Context. The accretion history of protostars remains widely mysterious, even though it represents one of the best ways to understand the protostellar collapse that leads to the formation of stars. Aims. Molecular outflows, which are easier to detect than the direct accretion onto the prostellar embryo, are here used to characterize the protostellar accretion phase in W43-MM1. Methods. The W43-MM1 protocluster hosts a sufficient number of protostars to statistically investigate molecular outflows in a single, homogeneous region. We used the CO(2-1) and SiO(5-4) line datacubes, taken as part of an ALMA mosaic with a 2000 AU resolution, to search for protostellar outflows, evaluate the influence that the environment has on these outflows' characteristics and put constraints on outflow variability in W43-MM1. Results. We discovered a rich cluster of 46 outflow lobes, driven by 27 protostars with masses of 1−100 M. The complex environment inside which these outflow lobes develop has a definite influence on their length, limiting the validity of using outflows' dynamical timescale as a proxy of the ejection timescale in clouds with high dynamics and varying conditions. We performed a detailed study of Position-Velocity diagrams of outflows that revealed clear events of episodic ejection. The time variability of W43-MM1 outflows is a general trend and is more generally observed than in nearby, low-to intermediate-mass star-forming regions. The typical timescale found between two ejecta, ∼500 yr, is consistent with that found in nearby protostars. Conclusions. If ejection episodicity reflects variability in the accretion process, either protostellar accretion is more variable, or episodicity is easier to detect in high-mass star-forming regions than in nearby clouds. The timescale found between accretion events could result from instabilities associated with bursts of inflowing gas arising from the close dynamical environment of high-mass star-forming cores.