This mean you need to enter another variables in the model to increase r2. In addition, the relationship between the Variable you enter in the model and the dependent Variable is weak.
Why might one obtain a low R-squared value for a linear model?
1. There actually is little or no association between your selected IVs and the chosen DV in the population.
2. One or more of your measures has poor technical quality (e.g., very low score reliability or validity), and therefore additional noise is being included in the data set, making it more difficult to discern relationships with accuracy.
3. Your model has omitted one or more salient variables from the IV list (e.g., the model is mis-specified). Note that mis-specification can be of two types: inclusion of non-associated variables, and omission of associated variables.
4. Your sample is atypical, or unrepresentative of the population.