One of the crucial factors for optimal performance of FRP (Fiber-Reinforced Polymer) composites in shear and flexural strengthening is the consideration of FRP-concrete bond. The short length of shear reinforcements often makes it difficult to achieve the necessary bond length for surface adhesion. Therefore, methods like Near-Surface Mounted (NSM), which involve embedding FRP bars or strips into grooves near the concrete surface, are employed for this type of strengthening. Determining the required anchorage length for the NSM method necessitates not only experimental studies but also numerical investigations to expand the results. The current experimental research aims to investigate the effect of NSM bond length. For this purpose, twelve concrete block samples (150 x 150 x 300 mm) were subjected to pullout tests. The parameters examined in these tests include the distance of the FRP from the edge, the bond length, and the dimensions of the strip. Additionally, some samples featured a portion of the FRP bonded to the concrete as a fillet outside the groove to compensate for shallow penetration depth. The results from these "fillet" samples were analyzed separately and then compared with the other samples. During these experiments, two failure modes were observed: concrete failure and FRP slip. The results indicated that in the samples constructed with a fillet, the fillet significantly contributed to load bearing capacity. Under identical conditions, the fillet configuration demonstrated the ability to withstand 50% of the load compared to the fully embedded (NSM) condition.