Optimisation of Inventory Operations
Proposal for a Data-Driven Replenishment Concept
Following the liberalisation of the Swedish pharmacy market, a data-driven concept for optimising inventory management was developed for a major pharmacy chain. Growing competitive pressure from private providers led to declining margins, necessitating a fundamental redesign of existing replenishment and inventory strategies.
The project aimed to design a robust, data-driven replenishment concept that reduces inventory while maintaining a high level of service to end customers. Existing replenishment practices were analysed using real demand and inventory data, and alternative approaches were developed and validated through simulation.
Initial Context
The project was initiated in the context of the liberalisation of the Swedish pharmacy market, which abolished the state pharmacy monopoly and enabled the entry of private providers. As a result, new competitors entered the market, leading to increased price pressure and declining margins for established providers. Against this backdrop, optimising internal cost structures became a key strategic priority, particularly in merchandise disposition.
Inventory management was identified as a central lever, as it significantly affects both capital commitment and delivery capability to end customers. Real historical inventory and demand data from a major Scandinavian pharmacy chain served as the empirical basis, enabling a quantitative analysis of existing disposition practices.
The project was conducted as part of a university consulting role-play. The underlying problem statement, data basis, and the structure of the Request for Proposal were modelled on a real-world project from the trade and distribution sector. The objective was to develop a well-founded proposal for a data-driven replenishment concept under these conditions and to quantitatively assess its impact.
Approach and Methods
-
The project was initiated as part of a university consulting role-play, which simulated a realistic tendering and competitive environment. Students, organised into groups, acted as independent consulting firms, competing for a project contract within a structured RFP process.
At the outset, the client presented the initial situation, including key target metrics such as the desired service level, perceived cost drivers, and existing planning practices. In addition, a structured Q&A session provided the opportunity to clarify operational, organisational, and economic conditions, for instance regarding demand patterns, branch structure, personnel costs, or existing ordering processes.
This phase primarily served to precisely define the problem, clarify assumptions, and establish the project focus, thereby forming the methodological basis for the subsequent development of a robust solution proposal.
-
Building on the insights from the RFP phase, a data-driven replenishment concept was developed and formalised in a structured proposal. Relevant literature on inventory management, ordering policies, and service level control was analysed and evaluated with respect to its applicability in the given context.
Based on these insights, a proprietary replenishment concept was designed, relying on the data-driven determination of reorder points and order quantities and enabling automated daily order recommendations at branch level. The objective was to replace existing manual inventory management, reduce personnel effort, and systematically address the inherent trade-off between inventory levels and delivery capability.
Beyond the conceptual design, the proposal also included a quantitative estimation of the expected effects, a structured representation of operational implementation, and an economic evaluation. This provided a robust basis for decision-making regarding the selection and further development of the concept.
-
Following the proposal phase, the developed replenishment concept was validated through a simulation-based proof of concept. The objective of this phase was to quantitatively assess the concept’s effectiveness and benchmark it against historical replenishment practices.
For this purpose, a simulation environment was implemented in Python, replicating historical system behaviour based on real demand and inventory data. Both the developed replenishment concept and the historical inventory management logic were exposed to identical demand patterns, enabling a consistent A/B comparison.
The simulation enabled a systematic analysis of key performance indicators, including service level, inventory levels, and ordering behaviour. The results of the proof of concept formed the basis for evaluating the developed approach and were subsequently presented and discussed within the context of the consulting role-play.
Solution Concept
The solution concept describes a replenishment system designed as a SaaS solution to support inventory planning. The objective was to develop a standardised and scalable concept that can be integrated into existing ordering processes and enables automated order recommendations. The data-driven determination of reorder points and order quantities forms the basis for optimising inventory levels and delivery performance.
-
The replenishment concept is based on an ABC classification of the product assortment. Tailored ordering strategies are applied to different product segments in order to reduce inventory levels while maintaining a high level of delivery capability.
As each branch operates with its own local inventory, replenishment decisions are made on a branch-specific basis. The underlying decision logic is uniform across all branches, while site-specific parameters such as demand patterns and delivery costs vary.
This separation of decision logic and parameterisation enables the concept to be consistently scaled across large branch networks without the need for branch-specific customised solutions.
-
The introduction of the replenishment concept was designed as a phased implementation approach to minimise risks and to assess its impact in a controlled manner. The initial step consisted of a simulation-based proof-of-concept phase, in which the concept was validated using historical data.
Building on this, a pilot phase in selected branches was planned to evaluate practical feasibility and to identify necessary operational adjustments. Following successful piloting, the concept was intended to be gradually rolled out across the entire branch network.
Due to the standardised disposition logic and the parametrised configuration at branch level, the roll-out is scalable and can be implemented with manageable organisational effort.
-
The pricing and revenue model followed a value-based approach. The decisive factor was not implementation effort, but the economic value generated through reduced inventory levels, lower process costs, and improved delivery capability.
The initial project phase was designed so that the investment would amortise within one year. Ongoing operation, support, and further development of the replenishment system were covered by recurring licence and support agreements.
The model thus combines a short payback period with long-term stability, enabling low-risk implementation and sustainable operation of the concept.
Outcome and Impact
A well-founded Proof of Concept providing a clear business case for efficiency gains and inventory optimisation
As part of the Proof of Concept, the developed replenishment concept was evaluated against historical replenishment practices using simulation. Both scenarios were based on identical demand and item data and differed solely in the applied replenishment logic. The developed concept demonstrated the following effects:
Stabilisation of the service level at defined business targets
Reduction in average inventory levels
Lower inventory volatility through more consistent order quantities
Reduced operational effort through automated replenishment logic
The results demonstrate that a data-driven replenishment concept can generate substantial economic benefits while significantly reducing operational effort in replenishment. The Proof of Concept provided not only quantitative evidence of impact but also a robust basis for management and investment decisions. The project illustrates how a coherent, data-driven proposal with measurable economic value can be derived from a clearly defined problem statement and real operational data.