Effective arranging of numerical data and design of associated computational algorithms are important for any area of mathematics for teaching, learning and research purposes. Usage of various algorithms for the same area makes mathematics teaching goal-oriented and diverse. Matrices and linear-algebraic ideas can be used to make algorithms visual, two dimensional (2D) and easy to use. It may contribute to the planned educational reforms by teaching school and university students deeper mathematical thinking. In this article we give novel data arranging techniques (2D and 3D) for matrix multiplication. Our 2D method differs from the standard, formal approach by using block matrices. We find this method a helpful alternative for introducing matrix multiplication. We also give a new innovative 3D visualisation technique for matrix multiplication. In this method, matrices are positioned on the faces of a rectangular cuboid. Computerized implementations of this method may be considered as student project proposals.