Repetition of a message should increase how persuasive its effect is, at least it should according to research conducted by Zajonc (1968), who suggested this work via the mere exposure mechanism. People prefer stimuli (messages, faces, paintings etc) that they are more familiar with (have been more exposed to) and so by repeating the message, you expose them to it more and so increase their liking for it. One of Zajonc's (1968) experiments used made up 'Chinese' characters as stimuli, and asked participants to rate the 'goodness' of their meaning (how positive they thought the word the character represented would be). This was used as a measure of preference, assuming prefered words would be given higher ratings for 'goodness' of meaning. Zajonc (1968) predicted higher frequency characters (those participants had been exposed to more) would receive higher 'goodness' of meaning ratings, which they did find and are reflected in the table below.
The go compare advert's use of repetition does not increase positive feelings towards financial comparison websites or opera singers, a civilian casualty of this terrible advert. Rather the repetition becomes annoying and distracts from the actual message of the advert. Repetition in advertising has been found to be counterproductive in some situations as it leads to 'wear out' - the viewers become bored of the message (Appel, 1971) and this may account for why the go compare advert does not have the desired affect.
Appel, V. (1971). Advertising wear out. Journal of Advertising Research, 11, 11-13.
Zajonc, R. B. (1968). Attitudinal effects of mere exposure. Journal of Personality and Social Psychology, 9, 1-27.