The soil microbiome can increase crop resilience to both abiotic and biotic stress, and there is growing interest in uncovering the mechanisms by which we can shape plant associated microbiomes to increase crop yields within stressful environments. Through rhizodeposits, plants influence the composition of microbial communities and abiotic conditions in the rhizosphere, potentially generating plant-soil feedbacks which can increase nutrient availability and competitive ability against neighboring plants. Invasive plants have repeatedly been shown to drive plant-soil feedbacks that increase their ability to adapt to a wide range of environmental conditions and suppress neighboring plants. Using invasive plants as model species, we discuss what is currently known about the mechanisms that generate the plant-soil feedbacks that increase plant productivity, competitive ability, and resilience. Specifically, invasive potential is enhanced through 1) positive direct feedback loops which occur within a species, and can occur through enhanced mutualistic associations; or 2) negative indirect feedback loops, when feedbacks affect heterospecific plants through either allelopathy, disruption of mutualistic associations, and increased pathogen abundance. Knowledge on invasive plant exudate-microbe interactions may increase cropping system resilience through breeding superior crop genotypes, or potentially through soil amendments that disrupt weed-microbe interactions. We argue that because the soil biotic and native plant community often evolve in response to negative indirect feedback loops, focusing breeding efforts on positive direct plant-soil feedbacks, such as those that increase mutualistic associations and nutrient availability in the rhizosphere, are likely to lead to long-term stress resilient crops. Future research should explore to what extent upregulating production of specific exudates in non-invasive crop species generate the same plant-soil feedbacks responsible for invasive plant success.