In this paper, a decoding procedure is developed for both the compress-and-forward (CF) and the short message quantize-and-forward (QF-SM) relaying schemes. This procedure is based on exploiting a common feature of the mapping used by both schemes to determine the relay output from the estimate of its input. Using this decoding procedure, it is shown that the CF scheme achieves higher rates than both the conventional CF and the QF-SM schemes. The advantage of the proposed procedure is demonstrated in a Gaussian multicast one-message two-destination relay network and a multicast two-message three-destination network.