The gamma distribution is commonly used in Bayesian models, but finding a good prior for the shape parameter can be tricky. Existing methods for this are slow or complicated, especially in models with many shape parameters. However, it has been discovered that the full conditional distribution of the gamma shape parameter can be approximated well by another gamma distribution. A new algorithm has been developed to quickly and easily find this approximation, which has been shown to be fast and accurate in various scenarios. This approximation can also be used as a proposal distribution for Metropolis-Hastings if exact results are needed.