First, what's for a Greek or a Roman a tribe?Initially, it's a political sub-divisions of their own peoples : phylai in Greece (especially in Athens), tribus in Rome. The existence of these tribes was understood as super-familial groups with a shared ancestry and a common founding ancestor. They weren't essential bodies as they were themselves divided, and they could be recomposed (Roman tribes could be removed or added following needs) and weren't communities in the modern sense of the word.Still, they formed an important part of the political life of classical cities for electoral and exchanges purposes, at least initially : but even when this purpose declined, it tended to remain part of a citizen's formal identity.

Thing is, the word is rarely used to describe Barbarian realities : most occurences of "tribes" in translations can be misleading, as it translate different concepts in one broad anthropologic meaning.For instance, Diodorus of Sicily only use it to describe Athenian history. Caesar doesn't use it to describe Gauls, Tacitus exclusively use "gens" to label Germanic people, etc. It can make texts more understable for a broad reading , but can easily be an obstacle to understand how they percieved these peoples or sub-peoples.Eventually, the broad use of tribe is more enlightening about the period translators used it than about classical's understanding of foreign peoples.

Of course, gens, civis, natio, etc. are, like "tribe", Greco-Romans concepts they familiar with used to decribe realities they're not the best they can. It doesn't help that their meaning evolved with time, and what could be described as a civitas/demos in the IIIrd century BCE might quite well not have been so two centuries later.But as we have essentially only Greco-Romans written sources avaible (with rare exceptions such as mention of teutas in Gaulish written dedications), it's really important to know the context and the original words being used in the text, because they're more precise in the authors' context.

I know it can be a frustrating answer, but there is not really a standardized understanding of foreign peoples in classical authorship, either in vocabulary or in concept (especially as classical anthropology tends to consider Barbarians as a-historic peoples compared to them, meaning they don't really hav an history and broadly are what they used to be, or remained in a state comparable to archaic history of civilized peoples).Nevertheless, comparisons are possible : when Caesar describe Gauls as being divided in cities, there's a similar vocabulary to the one used to describe Greek history.

It might be better to rather use "peoples" than "tribe" to label Barbarian groups when not really sure if they formed an independent group, part of a coalition or a sub-division of another people. Even classical peoples weren't really sure about what distinguished them, and in a Barbarian context, what could be considered as a tribe had of course slight different meaning (again with the Gaulish exemple, teuta could mean a citizenship as in τοουτιουϲ ναμαυσατις, tooùtious namausatis : tribesman/citizen of Nimes)

Basically, "tribe" is an extremely broad name which covers a lot of diverging concepts to the point being virtually a synonymous for any human group worth of mention in classical texts. It have some convenient use, for simplicity(s sake, but can't be a basis for understanding Barbarians' social realities.