Automated recommendation systems have been increasingly adopted by companies that aim to draw people attention about products and services on Internet. In this sense, development of distributed model abstractions such as MapReduce and GraphLab has brought new possibilities for recommendation research tasks due to allow us to perform Big Data analysis. Thus, this paper investigates the suitability of these two approaches for massive recommendation. In order to do so, the Alternating Least Squares (ALS), which is a Collaborative Filtering algorithm, has been tested using recommendation benchmark datasets. Results on RMSE show a preliminary comparative performance analysis.