The significance of the illocutionary component of speech acts for Habermas is not just that through them one can best elucidate the fact that utterances are “speech actions.” The significance of the performative component of speech acts is more central than this for Habermas. The significance of the illocutionary component for Habermas lies rather in the “peculiarly generative power of speech acts”:
“It is to this generative power that I trace the fact that a speech act can succeed or fail. . . .  the generative power (of a speech act) consists in the fact that the speaker, in performing a speech act, can influence the hearer in such a way that the latter can take up an interpersonal relation with her.” (OPC: 56-57, emphasis added).
The word “influence” in the above quote is absolutely crucial for Habermas. Habermas takes “influence” here in non causal sense. Influence for him here is rational and not causal. A part of Habermas’ project of theory of communicative action, formal pragmatics and theory of meaning is to show that this causality of reason can be explained without any reference to Kant’s otherworldly realm of intelligibility.
The above difference between rational influence and causal influence is the basis for Habermas distinction between 'perlocutionary effects' and 'illocutionary success,' (which in turn is the basis for Habermas' crucial distinction between communicative action and strategic action).
"Perlocutionary effects, like the successful results of teleological actions generally, may be described as states in the world brought about through intervention in the world. By contrast, illocutionary successes are achieved at the level of interpersonal relations on which participants in communication to come to an understanding with one another about something in the world. In this sense, they are not innerworldly but extramundane. At most, successful illocutionary acts occur within the lifeworld to which the participants in communication belong and that forms the background for their processes of reaching understanding. They cannot be intended under the description of causally produced effects" (OPC: 127).
The fact that the above crucial distinction is related to Habermas' effort to prove the Kantian notion of causality of reason from within normally goes unnoticed in Habermas scholarship. Charles Nussbaum is the only Habermas scholar I know of who has paid some attention to this.