Environmental challenges like climate change and ecosystems degradation are social dilemma of public goods contribution, plagued by free-riding. In this project we expand on the idea of environmental club goods or 'crowdaction' as a viable solution based on endogenous cooperation driven by social influence. We propose a discrete choice model in a population framework, named the Garden Model, which allows to analyse analytically and experimentally possible learning pattern towards cooperation. We compare different settings: one is an entry game, where the externality of social interactions can work as an endogenous threshold - critical mass - for the decision to join the Garden club. A second version confronts the players with a double stage decision: joining the club and how much to contribute. The latter version allows for a general setting where contributions can be negative, and describe exploitation of the public good. The different settings of the model entail positive feedback scenarios, with possible multiple equilibria and negative feedback scenarios, with possible periodic dynamics. The ultimate goal of the project is to evaluate aggregate contributions, measured as the Garden beauty. The setting with positive feedback can present the counterintuitive scenario where for positive shocks on marginal contribution benefits, individual contributions increase, while the resulting Garden beauty decreases. The model is designed for the different equilibrium outcomes to be tested in laboratory experiments.