The BRW Typology

The study of infrastructure transitions has a long-standing analytical vocabulary. The multi-level perspective distinguishes niches, regimes, and landscapes; strategic-niche-management theory describes how niches protect and grow; technological-innovation-systems theory maps the functions that support new technologies. Each of these frameworks is useful, and each has helped shape the Lab's thinking. What the Lab has found consistently, in close-to-practice comparative work, is that these frameworks describe what is happening more precisely than they prescribe what a niche actor can actually do. The BRW typology is an attempt at the second.

The typology is a practitioner's tool as much as an analytical one. It distinguishes three — only three — strategies that a niche infrastructure actor can pursue with respect to an entrenched legacy regime. Each strategy has specific conditions under which it is viable, specific failure modes, and specific downstream consequences for the shape of the transition. A practitioner who correctly identifies which strategy is available, which is being pursued, and which is being blocked is a practitioner who can act more intelligently than one who has not done that analysis.

BRW is a niche-navigation typology. It answers a specific question: given a new infrastructure system attempting to establish itself in a context where a legacy regime is already present, what strategic postures are available to the niche actor with respect to the legacy regime? The typology's claim is that the answer falls into three categories, that the categories are exhaustive at the conceptual level, and that most real-world niche actors pursue a mix of two or three of them simultaneously.

BRW is not a prescription for which strategy to pursue. No strategy is inherently superior; each has specific conditions of viability, and a good strategic analysis identifies which configuration of strategies fits the specific context being analysed. BRW is not a theory of why transitions succeed or fail; it is a framework for reading what niche actors are doing and for comparing how different actors in the same sector are positioning themselves differently.

BRW is not a replacement for the multi-level perspective, strategic-niche-management, or other transitions-studies frameworks. It is complementary. MLP and SNM describe the architecture of the transition at a macro level; BRW describes the actor-level strategic choices that populate that architecture. A fully developed transition study uses both.

2.1 Bypass

A bypass strategy is one in which the niche actor constructs a parallel infrastructure that does not depend on the legacy regime's chokepoints. The niche proceeds around the incumbent system rather than through it, building its own distribution, its own customer interface, its own payment relationship, and — critically — its own set of dependencies that the legacy regime does not control.

The canonical bypass case in the Lab's work is battery-swap infrastructure for electric motorcycles. A swap station is a greenfield physical asset. It does not require cooperation from petrol retailers, it does not compete for floor space with an incumbent product line, and it does not need the incumbent regime's distribution network to reach its customers. The rider's fuel purchase — historically a petrol-station transaction — becomes a swap-station transaction, and the petrol retailer has no point of leverage on the new transaction at all.

Bypass strategies work when four conditions coincide. First, the capital cost of building the parallel infrastructure is affordable — typically because the parallel infrastructure is smaller, cheaper, or more geographically targeted than the legacy one. Second, the niche actor has a path to customers that does not require the legacy regime's cooperation — typically through an independent distribution channel. Third, the legacy regime does not control an upstream dependency that would enable it to block the bypass — e.g., does not own the energy supply to the swap network. Fourth, the policy environment permits the bypass — not all jurisdictions permit what amounts to a parallel supply network for a regulated service.

Bypass strategies typically fail in three distinct ways. They fail when the parallel infrastructure is sub-scale (too few swap stations in the right places); when the niche actor cannot reach customers without the legacy regime's cooperation (as, for example, if battery-swap required partnership with petrol-station landlords); or when the legacy regime retains control of an upstream dependency that lets it choke supply (as, for example, if battery-swap depended on battery imports that the incumbent could block through regulatory or commercial channels).

2.2 Repurpose

A repurpose strategy is one in which the niche actor reuses an asset, channel, or capability originally built to serve the legacy regime, redirecting it to new ends. The niche does not construct its own parallel infrastructure from scratch; it co-opts existing infrastructure, typically by striking a commercial or technical arrangement that lets it run its own service over someone else's substrate.

The canonical repurpose case is a pay-as-you-go electric-motorcycle finance product layered above a mobile-money payment rail originally built for peer-to-peer cash transfers. The PAYGO operator does not build its own payment rail; it runs over M-Pesa. The repurpose is deliberate and effective: M-Pesa's reach is enormous, its reliability is trusted, and the marginal cost of adding a new product category is low. Another repurpose example: the layering of electric-motorcycle asset finance above an existing motorcycle-finance channel originally built for petrol motorcycles. The financing-channel infrastructure — credit scoring, operator field presence, collections processes — was built for one product and is redirected to another.

Repurpose strategies work when three conditions coincide. First, the existing infrastructure is genuinely capable of carrying the new service — technically, commercially, and in terms of user trust. Second, the cost of the repurpose arrangement is lower than the cost of building a parallel infrastructure. Third, the owner of the repurposed infrastructure has interests that are at least compatible with, and ideally aligned with, the niche actor's success. When these conditions hold, repurpose can produce adoption far faster than bypass because it does not require greenfield build-out.

Repurpose strategies fail when the repurpose arrangement becomes a source of lock-in in its own right — when the niche actor becomes so dependent on the repurposed infrastructure that the owner of that infrastructure can extract rent, change terms, or discontinue access. The Lab's reading of closed-loop mobile-money architectures is that they are, in effect, successful repurposes that have evolved into lock-in — which is why our finance programme treats the payment-layer question with the seriousness it does.

2.3 Weaken

A weaken strategy is one in which the niche actor (often in concert with other actors, including regulators, advocacy groups, and commercial partners) operates against the legacy regime's lock-in directly. The niche does not build around the legacy (bypass) or borrow from it (repurpose); it works to remove the barriers to entry, switching costs, regulatory protections, or technical chokepoints that keep the legacy system in place.

The canonical weaken case in the Lab's work is regulatory intervention that reduces effective subsidy for fossil-fuel motorcycles — import-duty differentials that favour electric, emissions zoning that restricts ICE motorcycles in specific areas, fuel-subsidy reforms that raise the effective price of petrol. Each of these is a weaken-strategy move: it does not directly advance the electric alternative but does remove advantages the legacy enjoys. Another example: data-portability regulation that reduces the switching cost for riders moving between operators, or interoperability standards that require closed-loop operators to expose APIs to competing networks.

Weaken strategies are distinctive because they often cannot be pursued by the niche actor alone. They typically require regulatory action, coalition-building with other niche actors, or advocacy work that builds public support for dismantling the legacy regime's protections. This makes weaken strategies politically demanding — they involve the niche actor in policy work that bypass and repurpose strategies can largely avoid — but also high-leverage when they succeed, because they can shift the playing field for the entire niche population at once.

Weaken strategies fail when the legacy regime's political defences are stronger than the coalition pushing for change; when the regulatory changes sought are technically correct but practically unenforceable; or when the weakening of the legacy produces short-term welfare costs (higher prices, service disruption) that undermine public support for the transition.

Real niche actors rarely pursue a single strategy. The Lab's reading of the Kenyan electric-mobility case is that the most successful actors combine two or three strategies in a specific configuration that the analyst can read from their operational choices. A brief sketch of how this works in practice:

Swap-network operators (Spiro, Ampersand) run predominantly on bypass — their swap-station networks are greenfield and do not depend on petrol retailers. But they also run on a secondary repurpose of the mobile-money payment layer, which they did not build. And they benefit from weaken-strategy regulatory moves they did not themselves pursue (import-duty differentials on electric motorcycles).

PAYGO financiers (M-KOPA) run predominantly on repurpose — their core infrastructure is a financing channel originally built for solar home systems and now redirected to productive-asset electric mobility. They benefit from the bypass strategies of partner swap operators and from the weaken-strategy regulatory environment, without themselves primarily pursuing either.

Ride-to-own operators (ROAM / 4G Capital) occupy an interesting hybrid position — they run the repurpose of existing motorcycle-finance channels, but also engage in genuine bypass work in constructing their own after-sales, parts, and servicing networks that do not rely on the petrol-motorcycle servicing infrastructure.

A productive way to read a sector, in the Lab's view, is to construct a strategy matrix — a table of actors versus BRW strategies — and populate it with the operational evidence. The matrix typically reveals that different actors are pursuing different mixes, that those mixes track identifiable variations in scale and capability, and that the sector-level outcome depends on which mix becomes dominant. This is one of the BRW typology's most useful analytical applications.

BRW is developed in dialogue with two established bodies of theory, each of which it extends rather than replaces.

The multi-level perspective (MLP) on socio-technical transitions — developed by Geels, Schot, and collaborators — distinguishes three levels at which transitions operate: niches (where novelties incubate), regimes (where established systems maintain stability), and landscapes (where long-term trends exert pressure on regimes and niches alike). MLP is the dominant analytical framework in transitions studies and provides the macro-level architecture within which BRW operates. BRW adds an explicitly actor-strategic layer to MLP's typically structural account: given that a niche exists and a regime exists, BRW says how the niche can proceed against the regime.

The six-domain lock-in taxonomy — developed by Helmrich et al. in the context of infrastructure path dependency — distinguishes six domains in which legacy infrastructures produce lock-in effects: physical-technical, institutional, behavioural, cognitive, political, and economic. Each of the three BRW strategies engages these domains differently. Bypass works primarily against physical-technical and economic lock-in by building parallel physical infrastructure with its own economics. Repurpose works primarily against behavioural and cognitive lock-in by using existing channels that users already trust. Weaken works primarily against institutional and political lock-in by changing the rules of the regulatory game. A mapping between BRW strategies and Helmrich domains is one of the more analytically productive exercises for a researcher using BRW.

The Lab's own contribution sits specifically in the mapping between strategy choice and operational-evidence reading. We do not claim to have invented the three strategies as pure categories; variants of each appear in existing transitions literature. What we claim to have done is consolidate them into a parsimonious three-category typology that a practitioner can apply directly, with a documented evidence base in the Kenyan e-mobility case.

A typology is only as good as the empirical work that produced it. BRW was developed through a specific sequence of research activities whose methods are worth documenting in full, both for transparency and for peer researchers considering extension or replication.

5.1 Primary field research

The initial evidence base is a programme of semi-structured interviews conducted in Nairobi and surrounding areas across 2025–26, under a research protocol approved by the TU Delft Human Research Ethics Committee (HREC). The interview population comprises three distinct cohorts:

• Rider cohort — approximately fifty boda-boda operators, spanning those operating petrol motorcycles, those operating ride-to-own electric motorcycles, and those operating BaaS-configured electric motorcycles. Recruited through a combination of station-based recruitment at swap stations and snowball sampling through the initial interview pool.
• Operator cohort — ten to fifteen semi-structured interviews with staff at the four main industry-actor categories (PAYGO financiers, ride-to-own financiers, BaaS operators, and concessional-finance intermediaries), covering both commercial and operational roles.
• Regulatory and institutional cohort — five to ten interviews with regulators, policy actors, and institutional partners relevant to the e-mobility transition.

Interviews were conducted in English or Swahili (with interpretation support where needed), recorded with explicit consent, transcribed, and coded using a combined deductive-inductive thematic approach. Ethical considerations — consent, data minimisation, anonymisation, cross-border data handling under GDPR and the Kenya Data Protection Act 2019 — are documented in the HREC-approved protocol and summarised in the regulatory-frameworks entry.

5.2 Comparative case analysis

Alongside the interview work, the typology was refined through comparative case analysis of the four main industry actors (Spiro, Ampersand, ROAM/4G Capital, M-KOPA) and the three financing-architecture categories that comprise them. Secondary sources — trade press, operator disclosures, academic publications, policy documents — triangulate the interview evidence.

The comparative method is explicitly configurational: each case is read not for its individual features but for the specific configuration of BRW strategies it pursues, and the cross-case variation is where the analytical traction comes from. A case that runs purely on bypass would not contribute new information to the typology; the useful cases are those that mix strategies in distinguishable ways.

5.3 Transaction-data analysis

Where design-partner operators have shared anonymised transaction data, the typology's empirical claims are checked against actual operational patterns. A BRW analysis that claims a particular actor is pursuing repurpose of a payment rail should be legible in the transaction data as a specific pattern of payment flow. Where the transaction data contradicts the qualitative reading, the qualitative reading is revised.

5.4 Iterative refinement

The typology as published here is version 2.0. Version 1.0, developed in 2025 through the initial phase of the MSc thesis work, included four strategies rather than three; a fourth category (accommodate, in which the niche absorbs some of the legacy regime's constraints rather than challenging them) was merged into repurpose once comparative work showed that the distinction was not analytically productive. The typology remains open to further revision as the evidence base grows; substantive revisions will be documented with version history.

For a researcher or practitioner wanting to apply BRW to a new sector or case, the Lab recommends the following sequence.

1. Identify the legacy regime and its lock-in structure

   Begin with the Helmrich six-domain taxonomy. Which of the six domains of lock-in are active in the case being studied? The answer shapes which BRW strategies are most available to the niche.

2. Identify the niche actors and catalogue their strategy mix

   Construct the strategy matrix described in §3 above. For each significant niche actor, identify which of the three strategies they pursue and in what mix. The matrix is typically more revealing than any single-actor analysis.

3. Test the strategy mix against operational evidence

   A claim that an actor pursues bypass should be legible in the form of parallel infrastructure investment. A claim of repurpose should be legible in partnership agreements and dependency relationships. A claim of weaken should be legible in policy engagement and coalition-building. Where the claim and the evidence do not align, the claim is wrong.

4. Identify the conditions of viability for each strategy in the specific case

   Return to the conditions outlined in §2 above. Which conditions of viability are met in this case for each strategy? Which are not? The answer tells you which strategies are live options for the niche and which are foreclosed.

5. Read the sector-level trajectory

   Given the strategy mix of the niche actors and the conditions of viability in the specific case, what does the sector-level trajectory look like? Which actors are positioned to succeed? Which will fail? Where will the transition run fast and where will it stall? This is where the typology's analytical yield is largest.

BRW is, at version 2.0, a working typology rather than a closed theory. Several limitations are known and are the subject of ongoing work.

Boundary cases. Some niche strategies do not fit cleanly into any of the three categories. A sector that is simultaneously being bypassed by new infrastructure and repurposing different parts of the legacy may be better described as two distinct transitions than as one BRW configuration. The typology does not yet have a clean decision rule for when to treat a case as one transition versus several.

Cross-strategy interaction. The typology says relatively little about how the three strategies interact when pursued in combination. The Lab's working hypothesis is that bypass and repurpose can be complementary (a swap-network operator repurposing mobile-money payments is pursuing both), but that weaken often requires disengagement from repurpose relationships that would be compromised by regulatory confrontation with the incumbent. The interaction question deserves dedicated treatment.

Temporal dynamics. The typology is, as currently specified, a static reading of an actor's strategy at a point in time. Real actors shift strategies as their position in the transition matures. A dynamic version of the typology — one that tracks strategy evolution over time — is a natural extension and is in preparation.

Extension beyond infrastructure transitions. BRW was developed for infrastructure transitions specifically. Whether it extends to other forms of socio-technical transition — consumer-product categories, service-sector transitions, institutional transitions — is an open question. The Lab's comparative work across finance, electrification, and water is the closest test to date; extension further afield is future work.

The BRW typology is the Lab's signature analytical contribution and is published here, with full documentation, for open use by other researchers and practitioners. Citation is welcome. Extension — applying BRW to new sectors, refining the strategy definitions through further cases, or developing the theoretical grounding further — is actively invited; readers pursuing this work are encouraged to contact the Lab so that extensions can be indexed in future versions of this document.

A formal working-paper version of the typology, with full references, empirical case documentation, and comparative analysis, is in preparation and will be published through the articles index during the 2026 publication cycle.