Advertising relies on repetition because some repetition increases likability. 

While established brands can coast a little - you never see advertising for Dinty Moore Beef Stew because it is awesome - new brands rely on gaining popularity due to excessive exposure and repetition. 

When many people first saw the term "Google", they confused it with "Goober", the lazy man's peanut butter and jelly, for slackers who don't even care enough to spread them separately and that did not help likability, but repetition and excessive exposure made Google the top brand on the Internet, and Google has made Wikipedia a top brand on the Internet despite a name that annoyed people, again through excessive exposure and repetition even though it gets far more wrong than correct. 

Psychologists posit that the psychological mechanism behind this repetition effect is the easiness with which we perceive information. Repeatedly perceived information is easier to process for the brain, which saves capacity, and that 'feels' positive. 

Recent research by Sascha Topolinski and Fritz Strack from the Universität zu Köln (or University of Cologne, if you think the French just get to make everything French whenever they want) has concluded that this feeling of easiness and ensuing repetition effects actually stem from ... the mouth. 

They say that each time we encounter a name, the lips and the tongue automatically simulate the pronunciation of that name. This happens covertly, without our awareness and hopefully without actual mouth movements. During inner speech, the brain attempts to utter the novel name. When names are presented repeatedly, this articulation simulation is trained and thus runs more easily for repeated compared to novel names.

Crucially, if this inner speech is disturbed, for instance during chewing gum or whispering another word, the articulation of words cannot be trained and the repetition effect vanishes. People who are chewing something are immune to word repetition, they do not prefer familiar words over novel ones. 

They applied this to the real-world scenario of advertising in movie theaters. There, people usually consume popcorn and other snacks during watching commercials, which they believe disturbs the inner articulation of brand names. 

In two field studies, participants were invited to a real movie theater and were presented a block of commercials and a movie later on. The commercials were real commercials for existing products that were, however, foreign to the German participants, for instance, the Skandinavian butter LURPAK and a body lotion from INNISFREE.

Half of the participants received popcorn to eat freely during the cinema session. For them, the mouth was occupied with nibbling and chewing the popcorn so the mouth muscles could not engage in inner speech when watching the ads for the novel brands. The other half of the participants only received a small sugar cube at the beginning of the session so that they also had some sweet taste experience. The sugar cube, however, dissolved quickly in their mouth so that the mouth muscles were free to simulate the pronunciation of the brand names. 

In the first study, which involved 96 participants, the participants were invited to the lab one week after the cinema session. They were presented with images of products. Half of these products had been advertised in the cinema session, the other half were completely novel products. Participants were asked to indicate the products that thy likes, and their physiological responses were measured.

Those participants who had only received a sugar cube and could thus internally train the brands’ articulation demonstrated that there was a clear advertising effect. They preferred advertised over novel products and also showed positive physiological responses of familiarity for advertised products. However, those participants who had eaten popcorn while watching the commercials one week before showed no such advertising effect.

In a second study with another 188 participants, the popcorn procedure and commercial session was also carried out. But this time, participants were asked for real consumer choices one week later. They were given a small amount of money that they should spend on buying a skin lotion and donate to charity. Specifically, they were presented with six different lotions (with different brand names) and six charity foundations with fictitious names.

Three of the lotions and three of the charity foundations had been advertised in the earlier cinema session. Participants who had eaten a sugar chose the advertised products more often: they were more likely to buy the advertised lotions and donated their money for the advertised charities. However, the participants who had eaten popcorn did not show this effect.

Source: University of Koeln