To Vary or Not? The Effects of Ad Variation on the Web By Sang Y ...

curve. Early studies indicate that attitudes toward the ad (Aad), attitudes toward the brand (Abrand), purchase intentions as well as positive cognitive responses peak at the third exposure (Cacioppo & Petty, 1979, 1980; Calder & Sternthal, 1980; Belch, 1982).

Contemporary researchers, on the other hand, argue that an inverted U-curve or S-shaped curve is rare in the real world and that the more frequently observed model is a concave model (Sissors & Bumba, 1996; Cannon & Riordon, 1996). According to this model, frequency effects experience diminishing returns after a certain point of exposure and reach a maximum point but the impact of frequency can be maintained if frequency continues. Thus, unlike the inverted-U model and S-shaped model, the concave model does not suggest a rapid decline in advertising effectiveness after the tedium point.

From a theoretical perspective, the frequency models discussed thus far are somewhat contradictory to encoding specificity and transfer-appropriate processing hypotheses which predict that people remember better when the same stimulus is repeated. This may be because these two hypotheses do not account for tedium which may occur with repeated exposure.

Frequency and Ad Variation

Recent research on the effects of ad variation is largely based on the fact that frequency eventually induces tedium. Specifically, one of the primary propositions of ad variation is that while people may experience tedium as a result of repeated exposures to the same ad, varied ads will reduce or delay tedium and thereby enhance advertising effectiveness (Haugtvedt, Schumann, Schneier, & Warren, 1994; Schumann, Petty, & Clemons, 1990; Schumann, & Clemons, 1989; Grass & Wallace, 1969). Furthermore, some researchers suggest that even in low frequency conditions where tedium is not expected (e.g., frequency 2 conditions), varied ads are more effective than non-varied ads (Unnava & Sirdeshmukh, 1994; Unnava & Burnkrant, 1991).