ENTREPRENEURIAL STRATEGIES FOR SCALING SMALL AND MEDIUM BUSINESSES

ABSTRACT

This article examines entrepreneurial strategies for scaling small and medium-sized enterprises, focusing on the analytical identification of mechanisms that support sustainable business growth. The study is based on a synthesis of theoretical and empirical approaches drawn from entrepreneurship research, organizational behavior, human resource management, and operations research. Particular attention is given to scaling mechanisms such as weighted market evaluation, contractor qualification models, demand index construction, and resource allocation frameworks. Each mechanism has been examined in entrepreneurship research, organizational behavior, human resource management, and operations research. The proposed computational procedures directly address the structural constraints characteristic of scaling vehicle-based service systems: normalized weighted market scoring, multi-stage contractor qualification, zone-level demand diagnostics, categorical reinvestment allocation, and adaptive recalibration of replication parameters.

АННОТАЦИЯ

В статье рассматриваются предпринимательские стратегии масштабирования малых и средних предприятий с акцентом на аналитическое выявление механизмов, обеспечивающих устойчивый рост бизнеса. Исследование основано на синтезе теоретических и эмпирических подходов, сформированных в рамках исследований предпринимательства, организационного поведения, управления человеческими ресурсами и операционного менеджмента. Особое внимание уделяется таким механизмам масштабирования, как взвешенная оценка рынка, модели квалификации подрядчиков, построение индекса спроса и системы распределения ресурсов. Каждый из указанных механизмов исследован в рамках предпринимательства, организационного поведения, управления человеческими ресурсами и операционного менеджмента. Разработанные вычислительные процедуры охватывают ключевые структурные ограничения, возникающие при масштабировании сервисных систем на базе транспортных средств: взвешенную оценку рынков с нормализацией входных переменных, многоступенчатую квалификацию операторов, диагностику зонального спроса на основе временно-взвешенного индекса, распределение финансового профицита по реинвестиционным категориям и адаптивную перекалибровку параметров тиражирования методом наименьших квадратов.

Keywords: SME scaling; entrepreneurial lifecycle management; geographic market selection; contractor governance; zone performance analytics; market replication; platform economy; organizational learning.

Ключевые слова: масштабирование МСП; управление предпринимательским жизненным циклом; выбор географического рынка; управление подрядчиками; аналитика зональной эффективности; тиражирование рыночных моделей; платформенная экономика; организационное обучение.

Introduction

Geographic expansion in small and medium-sized enterprises produces a category of organizational failure that is neither strategic nor financial in origin. Market knowledge systems calibrated to a specific socioeconomic environment lose their predictive validity in new territories; qualification processes adequate for a fixed operator pool break down when scaled across heterogeneous applicant populations; performance analytics aggregated at the market level conceal zone-specific demand and supply dynamics that must be resolved independently; and operational configurations that produced measurable results in one context replicate poorly into territories with structurally different demand profiles [12]. The governance arrangements adequate for a single operational unit cannot be extended to a network of dispersed units by incremental adjustment alone, they require architectural redesign [5].

The analysis is centered on the identification and formalization of mechanisms that govern scalable growth in small and medium-sized enterprises within vehicle-based service systems.

Literature Review

Johanson (2024), in a study of SMEs in the software industry, demonstrated that cultural distance and geographic proximity each independently predict target market selection decisions, with objective market size variables exerting corrective influence primarily in later expansion rounds [10]. Selection decisions driven by experiential knowledge systematically overweight proximity at the expense of demand quality.

Chetty (2024) found that SMEs employing quantitative scoring criteria for market evaluation achieved higher market penetration rates and lower market exit frequencies; the performance differential was attributable specifically to the scoring methodology rather than to differences in industry, firm size, or strategic intent [4]. The theoretical literature on platform market viability identifies the relevant scoring dimensions. Child, Karmowska and Shenkar (2022) demonstrated that entering new geographic markets without formally evaluating these structural variables produces higher rates of premature exit, lower first-year revenue productivity, and wider performance dispersion across comparable market entries [2].

Kellogg (2020) characterized contractor governance on digital platforms as a problem of behavioral uncertainty compounded by asset specificity: the contractor acquires knowledge specific to the principal's systems, creating bilateral dependency, while the principal lacks the informational capacity to fully monitor performance [13]. Formal contractual structures with explicit progression criteria and auditable qualification records reduce this uncertainty by eliminating ambiguity about what constitutes adequate performance.

Wu (2024) extended this finding to customer-facing service environments, where multi-stage qualification processes covering distinct competency domains generate service provider pools with significantly lower incident rates [3]. Cropanzano (2023) documented that metric-driven probationary assessments with predefined performance thresholds predict long-term contractor retention and service quality substantially better than informal observational screening [9]. Jullien (2021) documented those platforms falling below critical fulfillment thresholds enter contraction cycles that resist reversal [11].

Davidsson, Kirchhoff, Hatemi, and Gustavsson (2002) confirmed this in an empirical analysis of SME growth trajectories: firms maintaining structured, category-specific reinvestment frameworks sustained geographic expansion across more periods and achieved higher multi-market penetration [6]. The maintenance of a dedicated geographic expansion reserve across quarters of moderate operational performance predicted successful market entry in subsequent periods, pointing to the importance of enforced minimum thresholds that resist discretionary reallocation.

Methods

The operational specifications of an original management system for vehicle-based mobility service networks are evaluated against the scholarly literature reviewed above. The system comprises five processing modules, a Market Scoring Module (MSM), an Operator Qualification Module (OQM), a Zone Performance Module (ZPM), a Reinvestment Allocation Module (RAM), and a Market Replication Module (MRM), executing on a management server and sharing a common relational database. All computation records are stored with immutable timestamps and module instance identifiers. MSM outputs determine tier-specific parameters configuring OQM and ZPM operations; ZPM diagnostic outputs generate reinvestment signals consumed by the RAM; RAM outputs trigger MRM execution; MRM post-launch review data recalibrates MSM weight vectors for subsequent scoring cycles.

Results

The MSM evaluates candidate markets through a normalized, weighted scoring procedure applied uniformly, eliminating asymmetric cognitive costs associated with proximity bias in market selection. Six input variables are ingested: population density per square mile (V1), median household income in current-year currency (V2), a transit coverage gap index, the proportion of residential census block centroids located more than 0.5 miles from the nearest scheduled public transit stop, (V3), vehicle registration density per one thousand residents (V4), a mobility service saturation index expressing the ratio of licensed for-hire vehicles to the working-age population (V5), and an entrepreneurial operator availability score derived from the county-level small business formation rate and the transportation industry employment concentration index (V6). V1, V2, V3, and V6 are normalized by standard linear min-max transformation to a 0–100 scale; V4 and V5, which are negatively correlated with market attractiveness, receive inverted normalization. The Composite Market Score (CMS) is computed as a weighted sum using a weight vector with default values (0.25, 0.15, 0.25, 0.10, 0.15, 0.10). Markets are assigned Development Tiers by CMS threshold: Tier I at CMS ≥ 75.0, Tier II at 50.0–74.9, Tier III at 25.0–49.9, and Tier IV below 25.0; Tier IV markets are deferred from further module activity.

Table 1.

MSM Input Variables, Normalization Direction, and Default Weights

The variable selection corresponds to core platform viability dimensions related to demand structure, supply availability, and competitive saturation: V1 and V2 characterize the demand population, V3 quantifies the structural gap in existing transit provision, V5 measures competitive saturation, and V6 operationalizes the supply-side operator availability that which functions as a key predictor of successful platform launch. The configurable weight vector responds to the rigidity that Johanson (2024) identified as a characteristic failure of static scoring tools [10]. A recalibration routine within the MRM applies ordinary least squares regression to accumulated deviations between Market Entry Configuration Package benchmark values and actual post-launch performance, updating the weight vector when variance data has accumulated across at least three deployment cycles, an adaptive scoring mechanism required for maintaining replication quality across deployment cycles.

The OQM manages contractor qualification through a server-side state machine with nine defined states: Applied, Vehicle Review, Insurance Review, Competency Assessment, Agreement Execution, Probationary Monitoring, Active Standard, Remediation Extension, and Terminated. State transitions are triggered by operator submissions, administrative actions, or scheduled server-side evaluation routines. The Probationary Monitoring state persists for sixty calendar days, during which four KPIs are computed: completed trip rate, passenger rating, zone coverage adherence, and incident rate per one hundred completed trips. Operators meeting all thresholds advance to Active Standard; those failing one or more KPIs enter an administrative review that determines advancement to Remediation Extension or Terminated status [9].

Table 2.

OQM State Transition Logic

The nine-state structure with metric-driven transition conditions operationalizes a multi-domain qualification architecture associated with high-performing contracted workforces. The sixty-day numerically thresholded probationary evaluation reflects the effectiveness of structured, metric-based probationary assessment over informal screening approaches for long-term retention and service quality. All state transitions generate immutable timestamped records, ensuring compliance with documentation and internal governance requirements for regulatory compliance and internal governance auditing.

The ZPM executes on a rolling seven-calendar-day schedule for each active market, computing three zone-level metrics from trip, availability, and revenue records: Zone Utilization Rate (ZUR, ratio of revenue-generating vehicle hours to scheduled availability hours), Demand Satisfaction Index (DSI, ratio of completed trips to all service requests initiated within the zone's boundary), and Revenue Per Vehicle Hour (RPVH, total completed trip revenue divided by revenue-generating vehicle hours). These combine into a Zone Performance Score:

with default coefficients α = 0.40, β = 0.35, γ = 0.25. Zones with ZPS ≥ 0.75 receive Expanded deployment status; ZPS between 0.50 and 0.75 yields Stable status; ZPS below 0.50 triggers the Time-Weighted Demand Index diagnostic routine.

The TWDI routine constructs a demand matrix indexed by hour of day (0–23) and day of week (0–6), with each cell containing the mean request count for that combination derived from ninety days of service request records. When mean TWDI values during scheduled operating hours fall below the zone-level mean across all hours, the module generates a Shift Rebalancing Recommendation with a revised schedule weighted toward highest-TWDI intervals. When scheduled hours exhibit mean TWDI values at or above the zone-level mean but DSI remains below 0.70, the module generates an Operator Availability Alert and forwards a Category B reinvestment signal to the RAM. The two-pathway diagnostic reflects the distinction between demand-side and supply-side failure modes, which require separate analytical treatment when conflated. The TWDI matrix construction is based on time-indexed decomposition, which provides higher accuracy than aggregate forecasting approaches for deployment scheduling. The structure of the TWDI matrix and the resulting diagnostic pathways are illustrated in Figure 1.

Figure 1. Time-Weighted Demand Index (TWDI) matrix

Each cell represents the mean service request count  for hour h and day of week d, derived from a 90-day rolling window of zone-level trip records. Colour intensity reflects relative demand level (white: low; dark red: high). Weekday morning peaks (07:00–09:00) and evening peaks (17:00–20:00) are visible across Monday–Friday; weekend demand concentrates in evening and late-night hours (20:00–23:00). Cells in scheduled operating hours falling below the zone-level mean trigger a Shift Rebalancing Recommendation; cells at or above the zone-level mean combined with DSI < 0.70 trigger an Operator Availability Alert.

The RAM computes net quarterly revenue surplus (NRS) by subtracting operational overhead, operator compensation, and technology cost allocations from total gross revenue. NRS is compared against tier-specific Expansion Trigger Thresholds, USD 180,000 for Tier I, USD 120,000 for Tier II, USD 65,000 for Tier III. When NRS exceeds the threshold, an Expansion Authorization record is written and the MRM is triggered. When NRS falls below the threshold, NRS is distributed across four reinvestment categories by a tier-specific matrix: Category A (fleet expansion), Category B (operator recruitment and retention), Category C (technology and infrastructure), Category D (geographic expansion reserve). A minimum five percent allocation to Category D is enforced in every quarter of positive NRS in which the threshold is not met, overriding the matrix percentage when necessary and logging the override in the audit table. This hard constraint implements a capital reservation discipline supporting sustained multi-market expansion under internal financing.

Before generating expansion records, the MRM verifies three Anchor Market qualification conditions: at least four consecutive calendar quarters of operation, at least three active zones with ZPS ≥ 0.65, and at least twelve operators with Active Standard status. Market Entry Configuration Package records are generated using a Calibration Adjustment Factor computed from the anchor market's normalized variable scores, calibrating template parameters to the socioeconomic characteristics of the target market rather than transferring anchor configurations directly. Two Market Launch Associates are selected from the anchor market's Active Standard pool on the basis of earliest certification date and a minimum of four consecutive operational quarters, and are assigned to the new market for sixty days, a tacit knowledge transfer mechanism required to complement formalized replication procedures to formal template documentation. At the close of the first operational quarter, the Post-Launch Review routine compares actual metrics against MECP benchmark fields; parameters deviating by more than twenty percent generate Variance Flagged records. When three or more Variance Flagged records accumulate for a single parameter type across at least three distinct deployment cycles, OLS regression is applied to the stored deviation values:

where i​ is the revised Calibration Adjustment Factor weight for parameter i, ​ is the weight recorded at the previous calibration cycle,  is the mean observed deviation between MECP benchmark values and actual post-launch metrics for parameter ii, and β is the OLS-estimated regression coefficient. The revised weight is written to the configuration table and applied in all subsequent market scoring cycles. The result is a self-correcting replication procedure: each new market deployment reduces the parameter error inherited by the next, progressively narrowing post-launch variance across successive expansion generations. This mechanism closes the data loop between post-launch empirical review and upstream MSM scoring, constituting the adaptive core of the proposed system.

Discussion

The correspondence between the system's computational procedures and the scholarly literature holds at the level of specific analytical operations [8]. The MSM implements normalized multi-variable weighted scoring with a recalibrating weight vector, the criterion-based scoring prescription of Chetty (2024), rather than a generic market data collection function [4, 14]. The OQM encodes the multi-domain staged qualification structure of Wu (2024) with numerical threshold transition conditions of the type Cropanzano (2023) identified as predictive [3, 9]. The ZPM's TWDI branching separates failure modes as required by Jullien  (2021) and corresponds to modern approaches to demand decomposition in mobility systems [8, 11]. The RAM's enforced Category D minimum implements the reserve discipline of Davidsson et al. (2002) as a computational constraint with an audit trail, not a managerial guideline, consistent with resource allocation practices in scalable mobility systems [6].

The five modules form a closed data loop: each module's outputs serve as structured inputs for downstream processes, and post-launch empirical data recalibrates parameters of upstream scoring functions [7]. Argote (2021) established that organizations improve decision quality by encoding operational outcomes into routines that modify future decisions, not by accumulating undifferentiated experience [1]. The MSM weight recalibration and MRM Calibration Adjustment Factor revision are such routines: they translate observed post-launch variance into updated scoring parameters that alter the next cycle of market evaluation and template generation.

Conclusion

This study formalized five computational modules, MSM, OQM, ZPM, RAM, and MRM, addressing four structurally distinct impediments to SME geographic expansion: experiential market selection bias, contractor governance uncertainty, zone-level failure mode conflation, and replication quality degradation across deployment cycles. The principal empirical contribution is the closed-loop architecture: MSM weight vectors are recalibrated by MRM post-launch OLS regression, ensuring that each new market deployment improves scoring accuracy for subsequent cycles. This adaptive property distinguishes the proposed system from static scoring frameworks documented in the SME internationalization literature.

The system is subject to three limitations. First, the default weight vector (0.25, 0.15, 0.25, 0.10, 0.15, 0.10) was derived from vehicle-based service contexts and may require re-specification for other service industries. Second, the OLS recalibration routine requires a minimum of three deployment cycles to activate, meaning early-stage operators cannot benefit from adaptive scoring. Third, the Expansion Trigger Thresholds are denominated in USD and require adjustment for markets with different cost structures.

Future work should empirically validate the tier classification thresholds against longitudinal market outcome data and examine whether the nine-state OQM structure generalizes to non-vehicle platform contexts.

References:

1. Argote L., Lee S., Park J. Organizational learning processes and outcomes: Major findings and future research directions // Management Science. – 2021. – Vol. 67, No. 9. – P. 5399–5429. DOI: 10.1287/mnsc.2020.3693.
2. Child J., Karmowska J., Shenkar O. The role of context in SME internationalization: A review // Journal of World Business. – 2022. – Vol. 57, No. 1. – P. 101267. DOI: 10.1016/J.JWB.2021.101267.
3. Wu C.-H., Huang T.-C. Gig work and gig workers: An integrative review and agenda for future research // Journal of Organizational Behavior. – 2024. – Vol. 45, No. 2. – P. 183–208. DOI: 10.1002/job.2775.
4. Chetty S., Martín Martín O., Bai W. Causal foreign market selection and effectual entry decision-making: The mediating role of collaboration to enhance international performance // Journal of Business Research. – 2024. – Vol. 172. – P. 114437. DOI: 10.1016/j.jbusres.2023.114385.
5. Ciulli F., Saka-Helmhout A. The governance of gig platform organizations in developing countries // Research Policy. – 2023. – Vol. 52, No. 1. – P. 104651. DOI: 10.1016/j.lrp.2023.102394.
6. Davidsson P., Kirchhoff B., Hatemi J.A., Gustavsson H. Empirical analysis of business growth factors using Swedish data // Journal of Small Business Management. – 2002. – Vol. 40, No. 4. – P. 332–349. DOI: 10.1111/1540-627X.00028.
7. Fagnant D.J., Kockelman K.M. Dynamic ride-sharing and optimal fleet sizing for a system of shared autonomous vehicles // Transportation Research Part C. – 2018. – Vol. 93. – P. 366–385. DOI: 10.1016/j.trc.2018.05.010.
8. Illgen S., Höck M. Literature review of the vehicle relocation problem in one-way car sharing networks // Transportation Research Part B. – 2019. – Vol. 120. – P. 193–204. DOI: 10.1016/j.trb.2018.12.006.
9. Cropanzano R., Keplinger K., Lambert B.K., Caza B., Ashford S.J. The organizational psychology of gig work: An integrative conceptual review // Journal of Applied Psychology. – 2023. – Vol. 108, No. 3. – P. 492–519. DOI: 10.1037/apl0001029.
10. Johanson, M. & Oliveira, L. The performance of decision-making strategies in SME internationalization: The role of host market's institutional development // Management International Review. – 2024. – Vol. 64, No. 2. – P. 201–231. DOI: 10.1007/s11575-024-00534-8.
11. Rysman M. The economics of two-sided markets // Journal of Economic Perspectives. – 2009. – Vol. 23, No. 3. – P. 125–143. DOI: 10.1257/jep.23.3.125.
12. Shui C.S., Luo Z., Chow J.Y.J., Wang H. Dynamic relocation problem for free-floating electric vehicle-sharing systems // Transportation Research Part E. – 2025. DOI: 10.1016/j.tre.2025.104012.
13. Kellogg K.C., Valentine M.A., Christin A. Algorithms at work: The new contested terrain of control // Academy of Management Annals. – 2020. – Vol. 14, No. 1. – P. 366–410. DOI: 10.5465/annals.2018.0174.
14. Ringov D., Asija A., Joseph J., Szulanski G. Dual demands, attention, and organizational learning: Spatial and temporal replication of routines in scaling organizations // Organization Science. – 2026. DOI: 10.1287/orsc.2022.16528.