Symmetry breaking, both spontaneous and explicit, will be approached using Ward identities as the main tool. We will first recall how Ward identities encode the spectrum of gapless, or nearly gapless modes (a.k.a. Goldstone bosons), emphasizing when UV or IR data is needed. We then proceed to consider the holographic perspective on this problem. We will discuss (a selection of) several models: a global U(1) symmetry, including the ambiguous case of two boundary dimensions; the simultaneous breaking of translations and scale invariance; supersymmetry breaking; models where Lorentz invariance is absent, such as in presence of a chemical potential or for theories with Lifshitz scaling. In all these cases we will first reproduce the expected Ward identities, and then try to go beyond and determine interesting IR data.
