Journal of Advertising Research, 37 (2), 33-45.

Briggs, R. and N. Hollis, (1997).

Briggs and Hollis conducted a fairly early experiment into whether banner ads are effective independent of click-through rates.  They compared click-through to entering a car dealership the day after seeing a car advertisement: ideal, but not likely because of variations in audience motivations and intentions towards products.  This recognition of the limits of click-through comes even before rates had plummeted in the 2000's to below 1%.  Briggs and Hollis instead measured the effects a single banner ad exposure had on brand awareness by conducting a controlled experiment over the internet with a random sample of HotWired users.  Participants were directed to a web page with a survey that collected their demographic information, with a banner ad at the top.  The next day, participants who had completed the survey were asked to fill out another survey which collected data targeted to the ad they have viewed.  The results indicated that even a single exposure can have an effect on brand awareness.  In the most compelling of the brands tested, a men's apparel brand with almost no previous advertising exposure, brand awareness increased 200 percent.  As this demonstrates, banner ads are especially effective with brands which are relatively new, have minimal previous advertising, or for some other reason have low brand awareness initially.  As awareness increases due to exposure, returns diminish, though ads continue to have some effect.

This study provides an early assessment of the effects internet advertising can have on consumer attitudes towards specific companies independent of its effect on behaviors like click-through.  Despite the fact that it is a relatively old study and the less than ideal sample which could threaten generalizability, it is greatly cited by other authors, and therefore worth examining.  It depicts how advertisers may still achieve some of their goals, despite consumers' negative attitudes toward internet advertising as a whole.  If consumers see banner ads, even briefly and without strong interest in them, brand awareness may still increase, as seen further in Dreze and Hussherr.  Is it this finding which prompts the authors to state confidently, "Our results tell us, simply and unequivocally, that [banner advertising] works."  Considering the mixed results of later studies, this is likely an overstatement, but the claim that banner ads do have some effect independent of attention and positive affect still holds true.