Digital pull logistics in the Ukraine conflict: dynamic capabilities and institutional theories applied to coalition humanitarian operations

Coordinating logistical efforts after disasters and complex emergencies remains one of the central challenges in humanitarian supply chain (SC) management. Relief operations often suffer from duplication of effort, mismatches between needs and deliveries and significant delays in aid distribution. These difficulties stem from fragmented information flows, diverse donor objectives and the absence of cohesive coordination mechanisms (Day et al., 2009; da Costa et al., 2014; OCHA, 2012). While push-based approaches – where aid is dispatched according to forecasts and prepositioned supplies – can ensure rapid mobilization, they frequently lead to oversupply of inappropriate items and under-provision of critical needs (van Wassenhove, 2006; Balcik and Beamon, 2008; Wisetjindawat et al., 2014). Conversely, pull-based systems – structured around actual requests from the field – offer greater alignment with local demand but are rarely implemented effectively at scale in humanitarian contexts (Ti and Kinsey, 2023). Related research on shelter-supply networks similarly underscores the importance of responsiveness to local conditions and supplier relationships (Copping et al., 2022).

The conflict in Ukraine provides a compelling lens through which to examine coordination challenges in large-scale humanitarian and defense logistics. Following the escalation of hostilities in February 2022, as illustrated in Figure 1, more than 40 donor nations contributed humanitarian and security assistance (SA) across a global supply network spanning thousands of miles (EU Assistance to Ukraine, 2024). The speed and magnitude of these contributions – often mobilized through multiple national systems – created significant coordination and information-management difficulties.

In this context, coordination was structured through a donor-driven governance model in which major contributing states retained decision authority over resources while Ukraine articulated operational demand signals. This configuration offered important advantages – most notably speed of mobilization, centralized visibility and enhanced security control in an active conflict zone. At the same time, such an arrangement introduces structural asymmetries, including limited direct local voice in allocation decisions and potential constraints on transferability to more decentralized humanitarian settings. Recognizing both the strengths and limitations of this governance structure is essential to understanding how digital, pull-based platforms such as Korovai function in high-security coalition environments.

Similar problems of duplication and delayed delivery had been observed in earlier crises such as the Haiti earthquake and the Banda Aceh tsunami, where hundreds of uncoordinated donors produced overlapping aid flows and resource gaps (Tatham et al., 2017; Vohr, 2011). The Ukrainian response, however, differed from these prior events in both scale and duration, occurring amid an active conflict that demanded sustained, adaptive logistics under political and security constraints (Azarov et al., 2023).

It was within this environment that Korovai was introduced – a digital coordination platform jointly developed by the USA, the UK and Ukraine to centralize requests, track fulfillment and reduce redundancy across donor contributions, addressing coordination challenges within a SC that encompassed 41 donor nations (Ukraine Support Tracker, 2025).

The magnitude of this assistance effort was without precedent in modern humanitarian or defense logistics. The USA alone contributed approximately $46bn in military aid and $1.6bn in humanitarian support, while more than 30 partner nations together provided an additional $80bn across both domains (Masters and Merrow, 2024). Averaging more than $5bn per month in combined contributions, this scale far exceeded any recent disaster or complex emergency and created a level of logistical complexity that demanded a new coordination architecture such as Korovai.

Beyond aggregate government aid, Ukraine’s SC exhibits sensitivity to civilian donor income and price volatility. Kushnir et al. (2024) demonstrate how shifts in donor income directly influence demand for critical goods, creating financial fragilities that parallel those observed in humanitarian cash-transfer systems.

Recent humanitarian logistics (HL) research highlights persistent barriers to collaboration – including trust, information sharing and joint planning – as critical factors (CFs) affecting performance (Hudnurkar et al., 2014). Studies of civil–military coordination (Heaslip and Barber, 2014; Heaslip, 2013) and peacekeeping logistics (Grigoli et al., 2024) highlight persistent coordination and communication challenges across institutional boundaries. While recent research highlights the growing use of digital tools in humanitarian procurement and coordination (Moshtari et al., 2021), empirical analyses of digital or platform-enabled mechanisms remain limited. As Kushnir et al. (2024) observe, information flows form the critical linkage between financial contributions and material deliveries in Ukraine’s defense SC. Digital platforms, therefore, do not merely enhance visibility; they mediate the translation of donor resources into operational capability.

Scholarship has examined the tradeoffs among in-kind, cash and local procurement modalities in humanitarian response (Piotrowicz, 2018), yet little attention has been given to the digital infrastructures or coordination platforms that increasingly influence these choices in complex emergencies. This study, therefore, addresses a clear gap by examining Korovai’s implementation and assessing its impact on HL during the Ukraine conflict.

To interpret how digital coordination platforms reshape HL, this study draws on dynamic capabilities theory (DCT), which explains how organizations sense, seize and reconfigure resources under conditions of uncertainty (Teece et al., 1997), and institutional theory (IT), which emphasizes how legitimacy and shared norms influence organizational adoption and coordination (DiMaggio and Powell, 1983; Scott, 2014). Together, these perspectives provide complementary insights into the adaptive and legitimizing functions of the Korovai platform.

The contribution of this research is threefold. First, it provides unprecedented, first-hand empirical evidence of Korovai’s operation through semistructured interviews with personnel at the US European Command Control–Ukraine (ECC-U) and its successor, the International Donor Coordination Center (IDCC). Second, it situates Korovai within the broader literature on push versus pull logistics and on collaborative SC practices, extending theory to the context of a large-scale conflict. Third, it offers practical insights into how digital platforms can improve efficiency, transparency and coordination in humanitarian operations, while also highlighting risks such as donor dominance, cybersecurity and the digital divide. The study concludes by considering the transferability of Korovai’s principles to sudden-onset disasters and other humanitarian emergencies.

HL and the international aid effort in support of Ukraine both involve the management and distribution of relief supplies during crises. A key consideration in response efforts is whether to adopt a push or pull logistics strategy. Each strategy has distinct implications for efficiency, responsiveness and stakeholder coordination. In this paper, we review HL literature relevant to the response to the conflict in Ukraine and examine the applications of push and pull logistics systems to learn from the successes and shared issues of a military conflict and other disaster responses.

HL operates under severe constraints of uncertainty, urgency and resource scarcity. Two archetypal supply approaches dominate: push systems and pull systems. Push systems rely on forecasts, prepositioned stock and historical patterns of need (van Wassenhove, 2006; Balcik and Beamon, 2008). They enable rapid deployment of relief items but often create oversupply of inappropriate goods or under-provision of critical resources when conditions change quickly (Wisetjindawat et al., 2014; Klug, 2023). Pull systems, in contrast, are demand-driven, aligning supply with real-time requirements through feedback and data-sharing mechanisms to reduce waste and duplication (Dubey and Gunasekaran, 2016; Ti and Kinsey, 2023). Related research on shelter and supplier networks emphasizes the importance of network visibility and locally informed decision-making in achieving similar responsiveness (Copping et al., 2022).

Empirical evidence shows both the promise and limitations of these approaches. After Hurricane Katrina, for instance, large quantities of housing units and ice went unused due to mismatches between supply and need (Day et al., 2009). Likewise, the 2011 Japan tsunami and the 2015 Nepal earthquake revealed that prepackaged aid often failed to meet local or seasonal requirements, while pull systems were slowed by damaged infrastructure (OCHA, 2015; Vohr, 2011; Heaslip, 2013). These experiences have led to hybrid models that maintain readiness through prepositioning while integrating flexible, demand-driven mechanisms once local needs are reliably assessed (Schulz and Blecken, 2010; Upadhyay et al., 2022).

Push logistics emphasizes prepositioning resources and decisions based on anticipated demand derived from historical data and forecasting (Duran et al., 2011; Kovács and Spens, 2007; Yi and Özdamar, 2007; Minculete and Olar, 2016). This strategy prioritizes immediate availability but risks over- or under-stocking when demand shifts (Van Wassenhove, 2006; Behl and Dutta, 2019). Because information flows one-way from supplier to customer, decision-makers remain distant from actual demand signals and schedule-driven operations can become rigid (Ti and Kinsey, 2023; Klug, 2023).

Historical disaster responses demonstrate the value and limits of this approach. Following the Japan tsunami, accurate forecasting enabled rapid deployment but also produced surpluses of unsuitable goods (The New Humanitarian, 2012; Reuters, 2011). Humanitarian operations commonly rely on push logistics during the initial response phase when needs are uncertain and life-saving supplies must move quickly (Wisetjindawat et al., 2014). Early shipments typically include food, water and medical items, followed by more diverse requirements as assessments improve.

Push logistics also dominated the early Ukraine response. Prepositioned stockpiles and standardized aid packages facilitate rapid deployment of food, medical supplies and shelter materials (Balcik and Beamon, 2008; Tatham et al., 2017). Coordination among military and civilian responders during such deployments can introduce additional organizational and communication challenges (Heaslip and Barber, 2014). In Ukraine border regions, which initially had more predictable demand for essentials, the approach proved effective (Zinchenko, 2024). Yet mismatches soon emerged as needs diversified and weather and medical conditions changed. This is typical in humanitarian relief efforts (Jahre and Jensen, 2010; OCHA, 2015; Kovács and Spens, 2009). Experience from Haiti and Japan confirms that prepositioning must be paired with adaptive mechanisms to remain effective (Vohr, 2011; Reuters, 2011; Heaslip, 2013).

Pull logistics respond to actual demand as it arises, relying on real-time data and agile SCs (Dubey and Gunasekaran, 2016; Charles et al., 2010; Minculete and Olar, 2016; Manopiniwes and Irohara, 2017). Recent analytical work further illustrates this adaptability: Ozen and Krishnamurthy (2022) modeled dynamic relief-center networks that reallocate supplies in response to changing demand signals, a capability that parallels the digital pull mechanisms later exemplified by Korovai. This learner approach matches supply and demand more precisely and promotes flexibility through two-way communication between suppliers and beneficiaries (Zagursky and Slipukha, 2019). Modern militaries similarly employ pull systems to ensure timely, location-specific delivery (Ti and Kinsey, 2023).

As the Ukraine crisis evolved, pull logistics became increasingly essential. Digital platforms and data-sharing tools enhanced demand tracking and coordination (Charles et al., 2010; Arts et al., 2022; Becker et al., 2023). In heavily affected urban areas, local actors supplied detailed data on immediate needs, enabling international agencies to target aid more effectively (Heaslip and Barber, 2014; Holguín-Veras et al., 2012). Lessons from CARE International further stress the value of data-driven SCs (Duran et al., 2011; Tatham et al., 2017).

Although pull systems minimize waste, they can face delays when infrastructure is damaged or data are incomplete (Jahre and Jensen, 2010; Holguín-Veras et al., 2012). The Nepal earthquake response highlighted how such delays can exacerbate humanitarian suffering (OCHA, 2015; Yi and Özdamar, 2007). Accurate, timely information is thus critical to the success of demand-driven approaches (Holguín-Veras et al., 2012; Dubey and Gunasekaran, 2016; Pandey et al., 2024).

The strengths and weaknesses of push and pull strategies underscore the need for balance in HL, where unpredictability and urgency require both readiness and flexibility (Oloruntoba and Gray, 2006; Kovács and Tatham, 2009). Hybrid models integrate the rapid deployment of push systems with the responsiveness of pull systems, adapting as information improves (Schulz and Blecken, 2010; Upadhyay et al., 2022; Tatham and Houghton, 2011). The Ukraine response illustrates the complementary nature of both approaches and highlights the growing role of technology and collaboration in bridging the two – an evolution examined further in the following sections.

HL involves the movement and coordination of people, supplies and information under conditions of extreme uncertainty. Past large-scale disasters, such as the 2011 Japan tsunami and the 2015 Nepal earthquake, revealed persistent challenges in mobilizing resources and coordinating diverse actors across damaged infrastructure (OCHA, 2015; The New Humanitarian, 2012). Similar issues reappeared in the 2022 Ukraine response, where conflict disrupted SCs and generated massive humanitarian and defense-related demand (Zinchenko, 2024; UNHCR, 2023). These experiences highlight recurring themes – coordination, information flow, enabling technology, collaboration, trust and integration – that continue to shape HL and are equally relevant to SA operations.

Effective coordination remains the cornerstone of humanitarian operations. In most crises, the rapid influx of international donors and nongovernmental organizations (NGOs) overwhelms national and local capacities (Balcik et al., 2010; da Costa et al., 2014). Fragmented communication channels, redundant shipments and delayed needs assessments frequently undermine efficiency (Tatham and Pettit, 2010). The literature consistently identifies information asymmetry as a leading cause of duplication and resource waste (Kovács and Spens, 2009; Overstreet et al., 2011). HL studies, therefore, emphasize shared information systems and standardized reporting formats as prerequisites for coordinated response.

In Ukraine, as in earlier natural-disaster contexts, communication and coordination gaps among donors, host-nation agencies and implementing partners impeded the early flow of aid (Tart, 2023). These same coordination deficiencies motivated the eventual development of the Korovai platform, which sought to consolidate requests and commitments into a common digital environment.

Recent scholarship demonstrates that digital tools can significantly enhance visibility, transparency and adaptability in HL networks. Emerging technologies – including blockchain for traceability, artificial intelligence (AI)-based analytics and IoT-enabled monitoring – have improved decision-making and accountability (Pandey et al., 2024; Marić et al., 2022). Studies of humanitarian information platforms such as ReliefWeb, HDX and VOSOCC show how structured data exchange supports faster coordination and resource deployment (Moshtari et al., 2021; OCHA, 2015).

However, most systems remain designed for short-term disaster relief rather than sustained, high-security environments. Ukraine’s hybrid conflict required both military-grade data protection and humanitarian accessibility – demands that traditional platforms were not designed to meet. The literature thus underscores a gap in understanding how digital systems can simultaneously ensure security, interoperability and inclusivity, a challenge directly addressed by Korovai and its associated Sky Blue architecture.

Collaboration among humanitarian actors depends heavily on relational and cognitive trust. Trust reduces uncertainty, facilitates information sharing and encourages joint problem-solving (Heaslip and Barber, 2014; Hudnurkar et al., 2014). Multiple case studies have shown that when donors, military units and NGOs possess mutual confidence in each other’s intentions and data reliability, coordination improves markedly (Overstreet et al., 2011; Scholten and Schilder, 2015). Conversely, low trust often results in information withholding, fragmented planning and redundant logistics flows.

Building joint knowledge – defined as the shared understanding of goals, constraints and operational language – is equally critical (Hudnurkar et al., 2014). Joint knowledge creation typically occurs through regular information exchange, co-location and technological platforms that promote visibility and feedback. Korovai’s integrated dashboard and asynchronous update features later provided exactly this shared cognitive space, transforming dispersed donor activities into a more coherent logistics network.

Integration across procurement, transportation, warehousing and last-mile delivery enhances consistency and efficiency in humanitarian SCs (Oloruntoba and Gray, 2006). Optimized relief-center design is another critical element of such integration, determining how effectively prepositioned goods reach beneficiaries (Ozen and Krishnamurthy, 2018). Studies highlight the importance of connecting upstream donor decisions with downstream distribution outcomes through unified information systems (Pandey et al., 2024; Marić et al., 2022). In both humanitarian and defense logistics, integration ensures that supply allocation matches dynamically changing needs.

The literature also emphasizes interorganizational integration – the alignment of objectives, data standards and procedures across independent entities (Besiou and Van Wassenhove, 2020). Such alignment is difficult to sustain when organizations operate under different funding structures or political mandates. In Ukraine, the convergence of HL and SA processes magnified these complexities.

Figure 2 synthesizes insights from prior research and highlights the major thematic intersections between HL and defense-oriented SCs. Both domains rely on coordination, collaboration and trust as central enablers of effective operations. Information flow and enabling technologies form the connective infrastructure that links these domains, while integration ensures coherence across donors, governments and implementing partners. The areas of overlap represent opportunities for shared learning and system improvement, particularly in contexts like Ukraine, where military and humanitarian objectives converge.

The literature indicates that HL and SA share not only core principles but also persistent operational challenges. In both domains, efficiency depends on transparent communication, synchronized planning and reliable data exchange. What distinguishes Ukraine is the scale, duration and contested environment in which these logistics processes occur. Unlike short-term disaster responses, the Ukrainian operation has demanded prolonged coordination across military, governmental and civilian institutions.

While Ukraine’s SC centers on military aid rather than survival commodities, the same collaboration and information-flow mechanisms underpin its effectiveness. The overlap shown in Figure 2 underscores the importance of integrated digital coordination systems capable of operating across sectors with differing mandates. Lessons from HL therefore remain directly relevant to the analysis of Korovai’s design and implementation. These parallels justify extending HL concepts – trust, collaboration, information sharing and technological enablement – to the study of SA coordination during conflict.

The convergence of humanitarian and SA logistics in Ukraine underscores the need for analytical frameworks capable of explaining coordination across logistical, informational and financial domains within complex, multi-actor systems. Rather than treating this convergence as merely contextual, recent scholarship conceptualizes Ukraine’s defense supply system as a multilevel architecture spanning these interdependent domains, in which civilian actors function as structural intermediaries rather than peripheral contributors (Kushnir et al., 2024). In this framing, manufacturers, NGOs, donors and conscripts are embedded within an integrated network whose effectiveness depends not only on the movement of goods but also on the alignment of financial contributions and the integrity of information exchange.

Several digital and institutional frameworks have been developed to improve coordination in humanitarian operations, each reflecting distinct governance structures and technological capabilities. Table 1 compares representative examples (Sahana, HAP and HXL) with Korovai, highlighting their origins, governance models, core functionalities and known limitations.

As shown in Table 1, earlier platforms primarily addressed information standardization or accountability rather than real-time donor–recipient coordination, underscoring the need for adaptive systems such as Korovai. Unlike prior platforms that largely support decentralized humanitarian ecosystems and open data exchange, Korovai enabled real-time matching of Ukrainian operational demand signals with multinational donor commitments during an active armed conflict. Its donor-led control structure and integration within defense coordination channels allowed rapid synchronization under high-security conditions, but also depended on centralized authority, secure digital infrastructure and coalition compliance – features not uniformly present in more decentralized humanitarian settings.

The next section introduces DCT and IT as complementary perspectives for understanding how digital platforms like Korovai evolve under uncertainty while achieving cross-sector legitimacy and sustained coordination.

While previous research has emphasized the operational and technological enablers of coordination in HL, less attention has been paid to the organizational mechanisms that explain how coordination platforms evolve, adapt and gain legitimacy in complex environments. To address this gap, the present study draws on two complementary theoretical perspectives: DCT, which explains how organizations adapt and reconfigure resources under conditions of uncertainty (Teece, Pisano and Shuen, 1997) and IT, which examines how legitimacy and conformity to shared norms influence organizational behavior and technology adoption (DiMaggio and Powell, 1983; Scott, 2014). Together, these perspectives provide a dual lens for analyzing how the Korovai platform both enabled adaptive coordination and achieved acceptance across diverse institutional actors.

Viewed through this lens, digital coordination platforms such as Korovai can be interpreted as capability-building mechanisms that enhance collective responsiveness in volatile environments. Korovai exemplifies how a coalition can rapidly reconfigure coordination routines by replacing fragmented, spreadsheet-based communication with an integrated, real-time digital system. Through its ability to consolidate requests, commitments and delivery status across actors, Korovai allowed users to sense emerging needs, seize coordination opportunities and transform traditional processes into a dynamic, information-driven system. As such, DCT provides a useful framework for examining how digital platforms support adaptive coordination and organizational learning during humanitarian crises.

IT complements this adaptive view by highlighting how organizational structures and behaviors are shaped by normative, coercive and mimetic pressures within institutional environments (DiMaggio and Powell, 1983; Scott, 2014). In HL, organizations often adopt coordination mechanisms not only for efficiency but also to achieve legitimacy within a complex field of donors, governments and nongovernmental partners (Moshtari and Gonçalves, 2017). Legitimacy depends on the extent to which coordination tools align with shared expectations of transparency, accountability and interoperability across actors.

From this perspective, Korovai represents a field-level institutional response to competing pressures for coordination and control. Its development reflected coercive pressures from defense institutions for data security, normative expectations from humanitarian actors for inclusive collaboration and mimetic learning from existing platforms such as ReliefWeb and HDX that exemplified effective information sharing. By reconciling these institutional logics, Korovai achieved cross-sector acceptance and sustained participation among donor nations. IT thus helps explain not only how digital platforms like Korovai function but also why they achieve legitimacy and durability in complex, multi-actor humanitarian systems.

The preceding discussion highlights that while existing research has advanced understanding of collaboration and information flow in HL, much of it focuses on static or single-domain coordination mechanisms. Earlier platforms such as Sahana, HDX and ReliefWeb have demonstrated the potential of digital tools to improve information sharing, but they were primarily designed for peacetime disaster relief and lacked integration across defense and humanitarian actors. Few studies have examined how such platforms can evolve dynamically in armed-conflict contexts, where information security, interoperability and legitimacy pressures are far more pronounced.

Furthermore, prior research has explored collaboration factors influencing coordination effectiveness (Hudnurkar et al., 2014; Heaslip and Barber, 2014) and organizational agility (Besiou and Van Wassenhove, 2020). However, few empirical studies have examined how these relational and organizational insights translate into the design or performance of digital coordination systems. The literature also provides little evidence on how institutional pressures – such as defense security requirements, donor accountability norms and humanitarian transparency expectations – jointly shape the design and adoption of these systems. To date, no empirical study has explicitly examined donor-driven digital, pull-based platforms operating in high-security conflict environments or analyzed how dynamic capabilities and institutional pressures jointly influence their emergence, design and legitimacy.

This study addresses these gaps by analyzing the Korovai platform as a novel case of humanitarian–military coordination under active conflict. It applies DCT to explain how the platform enhanced adaptive coordination and IT to interpret how it gained legitimacy among diverse international partners. By combining these perspectives, the study contributes to HL theory by linking collaboration mechanisms, digital transformation and institutional alignment in complex, high-risk environments.

This study adopts a qualitative approach, using semistructured interviews and thematic analysis, appropriate for exploring the implementation of novel systems within complex humanitarian operations. This design allows for an in-depth understanding of interactions between actors, technologies and organizational processes that could not be captured through survey or archival data alone. Together, these approaches ensure that the investigation of the Korovai platform is both contextually grounded and analytically systematic, consistent with accepted standards in operations and HL research.

We used semistructured interviews to gather primary data. No collected information was classified at any level, and all of it could be published on open-source platforms. To preserve participant confidentiality, all interviews were conducted anonymously and the analysis excluded any quantitative or contextual details that might allow attribution to specific individuals, organizational entities or donor countries. All data and geolocation information used in the analysis were obtained from reputable open-source databases and cited appropriately. The interviewees ranged in position from the tactical to strategic level and were subject matter experts (SMEs) in their respective fields.

An exploratory sequential process guided the research. Preliminary document analysis provided background on international donor coordination and the evolution of the Korovai platform. This was followed by semistructured interviews with personnel engaged in the ECC-U and its successor, the IDCC. The interviews were triangulated with operational reports, open-source data and secondary analyses (e.g. Tart, 2023; Zinchenko, 2024; Ti and Kinsey, 2023) to enhance construct validity.

Interview questions were developed to gather expert insight on participants’ experiences. The questions used during the interviews are in  Appendix 1. Gathering expert knowledge was vital to ensuring that the quality of the primary information received was not compromised (Harrell and Bradley, 2009). Interviewees were coded P1–P13 in Table 2 to ensure anonymity. Interviews were semistructured to obtain vital details about Korovai and allow the interviewee to tell the story from their perspective (Harrell and Bradley, 2009). A semistructured interview allows for follow-up questions during the interview. These questions help to ensure that the information relayed is understood. Follow-up questions included:

The questions were developed to understand how and why, eliminate hearsay and elicit longer answers than “yes” or “no” (Harvard University, 2026). Information about the research objectives was communicated via e-mail and in person to enhance the data collection in the interviews.

The first questions posed to those interviewed allowed the interviewee to present the state of the Ukraine SA logistical operations from their perspective. These logistical processes described were all part of the Ukraine SA SC.

A purposive sampling strategy was used to identify participants with direct involvement in the coordination of humanitarian and defense logistics through the Korovai platform. Selection criteria included experience within the ECC-U or IDCC and active participation in logistics or information-sharing processes. Thirteen participants representing strategic, operational and technical functions were interviewed.

According to Participant 6 (P6), the ECC-U initially comprised 17 personnel. Two members of that original team were interviewed; both served in senior leadership and advisory roles, representing the perspectives held by the ECC-U at the escalation of the Russia–Ukraine conflict in February 2022. As operations expanded, the organization – later renamed the IDCC – grew to more than 200 personnel. Five interviewees were drawn from the IDCC’s current staff, all of whom held leadership or advisory positions and interacted regularly with the majority of IDCC members.

The interviewees’ experience with HL operations varied, although 11 of the 13 had participated directly in at least one HL operation during their careers. All but one were involved in logistical activities supporting the flow of aid to Ukraine; the remaining participant, affiliated with the US Agency for International Development, was included to provide a broader perspective on HL coordination. The interviewees’ involvement spanned from February 2022 to 2024, with engagement durations ranging from two weeks to one year and most participating for more than one month. This broad temporal coverage ensured that the full evolution of Korovai – and its impact on logistics support to Ukraine – was captured in the analysis.

We began by examining 28 CFs identified by Hudnurkar et al. (2014) as promoting SC collaboration, integration and information flow ( Appendix 3). Some of these CFs overlapped conceptually, for example, collaborative communication closely aligns with information sharing, while collaboration may also encompass information sharing or collaborative planning. Others, such as incentive alignment, reflected organizational motivations for monetary gain. Codes representing each CF were applied to the interview data to identify instances where participants referenced these factors.

Data analysis followed a thematic coding approach, which is well-suited for identifying patterns in qualitative data (Kunz, 2019). Open coding was first used to identify recurring concepts within the transcripts, followed by axial coding to group related codes into broader categories corresponding to established collaboration constructs such as trust, information sharing, joint knowledge and enabling technology (Hudnurkar et al., 2014). In line with guidance from Braun and Clarke (2006), a threshold of one-third of respondents was used to determine whether a recurring topic qualified as a key theme.

The semistructured interviews generated extensive qualitative data through open-ended questioning, which required systematic organization and interpretation. Thematic analysis provided a means of identifying meaningful patterns linked to the research questions (Braun and Clarke, 2006; Peel, 2020). Descriptive coding was used to capture words and phrases corresponding to the CF definitions and key constructs from the literature review. This approach “identif[ies] a role, process, action, place, or something narrative to describe the unit of data” (Castleberry and Nolen, 2018). The resulting coding framework is summarized in Table 3, with theme frequencies compared before and after the implementation of Korovai by the ECC-U. Graphs and tables were then developed to support the interpretation of thematic patterns. Interpreting results beyond simple description is essential to ensure analytical depth and theoretical coherence (Maguire and Delahunt, 2017).

Thematic saturation was used as the benchmark for data sufficiency. Interviews were continued until no new themes or substantive insights emerged – a condition indicating that additional data would yield diminishing returns (Guest et al., 2006). The final interviews largely reinforced earlier categories, suggesting that the thematic structure was both stable and comprehensive.

The research triangulates the data from the interviewees with information gathered during the literature review, as shown in Figure 3. Triangulation provides a method to prevent errors, increases the quality of research and aids in providing validity and reliability to conclusions (Harrison et al., 2017; Bruns, 1989).

Finally, ethical procedures were followed by providing interview questions and objectives to the Air Force Institute of Technology Human Research Protection Program (HRPP) office. The Institutional Review Board granted an HRPP exemption, which is in  Appendix 2.

The interviews collectively produced a narrative describing how Korovai evolved from an improvised response mechanism into a progressively capable digital coordination platform. Participants traced the platform’s emergence as new features were introduced, refined and institutionalized across the multinational aid effort. The findings are organized into five stages:

1. coordination challenges prior to Korovai;

2. the system’s conceptualization and development;

3. perspectives from coordination practitioners;

4. secure collaboration architecture; and

5. digital integration and predictive analytics capabilities.

The USA and partner nations began delivering SA on February 24, 2022. In the first few months, more than 66 million tons of ammunition, vehicles, food, water, tires and other supplies were transported through a globally dispersed SC (Tart, 2023). Financial flows mirrored the logistical scale. Between 2022 and 2024, donor governments collectively committed more than $90bn in military and humanitarian assistance to Ukraine, averaging over $5bn per month (Masters and Merrow, 2024). This extraordinary volume of resources intensified the coordination challenges faced by the ECC-U, amplifying the need for a unified digital system to manage contributions and avoid duplication. Establishing and managing such a worldwide supply network is inherently complex and prone to disorder (Day, 2014). The urgency of the Ukrainian crisis, however, left little time for deliberate coordination; rapid assistance was essential to prevent Ukraine’s defenses from collapsing and to deny Russia its strategic objectives (Becker et al., 2023). As is often the case in war, logistical challenges emerged early in the response, a reality that was described by personnel operating both on the ground and within coordination centers during the first months of 2022.

The initial effort was multimodal, involving aircraft, ships, trains and vehicles from across the globe to deliver aid to Ukraine’s borders (P1). Aircraft frequently arrived at airfields without prior notice to ground support personnel (P9), and the contents of many shipments remained unknown until the cargo doors were opened (P9). In addition, customs documentation was often incomplete or inaccurate, further delaying offloading operations (P11). This absence of information forced ground crews to adopt a reactive posture, significantly constraining throughput capacity (P9).

An Air Force Contingency Response Element (CRE) was positioned at key airfields critical to sustaining the SC flow. CRE personnel observed the coordination problems described above and developed creative ad hoc solutions to maintain logistical movement from the airfields to the front lines (P9). A tent at one of these airfields became an informal coordination hub where liaisons from various elements of the SA effort met to synchronize operations (P9). Excel spreadsheets were used to record shipment information and distribute updates across the network (P6). Senior staff and additional liaisons participated remotely through phone and other communication channels (P9).

Recognizing the growing complexity of operations and the need for structured coordination, senior military leaders established a formal logistics coordination cell to align donor activities with Ukrainian requirements, which became increasingly specific as the conflict evolved (P6). This cell, named the ECC-U, consisted of logistics SMEs from the US Department of Defense (DoD), the UK, Ukraine and several other partner nations (Morrison, 2022). Liaisons from contributing countries were also assigned to the ECC-U. In its initial phase, multiple coordination meetings occurred daily, during which requirements and commitments were manually matched and entered into spreadsheets (P6). The leadership of this cell subsequently mapped the SA information flow, as depicted in Figure 4.

The coordination process relied on the exchange of information through e-mails with attached Excel spreadsheets, in-person or virtual meetings and telephone communications (P7). Updates to requirements or commitments typically triggered additional rounds of e-mails and phone calls (P6). As illustrated in Figure 4, this information flow contained multiple points of potential failure, increasing the likelihood of data loss and miscommunication. For instance, if a country rescinded a transportation commitment, the ECC-U was required to re-coordinate logistics across the coalition. Each such adjustment necessitated further deconfliction meetings and additional correspondence (P6).

Over time, the ECC-U evolved into both a coordination and knowledge center for all SA logistics movements. However, Ukraine’s status as an active conflict zone meant that requirements were continually shifting. The fog and friction of warfare ensured that demand was unpredictable and frequently changing. This volatility created a persistent mismatch between the ECC-U’s communication processes and the rapidly evolving operational environment (P7). As requests became more specific and time-sensitive, the existing coordination model proved inadequate. The SC, therefore, needed to transition toward a more agile, pull-based logistics system capable of responding dynamically to real-time requirements (P6 and P7).

A few weeks into the initial response, ECC-U personnel proposed a digital coordination tool modeled on a wedding registry (P6 and P7). The analogy captured the principle of avoiding redundancy: just as a registry enables guests to view and fulfill specific needs without duplication, the proposed system would allow donors to view Ukraine’s logistical requirements and commit to supplying them (P6). The team subsequently outlined the process shown in Figure 5, which supported a pull-based logistics model grounded in transparency and asynchronous coordination.

To ensure the system’s resilience, ECC-U staff emphasized that it needed to function asynchronously, allowing real-time updates and decentralized access (P6). Senior leadership at US European Command engaged the Defense Digital Service (DDS), which specializes in rapid-response digital solutions for the DoD (Morrison, 2022). The UK’s equivalent digital service was also enlisted to support the design and implementation (P6, P7 and P2). Both organizations were briefed on the operational challenges and the overarching goal of developing a centralized digital coordination platform. The software company Atlassian was then contracted to assist with development (Morrison, 2022). Within weeks, the resulting digital system was launched and named Korovai, after the Ukrainian word for wedding bread, symbolizing unity and cooperation (P6, P7, P1 and P2).

Through Korovai, Ukrainian representatives could log in and list prioritized requirements – such as mortars, ammunition, food and water (P1, P2, P3, P5, P6 and P7). These requests were reviewed and aggregated by Ukraine’s General Staff according to operational priorities (P6). Partner nations and donor organizations could then view the prioritized list, commit to providing specific items and thereby remove those requirements from the active registry (P1, P2, P3, P5, P6 and P7). This mechanism significantly reduced redundancy across the multinational logistics effort.

A key feature of Korovai was its ability to treat transportation as a separate commitment (P6, P7 and P1). Donor nations could commit to either providing the goods, the transportation or both. When a nation could not deliver the goods it supplied, another could commit to transporting them safely (P6, P7 and P2). This modular structure improved coordination efficiency and reduced duplication of effort across the coalition.

Korovai’s shared visibility enabled all authorized users to track changes to requests and identify logistical disruptions in real time. This enhanced information flow allowed ECC-U personnel to resolve operational issues more rapidly (P2, P6 and P7). Increased visibility also ensured that critical equipment reached Ukrainian forces on time and allowed the coalition to deliver SA more effectively. The system introduced accountability by recording every donor commitment and fulfillment (P2) and consolidated all SA tracking into a single digital platform, replacing multiple spreadsheets that often contained incomplete data (P2, P6 and P7).

The system’s value was quickly demonstrated through an operational example: a vehicle donated to Ukraine became inoperable due to maintenance issues, and replacement parts were available only from the donating nation (P2). Using Korovai’s tracking functions, ECC-U staff identified the donor and obtained the required parts promptly, restoring a vital asset to operational status. This case underscored how Korovai’s information management capabilities contributed directly to maintaining logistical readiness in the ongoing conflict.

Information related to operational requirements during armed conflict must be protected, with access restricted to authorized users. At the same time, the coalition supporting Ukraine required an environment that enabled transparent information sharing among partner nations. Without such collaboration, logistical efficiency would suffer; however, open networks risked exploitation by malicious actors seeking to disrupt Ukraine’s defense or undermine coalition efforts. In addition, some donors wished to contribute anonymously to avoid public disclosure (P2, P6 and P7).

To address these competing needs, the DDS developed a security framework that preserved Korovai’s collaborative functionality while ensuring data protection (Morrison, 2022). A US Government contractor, Sky Blue, was engaged to implement this solution. Sky Blue provides a secure but unclassified environment that enables approved users to collaborate through Korovai and a suite of integrated applications, including Microsoft Teams, Confluence, Qlik and Power BI (Sky Blue, 2023). The resulting architecture is illustrated in Figure 6.

Sky Blue can be accessed from any laptop or mobile device and uses multi-factor authentication to maintain security, even in austere operational settings (Sky Blue, 2023). This capability allows field personnel to monitor the fulfillment of their requests in real time, while critical updates flow simultaneously to Ukrainian military leadership. Logistical SMEs and international liaisons at the IDCC in Germany can thus coordinate asynchronously within the secure Sky Blue environment (Sky Blue, 2023). The continuous information flow between field units and decision-makers enhances situational awareness and supports timely, informed decisions.

A further advantage of Sky Blue lies in its flexibility: new partner nations or donor organizations can be integrated within hours or days rather than weeks or months (Sky Blue, 2023). This agility has proven essential for maintaining a responsive multinational logistics network and sustaining the coalition’s capacity to adapt to evolving operational requirements.

The IDCC ultimately coordinated a single-recipient SC serving Ukraine through multiple airports and seaports, with onward movement by air, rail and road (P8 and P9). Early in the response, political sensitivities complicated cooperation, as the European Union, the USA and the UK initially relied on separate systems (P6). The adoption of Korovai as the common program of record unified these efforts, enabled Ukraine to assume a leadership role in prioritizing requests and reduced the risk of donors losing political capital (P6).

Sky Blue has played a pivotal role in advancing Korovai’s functionality. It links the platform to the US DoD database Advana, which aggregates operational data and presents it through interactive digital dashboards for users and decision-makers (Sky Blue, 2023).

Within the Sky Blue environment, a machine translation capability allows rapid translation of technical materials, a critical function in multinational logistics operations where language barriers can impede coordination (P3, P4 and P5). For example, accurate translation of aircraft loading and unloading procedures and equipment handling guidelines is essential for safety. The translation function enables manuals and operational documents to be converted efficiently between languages, reducing errors and enhancing interoperability when equipment from one nation transits through another’s transport network (P3, P4 and P5).

Sky Blue also integrates large language models and other AI tools that can analyze data drawn from Korovai and Advana to forecast emerging SA requirements for Ukraine (Sky Blue, 2023). These models facilitate predictive analytics for logistics planning, life-cycle assessment and end-use analysis (Sky Blue, 2023).

Predictive capability is vital in ensuring that Ukraine remains prepared for future operational needs. Sustained military operations depend on supplies arriving at the correct location and time, yet communication challenges in contested environments often obscure demand signals. Reliable forecasting reduces this uncertainty, thereby lowering mission risk. The integration of Korovai within the Sky Blue ecosystem offers a model for any global SC, enabling the rapid establishment of an information-sharing and analytics hub. What traditionally requires months or years to develop can now be configured within days, providing both process efficiency and strategic agility (P4 and P5).

Seven CFs were found to be a common theme across the 13 interviews: trust, relationship promoter (liaison), collaborative planning, information sharing, joint knowledge, integrated processes and enabling technology. These seven CFs were mentioned by at least one-third of the interviewees, a requirement set to identify key themes from the interview data.

The frequency and direction of sentiment related to CFs in the interview responses were recorded. The top and middle portions of Figure 7 show the score for each CF before and after, respectively, Korovai became the system used by the ECC-U. A comment was scored as negative when it had a negative connotation. For example, one interviewee stated that upon arriving in the theater:

Another stated, “There was no system or process in place to facilitate communication between Ukraine and donors.” Both comments were scored negatively for the “integrated process” CF. The acknowledgement of negative sentiment in the coding allows for a more careful quantification of the difference between sentiment before and after Korovai. Both positive and negative overall sentiment were expressed in the responses.

The before sentiment scores shown in Figure 7 indicate that the strongest positive sentiment toward collaboration on information flow was related to liaisons and information sharing. These CFs align with the narrative presented in Section 4. There were planning sessions and liaisons at the ECC-U headquarters and certain airfields. Those at the operational and tactical level tuned into these meetings and recounted the presence of liaisons and information being shared via phone and e-mail. The lack of a shared platform to obtain a demand signal from the user mirrors a focus on a push logistical approach. This approach requires very little collaborative planning since prediction drives requirements rather than a demand signal.

The sentiment scores of responses after the implementation of Korovai, shown in Figure 7, indicate a significant increase in sentiment toward almost all the CFs, suggesting that Korovai positively impacted collaboration and information flow. This interpretation is supported by the sentiments of nine of the 13 interviewees. They commented that using a shared digital interface increased the SA enterprise’s efficiency and effectiveness. These are both objectives for SA and HL operations, as described in Figure 2.

Sentiment related to information sharing and collaborative planning improved from an already positive position. Sentiment related to enabling technology improved from neutral to strongly positive after the Korovai introduction. Most dramatically, sentiment about integrated processes, trust and joint knowledge, which were negative before Korovai, improved to strongly positive among responses after Korovai was implemented. Communication must occur to create trust between two entities. A pull system design forces logistical players to communicate. Also, trust increases cohesion and collaboration within an organization (Delbufalo, 2012). All sentiment toward CFs improved after the implementation, except in the case of liaison, which was neutral. The overall positive trend is consistent with Korovai playing a significant role in establishing a more demand-driven process.

The findings from this study can be interpreted through two complementary theoretical lenses – DCT and IT – each offering distinct insights into how the Korovai platform reshaped humanitarian coordination practices. DCT highlights Korovai’s adaptive capacity under extreme operational volatility, while IT explains how the platform achieved legitimacy and broad adoption within a complex network of defense and humanitarian actors. Examining the platform through both perspectives enriches our understanding of how technological innovation and organizational alignment jointly determine effectiveness in HL.

From a dynamic capabilities perspective, Korovai represents an advanced manifestation of adaptive coordination in HL. Earlier platforms such as VOSOCC, HDX and Sahana primarily supported information dissemination and situational awareness, but their flexibility was limited to predefined reporting structures. Korovai, by contrast, embedded sensing, seizing and transforming functions directly into its operational logic. It enabled the coalition to sense evolving requirements in real time through a centralized request-commitment interface, seize coordination opportunities by reallocating resources dynamically and transform legacy coordination routines from fragmented e-mail chains into automated data exchanges. The integration of Sky Blue and Advana further enhanced this adaptive capability, introducing translation, predictive analytics and asynchronous collaboration that collectively strengthened the coalition’s agility under crisis conditions. Through DCT, Korovai thus demonstrates how humanitarian networks can reconfigure technological and organizational resources to maintain responsiveness in volatile operational environments.

Viewed through an IT lens, Korovai also functions as a legitimacy-building mechanism within the broader humanitarian–military field. Whereas earlier systems emerged within distinct institutional domains – ReliefWeb under the UN or HDX under OCHA – Korovai evolved at the intersection of defense and humanitarian norms. Its design reconciled coercive pressures for data security from military stakeholders with normative expectations of transparency and inclusivity from humanitarian partners. Adoption across multiple donor nations reflects both mimetic learning from established coordination models and the creation of new shared standards for digital governance in coalition operations. In this sense, Korovai extends beyond technological innovation: it represents the institutionalization of a trusted, cross-sector coordination framework that aligns diverse actors under a common set of rules and accountability practices.

Traditional humanitarian operations often delegate coordination authority to multilateral bodies such as the United Nations Office for the Coordination of Humanitarian Affairs (UNOCHA, 2026) rather than to the affected state. Korovai inverted this pattern by enabling Ukraine to lead coordination within a multinational donor coalition. This arrangement enhanced legitimacy through alignment of coercive, normative and mimetic pressures – an outcome consistent with IT’s emphasis on conformity to shared expectations across organizations.

Together, these theoretical perspectives underscore both the adaptive and legitimizing functions of digital coordination platforms in HL. For practitioners, Korovai illustrates how dynamic capabilities – specifically, rapid sensing of needs, reallocation of resources and transformation of coordination processes – can be operationalized through digital infrastructures. At the same time, the platform’s success depended on institutional alignment, balancing security, transparency and interoperability across diverse national and organizational mandates. Future coordination platforms can build on this dual foundation by integrating predictive analytics and user-driven design with governance structures that reinforce trust and accountability. In doing so, they can enhance both the agility and the legitimacy of humanitarian response systems in increasingly complex and contested environments.

However, the transferability of Korovai’s principles depends on specific enabling conditions. These include the presence of a recognized coordination authority, the availability of secure and interoperable digital infrastructure, donor commitment to a shared platform and adequate technical capacity among users. In settings where governance is fragmented, digital access is uneven or trust among actors is weak, implementation may face structural constraints. Moreover, reliance on centralized digital systems introduces tradeoffs, including concerns about data sovereignty, dependence on donor-controlled technology, cybersecurity exposure and the potential exclusion of actors lacking secure digital access. Recognizing these boundary conditions clarifies that Korovai represents a context-dependent coordination model rather than a universally transferable solution.

This study is based on qualitative data from a targeted sample of 13 participants drawn from the ECC-U and the IDCC. While the interviews provided deep operational insight, they do not represent every donor nation using Korovai. Several participants were US or allied defense professionals, which may shape their perceptions of coordination effectiveness and governance. Future research should therefore examine perspectives from a broader range of donor and recipient organizations to balance potential bias.

Because access to Ukrainian counterparts was limited for operational-security reasons, the study relied on secondary and open-source documentation to validate findings. Consistency among interviewees and triangulation with official after-action reports and open data support the credibility of the results, yet the evidence remains interpretive rather than generalizable. The analysis assumes that participants accurately described processes and system performance.

The research did not quantify measurable performance indicators such as lead time reduction, duplication avoidance or fulfillment rates. Interviewees and ECC-U documentation consistently described qualitative improvements in these areas, but systematic data collection was beyond the study’s scope. Future mixed-method studies could pair qualitative perspectives with digital platform log data to confirm the magnitude of efficiency gains.

Another limitation concerns the absence of cost and risk assessment. The analysis focused on coordination benefits but did not examine platform-development or training costs, cybersecurity exposure or long-term maintenance requirements. These aspects warrant dedicated evaluation, as several participants noted that continual system upgrades and user training were essential to sustaining performance.

Finally, the findings are context-specific. Korovai was developed within a SA framework rather than a purely humanitarian operation. While many coordination mechanisms – information sharing, demand signaling and trust building – are transferable to disaster-relief logistics, extrapolation to nonmilitary contexts must be cautious. The results should therefore be interpreted as conceptually transferable rather than statistically generalizable.

This study examined how Korovai, a multinational digital-coordination platform, enhanced information flow and collaboration in Ukraine’s SA logistics. The thematic analysis identified seven collaboration factors – trust, liaison relationships, collaborative planning, information sharing, joint knowledge, integrated processes and enabling technology – that improved markedly after the platform’s implementation. Together, they demonstrate that digital infrastructures can transform complex supply networks into responsive, demand-driven systems even under conflict conditions.

From a theoretical standpoint, the results illustrate how DCT and IT jointly explain Korovai’s evolution. Through DCT’s lens, the platform enabled coalition actors to sense emerging requirements, seize coordination opportunities and reconfigure legacy routines into adaptive, data-driven processes. IT clarifies how Korovai gained legitimacy by reconciling coercive pressures for data security, normative expectations for transparency and mimetic learning from existing humanitarian platforms such as ReliefWeb and HDX. The system’s acceptance across more than 30 donor nations reflects both adaptive capacity and institutional alignment – critical to sustaining collaboration in high-risk environments.

Beyond theoretical contribution, the findings have practical and policy relevance. Korovai demonstrates how shared digital environments can reduce redundancy, accelerate matching between needs and resources and promote accountability through transparent tracking. Comparable systems could benefit United Nations cluster coordination, large NGO consortia and government-to-government aid mechanisms. However, implementation requires clear data-governance standards addressing cybersecurity, data sovereignty and equitable access to technology to avoid reinforcing the digital divide. Establishing these governance frameworks should accompany any broader adoption of digital coordination platforms.

Future investigations should extend this work in three directions.

First, quantitative validation is needed to measure Korovai’s impact on coordination performance. Research combining platform-usage data with logistics metrics – such as average time from request to fulfillment, number of duplicate shipments prevented or percentage of unmet requests resolved – would provide empirical evidence of efficiency gains.

Second, comparative studies should evaluate Korovai alongside other humanitarian information systems such as Sahana, HAP, HXL and ReliefWeb to assess scalability, interoperability and governance structures. Such analyses could identify the design principles that enable integration across defense and humanitarian domains.

Third, future work should address resilience and risk. Incorporating cybersecurity, cost and training-burden assessments will clarify the sustainability of digital coordination in austere environments. Additional attention should be given to hybrid push–pull configurations: Korovai’s integration with the Advana database already hints at a model that can switch from pull to push when communication breaks down, thereby maintaining supply-chain continuity. Studying this hybrid functionality in both conflict and disaster contexts could yield broader insights into resilient humanitarian-logistics design.

Finally, research on transferability to complex emergencies – where conflict and natural disasters overlap – would deepen understanding of how digital coordination platforms operate amid fluctuating institutional mandates. Comparative fieldwork in contexts such as Syria, Yemen or postdisaster recovery zones could reveal how political, cultural and infrastructural differences affect adoption.

In sum, Korovai illustrates how digital transformation can strengthen humanitarian and SA logistics by coupling adaptive technological capability with institutional legitimacy. Continued interdisciplinary research combining operations analysis, organizational theory and information-systems evaluation will be vital to translating these lessons into sustainable practice across the humanitarian domain.

1. What was the main problem or problems that were occurring with the security assistance logistical operations in Ukraine when you first arrived?

2. What was the response to remedy these problems at the operational and strategic levels?

3. How effective were the responses/solutions?

4. In your opinion, what was the metric being used to define success?

5. In your opinion, what are the key factors for effective communications during a crisis event where you have multiple organizations or countries coming together?

6. In your opinion, what leads to a clear and concise command relationship(s) in the supply chain during a time of crisis?

7. In your opinion, what are the most common causal factors for backlogged cargo during the ongoing Ukraine–Russia conflict?

a. Are these similar to or different than causal factors one may face during a time of crisis, such as a natural disaster?

8. In your opinion, how do you measure success during crisis operations?

a. Is there a specific metric that is often used?

b. Is it different between agencies and/or countries?