Područje koje se bavi malim i srednjim poduzećima i njihovom informatizacijom je konstantno aktualno zbog značaja te skupine poduzeća za gospodarstvo cijelog društva. Ključni problem kod te kategorije poduzeća predstavlja nedostatak znanja, vremena i resursa za provođenje kvalitetne informatizacije. Brojna istraživanja potvrđuju da loše provedena informatizacija rezultira nezadovoljstvom i smanjenjem efikasnosti poslovanja poduzeća. Doktorat obrađuje početne korake pri informatizaciji poduzeća, odabir programskog rješenja za podršku poslovanju i radnje koje tom činu prethode: procjenu izvodljivosti i učinaka uvođenja programskog rješenja na poslovanje poduzeća, analizu i procjenu projektnih rizika i specifikaciju potrebnih funkcionalnosti programskog rješenja. Opisan je i formaliziran vlastiti model za predviđanje učinaka uvođenja programskog rješenja u poduzeće temeljen na sustavu podržanom znanjem koji uključuje i procjenu sklonosti promjenama u poduzeću. Formiran je skup parametara za opis informatičkih projekata. Predstavljena je vlastita metodologija za opis stanja u poduzeću i formiranje funkcionalnih zahtjeva za novo programsko rješenje, koja povezuje klasične metode višekriterijskog odlučivanja s modelom dubinske dekompozicije, koji sprema znanje o organizacijskim promjenama u poduzeću. Postupak prikupljanja ekspertnog znanja u svrhu predviđanja učinaka informatizacije na poduzeće i formiranje baze znanja o organizacijskim promjenama u svrhu definiranja specifikacije traženih funkcionalnosti novog programskog rješenja su ilustrirani prototipom razvijenim nad predloženom metodologijom. Valjanost metode je verificirana na nekoliko studijskih slučajeva, korištenjem dubinske dekompozicije poslovnih procesa u poduzeću. Main aim of this study is to aid Small and Medium Enterprises (SME) during the process of selecting appropriate Enterprise Resource Planning (ERP) System. Today's companies, independently on their size or industry vertical, use some kind of information system, both for automation of their own business processes and for communication with external partners. A well-implemented ERP system increases the company's overall efficiency. A third of companies experience an increase in efficiency greater than 20 percent after having implemented an ERP system. On the other hand, a poorly selected or implemented ERP can burden the company with unnecessary expenses, cause inefficient use of employees' time, create bottlenecks in business processes execution and, consequently, reduce the company's overall efficiency. Negative effects of ERP failure increase with the company's size. Two percent of large companies face bankruptcy as the result of poor ERP implementation. Common negative effects of ERP implementation in SMEs, often not anticipated at the project beginning, include initial drop in productivity and increased demands on employees, who must take part in implementation in addition to their usual work, organizational changes due to changes in business processes, with the resulting stressful environment and resistance to change. According to research, a large number of SMEs in Croatia are unsatisfied with their current information system. Improving business efficiency of SMEs is a permanently attractive issue, as it involves a group of companies that comprises a significant portion of almost any country's economy. However, SMEs in general seem ill prepared for the task of selecting and implementing a new ERP. Decision makers usually have insufficient knowledge, time or resources to make a qualified selection, and to make a successful implementation. Almost a half of ERP implementation projects into SMEs are never completed, with many of them written off as failures. This dissertation gives an overview of the steps that need to be taken when choosing an information system, from assessment of the current situation (description of as-is state), estimation project risks and feasibility, to specification of the new ERP's required functionalities. A proprietary model for predicting effects of new ERP implementation, supported by a knowledge base that includes a parameter set for storing knowledge and a software prototype that enables interaction with end users is herein presented and formally described. A proprietary methodology for describing the as-is state and automated generation of functional requirements for the new ERP, named In-depth decomposition model (DDM), supported by the Organizational Growth Algorithm, is also presented in this dissertation. DDM is used to store knowledge about organizational characteristics of companies in different industry verticals, namely about how their organizational structure changes with increase in number of employees. Organizational growth algorithm collects information about the end user's company and, by applying the knowledge stored in the DDM, automatically generates the company's organizational structure, its business process model and specification of functional requirements for the new ERP. In order to demonstrate practical application of the presented methodology, a software prototype with a graphical user-interface has been developed. After the introductory section, second and third chapter of this paper define main terms related to the field of software implementation, including change management, business process change, feasibility calculation, project risk factors, etc. Since ERP selection is in most cases a multi-criteria decision-making (MCDM) problem, common MCDM methodologies are analyzed. An introductory overview of the developed methodology with its conceptual and logical models is given in the fourth chapter. Fifth chapter defines the concept of susceptibility of an organization to changes in its environment, by establishing a connection with the organization's process maturity. Sixth chapter presents a proprietary model for automated assessment of feasibility of an IT project in a company. Seventh chapter presents the In-depth decomposition model, describing the methodology's four main steps: automatic generation of company's organizational structure using the Organizational Growth Algorithm, generation of process model, generation of functional requirements list, and assessment of various ERP systems across multiple knowledge bases. Validation of the entire methodology on real-life examples is shown in eighth chapter and conclusion with the directions for future research are given in ninth chapter.