Contrary to a wide-spread belief that alkali metal (AM) atoms intercalated into layered materials form single-layer structures only, recent experiments [Nature 564 (2018) 234] showed that multi-layer configurations of lithium are possible in bi-layer graphene. Using state-of-the-art first-principles calculations, we systematically study the intercalation energetics for various AMs (Li, Na, K, Rb, Cs) in bi-layer graphene and MoS2. We demonstrate that for bi-layer graphene as host the formation energy of multi-layer structures is negative for K, Rb and Cs and only slightly positive for both Li and Na. In view of the previous experimental data on lithium, a multi-layer of Na might therefore form, while it is well-known that single-layers of Na in graphitic hosts are energetically very unfavorable. In MoS2, multi-layer structures are considerably higher in energy than the single-layer ones, but the formation of the former can still occur, especially for the AMs with the lowest electro-negativity. To rationalize the results, we assess the charge transfer from the intercalants to the host material and analyze the interplay between the ionic and covalent bonding of AM and host atoms. While our theoretical effort primarily focuses on the fundamental aspects of AM intercalation, our findings may stimulate experimental work addressing multi-layer intercalation to maximize the capacity of anode materials in AM ion batteries.