Image steganography algorithms hide the secret data in the cover image and give the stego image in the output. The visual quality is degraded due to data hiding. In the literature, the flipping algorithm is the most preferred data hiding algorithm to reduce the variability. In this algorithm, the secret data is flipped if the variability between the cover and stego image is greater than 50%. In our work, we have enhanced this algorithm and designed an improved flipping algorithm. In the proposed algorithm, the cover image is processed in blocks and each block variability is determined. If the variability of the block is higher than the threshold value then secret data is flipped and hide in the cover image. Thus, in the proposed algorithm, in some blocks, secret data is hidden in the original form and some blocks in flipped form. In the last, we have performed the experimental results on the standard dataset images and various parameters calculated. The result shows that the proposed algorithm enhances the visual quality of the cover image as compared to the existing algorithms. However, in the proposed algorithm need to communicate information with the receiver, the data is hidden in the original or flipped form.