An acquisition channel is any reproducible way a company brings new users in. Each channel has its own economics (CAC, payback, scalability), its own audience, and its own sensitivities. Most successful startups eventually depend on one or two dominant channels rather than a portfolio of weak ones.
Gabriel Weinberg's 'Bullseye Framework' suggests testing a wide set of channels cheaply, then doubling down on the one that produces the best CAC and reach. Channel saturation, platform risk, and competitive bidding are the three ways a working channel goes wrong.