As the market entry happened just recently, the client does not yet have a lot of historical data on which forecasts could be based. The paper approaches two challenges. First, it identifies parameters which may influence the forecasting replenishment process. Secondly, it tries to identify feasible forecasting and replenishment models which may be implemented by the client.

The paper is split in two parts. First, the different business models are introduced and potential influencing factors were identified. This was done through a secondary literature research and by analysing information the author received from the client. The parameters were then assessed via a questionnaire by three key stakeholders who are involved in the process. This allowed to identify key parameters which could be used for the course of the paper. The second part of the paper is concerned with the identification of feasible forecasting and replenishment models based on the knowledge from the first part. This is done by a literature research and a procedure of exclusion based on the constraint of limited historical data availability.

In the first part, the paper created transparency of the various parameters which had been identified by the author. Based on the outcome of the questionnaire, the paper identifies three possible forecasting techniques which may be used by the client for demand forecasting. Those methods are the moving average method, the weighted moving average method, and the single exponential smoothing technique. Furthermore, the topics of ABC analysis, service level, min/max. inventory and safety stock have been introduced and analysed as those were considered collectively as important by the respondents of the questionnaire. Through the analysis within this paper, transparency was created towards different influencing factors for forecasting and replenishment. Furthermore, concrete applications have been introduced and a possible implementation plan was proposed by the author. Nevertheless, it has to be stated that the limited historical data at hand still remains a major challenge for the client. Therefore, the models have to be thoroughly tested before being implemented. Furthermore, the nature of the models does only allow for a baseline forecast by smoothing out major variants in past demand. Therefore, they can only be considered as an addition to the currently performed qualitative forecasting.