In enhanced oil recovery, not only the low‐tension performance, but also the robustness at optimum formulation is an important issue. The fourth part of our review series is dedicated to robustness, defined as the width of the zone exhibiting three‐phase behavior around the optimum formulation, whatever the scanned variable. It is first corroborated from a screening of the available data in the literature that the tension minimum is inversely proportional to the square of the three‐phase range in the HLD scale. However, since there is still an inaccuracy of about a factor 10 in the tension minimum, some significant improvement can be attained in some cases by increasing the three‐phase behavior width in two ways. The first approach consists of finding systems that are insensitive to some formulation variable such as temperature, surfactant mixture composition or concentration, and water‐to‐oil ratio. The second way is to produce an artifact through which the optimum formulation is produced twice in a scan. If the distance between the two events in the scan is reduced down to be zero, their corresponding three‐phase behavior zones merge and result in a wider WIII region with a low tension. Several cases of such events are reported: alkaline scans, anionic‐nonionic and anionic‐cationic mixture changes, linear change in composition in three‐surfactant mixture, partial precipitation from a surfactant mixture in a salinity scan, and excessive partitioning of polyethoxylated nonionics. More complex transitions with three effects in a single scan or three concomitantly scanned variables show even more possibilities in practice.