The paper suggests an approach to improving the efficiency of visualization tools devel-oped for solving numerous applied problems. It is shown that one of the obstacles in this di-rection is the discrepancy between the choice of the method of visual representation of the analyzed data and the capabilities of the user using visual analytics tools. As a way to over-come the common errors of visual communication, the transition to adaptable visualization is proposed. The concept of adaptable visualization is based on the semiotic model of visualiza-tion and develops the concept of human-machine interaction in the direction of attracting the cognitive capabilities of users to solve visual analytics problems.