From a cellular perspective, senescence has been considered a binary state, wherein cells are either senescent or not. This reductionist notion, often defined as irreversible growth arrest, has guided efforts to identify universal biomarkers and senolytics, but both have consistently eluded us. This outcome is not surprising, given that the biological nature of senescence may not be strictly irreversible; the accumulated evidence suggests that growth arrest can become unstable over time, with cells acquiring alterations, occasionally regaining proliferative capacity, or undergoing partial reprogramming, and exhibiting a heterogeneous spectrum of phenotypes ("senotypes") influenced by tissue types, stressors, temporal dynamics, and disease states. We propose that such a shift towards a dynamic spectrum of cellular states, is necessary to develop tailored strategies for context-specific signatures rather than a hypothetical state of cells that qualify for universal markers. The future of senescence research should thus focus on mapping, understanding, and utilising the spectrum of senescence states to mitigate its onset or modulate its progression.