The paper analyses a three-stage game where two integrated system suppliers first decide whether to provide compatible system components, then choose the product specifications of their two offered components, and finally set prices. Contrary to results with exogenous product characteristics, the achievement of compatibility does not improve social welfare here. Given the assumed distribution of the consumers’ preferences on the product space, consumers’ losses from the changed product designs under compatibility outweigh the gains from these changes and from the availability of two additional system goods. Nevertheless, one firm ensures compatibility, if compatibility costs do not exceed a certain threshold.