The most important regions of a reactive potential energy surface are the transition state region, which determines the thermal rate coefficient, and the reactant and product asymptotes, which determine the equilibrium constant for the reaction. However, all realistic potential energy surfaces will also have reactant and product van der Waals wells.
For example, the following sketch shows the reaction profile of the Cl + HD to DCl + H reaction on its ground adiabatic potential energy surface. There is a shallow T-shaped van der Waals well in the reactant valley owing to the Cl(2P3/2)...HD quadrupole-quadrupole interaction, and an even shallower linear van der Waals well in the product valley caused by the dipole-induced dipole interaction between DCl and H:
Since these wells are so shallow (their depth is less than one tenth the height of the reaction barrier), one might not expect them to have a very big effect on the reaction dynamics, but they do (or at least the van der Waals forces that give rise to the wells and persist at smaller distances to affect the shape of the reaction barrier do).
The following figure compares calculated cross sections for the production of DCl and HCl from Cl + HD(j=0) on two different electronic potential energy surfaces with the experimental measurements of Kopin Liu. The two surfaces have similar transition state regions and reactant and product asymptotes, but they differ in one key respect: the BW surface has physically correct van der Waals wells whereas the G3 surface does not:
The conclusion is obvious. Although the van der Waals wells themselves are very shallow (and consequently often ignored in the construction of reactive potential energy surfaces), the forces that give rise to them can have a very dramatic effect on the outcome of a chemical reaction. The calculated DCl/HCl product branching ratio differs by almost an order of magnitude between the two surfaces, and it is clear from the comparison with the experimental measurements which of the two is correct.
For more details, including a simple explanation for why the van der Waals forces between Cl and HD have such a dramatic (steering) effect on the calculated product branching ratio, please see Science 286, 1713 (1999).