at the organisational level appears at the innovation-portfolio level as the 70/20/10 input pattern. The ecosystem has tended to invest heavily in incremental improvement and too little in the transformational innovation that generates disproportionate returns. The correction period validated this: the ventures that demonstrated durable competitive positions were those that had made genuinely transformational bets early - Moniepoint on POS infrastructure and credit scoring, Flutterwave on cross-border payment rails, Sun King on PAYGO energy financing - rather than those that had optimised existing models.

Innovation advisory support at the scale-up stage should not be episodic but ongoing, structured around four components: an innovation audit documenting current and immediately planned initiatives across product, service, and internal systems; a scorecard evaluating the quality and strategic direction of current innovation activity; a forward pathways scan identifying areas for future revenue and growth aligned to investor objectives; and commercially oriented innovation plans that are actionable and roadmapped, not directional and aspirational. A growth playbook with joint, co-owned actions between support providers and ventures - governed in 30–60–90 day cycles - ensures alignment between what needs to be done and who is accountable for doing it.

The most ambitious African scaling ventures are not just building businesses. They are building markets - creating demand that did not exist, establishing distribution infrastructure that competitors subsequently use, and setting regulatory precedents that reshape the rules of the game for everyone who follows. The substantive treatment of the Meadows leverage-points framework - and where in the leverage hierarchy the highest-impact interventions sit - sits in Feedback Loops; the implication for scaling-support design is that this systems-level ambition requires a different kind of leadership development than operational management training provides.

The leadership-development literature has named what is required. Ronald Heifetz's Leadership Without Easy Answers - the foundational treatment of adaptive leadership - distinguishes technical challenges (where the problem and solution are both known and the work is implementation) from adaptive challenges (where the problem itself is contested, the solution is unknown, and the work is collective sense-making and behaviour change). African scaling ventures encounter both kinds. Most support is calibrated to technical challenges; the consequential systemic moves are adaptive in Heifetz's sense. Westley, Zimmerman and Patton's Getting to Maybe extends the analysis specifically to systems-change leadership in complex, multi-stakeholder environments. Their central finding: systems change happens through coalitions of actors holding different positions and interests, mobilised by leaders who can hold complexity without prematurely simplifying it. The capacity to do this is not a personality trait but a set of practices that can be developed and supported.

Systems innovation leadership development needs three interconnected components. At the individual level: the collaborative leadership skills that enable learning and trust-building across stakeholder groups who do not naturally share incentives - government, investors, partner businesses, and communities - and the capacity to empower action in others rather than concentrate it. At the community level: coalition-building and advocacy capabilities that develop alignment and mobilise action around shared objectives - the skills that allow a venture to become an industry standard-setter rather than one player among many. At the systems level: a working understanding of the complex systems shaping the venture's operating environment - the feedback loops, delay structures, and political-economy dynamics that determine whether structural conditions are changing in ways that open or close scaling pathways. The ventures that have achieved genuine systemic impact in African markets were led by founders who understood this architecture, not just their own business model.

The knowledge that experienced founders hold is the scarcest and most practically valuable resource in the African scaling ecosystem - and it is systematically dissipated because there is no efficient mechanism to capture, structure, and route it to the ventures that need it. The substantive treatment of the tacit-vs-codified-knowledge distinction (Polanyi, Nonaka & Takeuchi) and why the codification work is structurally underprovisioned sits in Data, Insights and Knowledge.

Etienne Wenger's Communities of Practice - the foundational treatment of how knowledge accumulates in distributed practitioner networks - names the institutional form that the African scaling ecosystem has not yet built. Communities of practice are groups of practitioners who share a domain, build relationships through sustained interaction around shared problems, and develop a shared repertoire of practices, tools, and language over time. Wenger's empirical finding: the most consequential professional knowledge does not transfer through formal training or codified documents. It transfers through participation in communities of practice - through the apprenticeship-mode learning that comes from sustained engagement with people who have done the work. The implication for African scaling is direct. Most accelerator and ESO alumni engagement amounts to database management. It should amount to something structurally different: sustained, relationship-based engagement with ventures as they grow; peer networks structured around operational challenges rather than sector affiliation; and formal mechanisms that convert experienced founder knowledge into accessible, searchable guidance.

The Endeavor brokerage-network architecture treated substantively in What Working Support Infrastructure Looks Like is the closest African approximation of a working community of practice at scale. The accessibility constraint - only Endeavor-selected portfolio ventures access it - is what limits its ecosystem-wide impact. The design challenge is building communities of practice that operate at greater scale without losing the sustained-engagement quality that makes them work.

AI now makes the peer-matching function genuinely scalable. Founders with specific operational challenges can be connected to those who have navigated similar situations - matching by challenge type, market context, sector, and stage - at a speed and scale that human-curated networks cannot match. The knowledge accumulated through years of founder experience, which currently dissipates when individuals move on or ecosystems fragment, can increasingly be structured and made searchable. Building AI-assisted alumni networks is not an enhancement to scaling support programmes. It is the mechanism that makes the knowledge embedded in the ecosystem's most successful founders available to the ventures most likely to benefit from it.